Hi,
I think about using lisp for a new project (purely non-commericial).
There is a need for a GUI (Buttons, Listboxes, scrollbars, grahical
output, ...) and it should run under Win32 (NT / XP / 2000) and Linux.
My question: is there a free (or very cheap) lisp implementation which
either has a build-in GUI or has an easy interface to Qt, GTK or
something similar that runs under Linux and Win32.
(If not I have to do the programming in C++, but maybe I could run
some subporcesses in Lisp which will do the logical operations. Is
this possible ?)
May thanks and best regards,
Johann Murauer
········@acm.org
> * In message <·················@news.highway.telekom.at>
> * On the subject of "Free Lisp with GUI"
> * Sent on Wed, 16 Jan 2002 16:32:23 GMT
> * Honorable ········@acm.org (Johann Murauer) writes:
>
> (If not I have to do the programming in C++, but maybe I could run
> some subporcesses in Lisp which will do the logical operations. Is
> this possible ?)
you can use a browser as the front-end GUI.
--
Sam Steingold (http://www.podval.org/~sds)
Keep Jerusalem united! <http://www.onejerusalem.org/Petition.asp>
Read, think and remember! <http://www.iris.org.il> <http://www.memri.org/>
You can have it good, soon or cheap. Pick two...
The easiest way I've found to do this with Common Lisp (not being a
guru by any means) is with Tcl/Tk via the lisp2wish file of Matthias
Lindner (c): <http://ww.telent.net/cliki/Graphics%20Toolkit>. You can
use whatever implementation you want and all your program logic can
stay in Lisp.
I even made a little sourceforge project demonstrating it in action:
<https://sourceforge.net/projects/clslideshow/>, a slideshow program
for gif and jpeg files, but it only works in Linux/Unix right now
because I didn't bother using portable pathnames.
So that's a very easy way to do it that I don't think get's enough
mention, especially if you're on Linux where wish is commonly already
installed, but your project may of course have needs that require a
different solution.
-Jon
(switch "at" and first period to mail me)
········@acm.org (Johann Murauer) writes:
> Hi,
>
> I think about using lisp for a new project (purely non-commericial).
> There is a need for a GUI (Buttons, Listboxes, scrollbars, grahical
> output, ...) and it should run under Win32 (NT / XP / 2000) and
> Linux.
You could try LispWorks from Xanalys. IIRC is there cross-platform
GUI-stuff (CAPI) part of the free Version. It works on Windows and
Linux and according to their homepage other Unices too.
>
> My question: is there a free (or very cheap) lisp implementation which
> either has a build-in GUI or has an easy interface to Qt, GTK or
> something similar that runs under Linux and Win32.
CAPI, CLIM should do the job...
>
> (If not I have to do the programming in C++, but maybe I could run
> some subporcesses in Lisp which will do the logical operations. Is
> this possible ?)
Of course you can intermix languages how to do that depends on the
implementation you're using.
Regards
Friedrich
On 16 Jan 2002 17:55:06 +0100, Friedrich Dominicus
<·····@q-software-solutions.com> wrote:
>> I think about using lisp for a new project (purely non-commericial).
>> There is a need for a GUI (Buttons, Listboxes, scrollbars, grahical
>> output, ...) and it should run under Win32 (NT / XP / 2000) and
>> Linux.
>You could try LispWorks from Xanalys. IIRC is there cross-platform
>GUI-stuff (CAPI) part of the free Version. It works on Windows and
>Linux and according to their homepage other Unices too.
You should ask why they want a free lisp.
If they want something for delivering open source and free software
then both Franz and LispWorks are not really useful.
You generally distribute your code as both source and executables.
( Do you really want the user to download Xanalys and compile your
address book? )
Producing freely distributed executables with both is ( my
understanding ) highly problematic.
······@interaccess.com (Thaddeus L Olczyk) wrote in message news:<·················@nntp.interaccess.com>...
> Producing freely distributed executables with both is ( my
> understanding ) highly problematic.
For LispWorks at least this is not so for Windows and Linux: you can
generate freely distributable executables for both these platforms
from the commercial product (possibly with some constraints - there
may be stuff like CORBA or something you can't include). I think
there are royalties for the Unix platforms.
--tim
On 17 Jan 2002 01:43:31 -0800, ··········@tfeb.org (Tim Bradshaw)
wrote:
>······@interaccess.com (Thaddeus L Olczyk) wrote in message news:<·················@nntp.interaccess.com>...
>
>> Producing freely distributed executables with both is ( my
>> understanding ) highly problematic.
>
>For LispWorks at least this is not so for Windows and Linux: you can
>generate freely distributable executables for both these platforms
>from the commercial product (possibly with some constraints - there
>may be stuff like CORBA or something you can't include). I think
>there are royalties for the Unix platforms.
>
>--tim
Well we begin with the subject-- the person was asking for *free*
lisps. But lets just stop for a moment and assume that a person was
willing to pay something.
Remember this is stuff that the person would like to give away ( it
was clipped but I did say open source/free ). He wants it on both
Windows and Linux ( and possibly other UNICES ) . That's $1800
out of his pocket. Compare that to perl, python, ruby,gcc etc where
you pay nothing.
The upshot of this is that a person who wants to write something and
make it open source ( say he wants to write somethng that finds and
indexs all his mp3s, since he goes to that trouble to write it for
himself, he would like to make it open source ) is better off not
using Lisp. This leads to the idea that Lisp is not used much for
development.
Thaddeus L Olczyk wrote:
> Remember this is stuff that the person would like to give away ( it
> was clipped but I did say open source/free ). He wants it on both
> Windows and Linux ( and possibly other UNICES ) . That's $1800
> out of his pocket. Compare that to perl, python, ruby,gcc etc where
> you pay nothing.
>
> The upshot of this is that a person who wants to write something and
> make it open source ( say he wants to write somethng that finds and
> indexs all his mp3s, since he goes to that trouble to write it for
> himself, he would like to make it open source ) is better off not
> using Lisp. This leads to the idea that Lisp is not used much for
> development.
If the project is not time critical CormanLisp would be a good alternative. As far as I am informed CormanLisp is
Shareware and therefore affordable. But I am not sure how stable CormanLisp actually is.
Otherwise, I do not think that it is the fault (yes, yes you did not say it) of the vendors that there are not
more "sophisticated" Lisp implementations out there.
But I agree with you that Python/Perl/Ruby is lurking. But otherwise (exept for all the GUI stuff) Lisp has got a
good tradition for source-code only; if I can load a file (in the hope without any error messages -- which quite
often does not happen) and compile and use the functions everything will be okay. It is not comparable to a
stand-alone image but every day millions of people are doing actually this in: Matlab, IDL, Yorick,
Mathematica,...
S. Gonzi
······@interaccess.com (Thaddeus L Olczyk) writes:
> The upshot of this is that a person who wants to write something and
> make it open source ( say he wants to write somethng that finds and
> indexs all his mp3s, since he goes to that trouble to write it for
> himself, he would like to make it open source ) is better off not
> using Lisp. This leads to the idea that Lisp is not used much for
> development.
My feet! Lots of open-source software is written in Perl, Tcl/Tk,
Python, PHP or other scripting languages where the user of the program
will have to install the interpreter to be able to use it. You can do
the same thing with Common Lisp if you use one of the free (!)
implementations like CMUCL, SBCL, ECL, OpenMCL, or CLISP. With CLISP
alone you will be able to cover most Unix platforms (including Mac OS
X), Linux, Windows, and even Amiga OS IIRC.
Most Linux distributions will install Perl by default, but you'll
usually have to install something like Python or Tcl/Tk yourself,
probably from an RPM that comes with the distro. It's just as easy to
install CLISP nowadays, it's part of SuSE, Mandrake, Debian, and
probably a couple of other distros. On other platforms like, say,
Windows or Solaris, I can see no difference between Perl and CLISP:
You'll have to download and install it before you can use the fabulous
open-source program that was written in that language.
Now, if you complain about not being able to create stand-alone
executables with Perl, you either have to do it yourself or bother
Larry Wall until he does it. You'll probably have to pay someone to do
it for you. With Common Lisp that's easier: You can just buy LispWorks
or AllegroCL and create a stand-alone executable from the same source
code that ran with CLISP or CMUCL (provided it's ANSI compliant).
Edi.
> My feet! Lots of open-source software is written in Perl, Tcl/Tk,
> Python, PHP or other scripting languages where the user of the program
> will have to install the interpreter to be able to use it. You can do
> the same thing with Common Lisp if you use one of the free (!)
> implementations like CMUCL, SBCL, ECL, OpenMCL, or CLISP. With CLISP
> alone you will be able to cover most Unix platforms (including Mac OS
> X), Linux, Windows, and even Amiga OS IIRC.
To be fair, most Lisp apps are not very easy for somebody to install
if they don't know what they're doing. Lisp delivery of small apps
still is something of a problem.
joelh
Joel Ray Holveck <·····@juniper.net> writes:
> To be fair, most Lisp apps are not very easy for somebody to install
> if they don't know what they're doing. Lisp delivery of small apps
> still is something of a problem.
Unless, of course, they use debian. :)
Now, instead of others complaining that common-lisp-controller only
works in debian, I think it might be better if someone actually made
c-l-c packages for other systems. Preferably systems they actually
use. That way you don't get debian users trying to make RPM packages
when they only installed Redhat to make these packages.
To summarize, PLEASE help port common-lisp-controller to platforms
other than debian. Redhat, Windows, and FreeBSD would be a good start.
--
-> -/- - Rahul Jain - -\- <-
-> -\- http://linux.rice.edu/~rahul -=- ············@techie.com -/- <-
-> -/- "I never could get the hang of Thursdays." - HHGTTG by DNA -\- <-
|--|--------|--------------|----|-------------|------|---------|-----|-|
Version 11.423.999.221020101.23.50110101.042
(c)1996-2002, All rights reserved. Disclaimer available upon request.
Rahul Jain <·····@sid-1129.sid.rice.edu> writes:
> Joel Ray Holveck <·····@juniper.net> writes:
>
> > To be fair, most Lisp apps are not very easy for somebody to install
> > if they don't know what they're doing. Lisp delivery of small apps
> > still is something of a problem.
>
> Unless, of course, they use debian. :)
>
> Now, instead of others complaining that common-lisp-controller only
> works in debian, I think it might be better if someone actually made
> c-l-c packages for other systems. Preferably systems they actually
> use. That way you don't get debian users trying to make RPM packages
> when they only installed Redhat to make these packages.
>
> To summarize, PLEASE help port common-lisp-controller to platforms
> other than debian. Redhat, Windows, and FreeBSD would be a good
> start.
For a CL only solution you can look at CL.CONFIGURATION (pardon the
shameless plug). It works in any CL under any OS (or it makes
provisions to) and makes installing a new package as easy as
cl-prompt> (load "package.conf")
#p"package.conf"
cl-prompt> (conf:setup "THE-PACKAGE")
"THE-PACKAGE"
cl-prompt> (mk:load-system "THE-PACKAGE")
Also, the intention is not to be limited to MK:DEFSYSTEM for the last
line.
Cheers
--
Marco Antoniotti ========================================================
NYU Courant Bioinformatics Group tel. +1 - 212 - 998 3488
719 Broadway 12th Floor fax +1 - 212 - 995 4122
New York, NY 10003, USA http://bioinformatics.cat.nyu.edu
"Hello New York! We'll do what we can!"
Bill Murray in `Ghostbusters'.
CL.CONFIGURATION can be found in the CLOCC.
Cheers
--
Marco Antoniotti ========================================================
NYU Courant Bioinformatics Group tel. +1 - 212 - 998 3488
719 Broadway 12th Floor fax +1 - 212 - 995 4122
New York, NY 10003, USA http://bioinformatics.cat.nyu.edu
"Hello New York! We'll do what we can!"
Bill Murray in `Ghostbusters'.
In article <··············@photino.sid.rice.edu>, Rahul Jain <·····@sid-1129.sid.rice.edu> wrote:
> Joel Ray Holveck <·····@juniper.net> writes:
>
>> To be fair, most Lisp apps are not very easy for somebody to install
>> if they don't know what they're doing. Lisp delivery of small apps
>> still is something of a problem.
>
> Unless, of course, they use debian. :)
>
> Now, instead of others complaining that common-lisp-controller only
> works in debian, I think it might be better if someone actually made
> c-l-c packages for other systems. Preferably systems they actually
> use. That way you don't get debian users trying to make RPM packages
> when they only installed Redhat to make these packages.
>
> To summarize, PLEASE help port common-lisp-controller to platforms
> other than debian. Redhat, Windows, and FreeBSD would be a good start.
>
Hi-
What's the common-lisp-controller? What is it for?
Where is it in Debian (my distro)?
TIA
Regs,
synthespian
synthespian <···········@uol.com.br> writes:
> In article <··············@photino.sid.rice.edu>, Rahul Jain <·····@sid-1129.sid.rice.edu> wrote:
>> Joel Ray Holveck <·····@juniper.net> writes:
>>
>>> To be fair, most Lisp apps are not very easy for somebody to install
>>> if they don't know what they're doing. Lisp delivery of small apps
>>> still is something of a problem.
>>
>> Unless, of course, they use debian. :)
>>
>> Now, instead of others complaining that common-lisp-controller only
>> works in debian, I think it might be better if someone actually made
>> c-l-c packages for other systems. Preferably systems they actually
>> use. That way you don't get debian users trying to make RPM packages
>> when they only installed Redhat to make these packages.
>>
>> To summarize, PLEASE help port common-lisp-controller to platforms
>> other than debian. Redhat, Windows, and FreeBSD would be a good start.
>>
>
> Hi-
> What's the common-lisp-controller? What is it for?
C-L-C is, in essence, a system to allow Common Lisp programs to load
Common Lisp libraries, portably. For example, in order to load CLX,
you should be able to use `(require :clx)'.
In order to allow that, it includes DEFSYSTEM (load files in the
proper order with dependency checking) and MAKE:OOS (`OOS' means
`operate on system'). Libraries provide a system file describing their
files and the order they should be loaded in. MAKE:OOS operates on
libraries in various ways (e.g., loading or compiling them). Fex.
`(make:oos :foosystem :compile)' will compile the system FOOSYSTEM
(into a directory for compiled files from that implementation, so that
you can compile the same library with different implementations), and
`(make:oos :foosystem :load)' acts similarly to `(require
:foosystem)'.
> Where is it in Debian (my distro)?
C-L-C currently works well only on Debian, IIANM. Add these lines to
/etc/apt/sources.list:
deb http://cclan.sourceforge.net/debian cclan lisp
deb-src http://cclan.sourceforge.net/debian cclan lisp
and run `apt-get install common-lisp-controller' as root. After that,
you can read cCLan's package list (from their WWW site,
<URL:http://cclan.sf.net/>) and use apt-get to install the various
packages and automagically compile them for registered Common Lisp
implementations on your system.
> TIA
> Regs,
> synthespian
>
>
hth,
--
BPT <···@tunes.org> /"\ ASCII Ribbon Campaign
backronym for Linux: \ / No HTML or RTF in mail
Linux Is Not Unix X No MS-Word in mail
Meme plague ;) ---------> / \ Respect Open Standards
Brian P Templeton <···@tunes.org> writes:
> > What's the common-lisp-controller? What is it for?
> C-L-C is, in essence, a system to allow Common Lisp programs to load
> Common Lisp libraries, portably. For example, in order to load CLX,
> you should be able to use `(require :clx)'.
>
> In order to allow that, it includes DEFSYSTEM (load files in the
> proper order with dependency checking) and MAKE:OOS (`OOS' means
> `operate on system'). Libraries provide a system file describing their
> files and the order they should be loaded in. MAKE:OOS operates on
> libraries in various ways (e.g., loading or compiling them). Fex.
> `(make:oos :foosystem :compile)' will compile the system FOOSYSTEM
> (into a directory for compiled files from that implementation, so that
> you can compile the same library with different implementations), and
> `(make:oos :foosystem :load)' acts similarly to `(require
> :foosystem)'.
>
> > Where is it in Debian (my distro)?
> C-L-C currently works well only on Debian, IIANM. Add these lines to
> /etc/apt/sources.list:
>
> deb http://cclan.sourceforge.net/debian cclan lisp
> deb-src http://cclan.sourceforge.net/debian cclan lisp
It would indeed be a Rather Wonderful Thing if some *BSD folks took
sufficient interest to set up a "CLC" set of Ports.
The first few would probably be the most difficult; after a few got
put together, it would probably grow from there.
There's little doubt but that this would be Exceedingly Helpful in
encouraging greater re-use of CL code.
At the moment, it is pretty much typical for the basic _usable_
environment for the "libre/gratis" CL environments to represent The
Stuff I've Hacked Into My Own Environment.
CCLAN extends that quite nicely, but that's unfortunately only true so
long as you're using Debian. (Sort of like old Henry Ford; "You can
have any color you want so long as it's black...")
--
(concatenate 'string "cbbrowne" ·@ntlug.org")
http://www.ntlug.org/~cbbrowne/unix.html
"Robot: Mechanical apparatus designed to do the work of a man."
-- Encyclopedia Galactica
Christopher Browne <········@acm.org> writes:
> Brian P Templeton <···@tunes.org> writes:
>> > What's the common-lisp-controller? What is it for?
>> C-L-C is, in essence, a system to allow Common Lisp programs to load
>> Common Lisp libraries, portably. For example, in order to load CLX,
>> you should be able to use `(require :clx)'.
>>
>> In order to allow that, it includes DEFSYSTEM (load files in the
>> proper order with dependency checking) and MAKE:OOS (`OOS' means
>> `operate on system'). Libraries provide a system file describing their
>> files and the order they should be loaded in. MAKE:OOS operates on
>> libraries in various ways (e.g., loading or compiling them). Fex.
>> `(make:oos :foosystem :compile)' will compile the system FOOSYSTEM
>> (into a directory for compiled files from that implementation, so that
>> you can compile the same library with different implementations), and
>> `(make:oos :foosystem :load)' acts similarly to `(require
>> :foosystem)'.
>>
>> > Where is it in Debian (my distro)?
>> C-L-C currently works well only on Debian, IIANM. Add these lines to
>> /etc/apt/sources.list:
>>
>> deb http://cclan.sourceforge.net/debian cclan lisp
>> deb-src http://cclan.sourceforge.net/debian cclan lisp
>
> It would indeed be a Rather Wonderful Thing if some *BSD folks took
> sufficient interest to set up a "CLC" set of Ports.
>
> The first few would probably be the most difficult; after a few got
> put together, it would probably grow from there.
>
> There's little doubt but that this would be Exceedingly Helpful in
> encouraging greater re-use of CL code.
>
> At the moment, it is pretty much typical for the basic _usable_
> environment for the "libre/gratis" CL environments to represent The
> Stuff I've Hacked Into My Own Environment.
>
> CCLAN extends that quite nicely, but that's unfortunately only true so
> long as you're using Debian. (Sort of like old Henry Ford; "You can
> have any color you want so long as it's black...")
Heh. I think it's more like ``Yes, you have to have our logo on the
hatch of this tank, and yes, you only get two choices for the basic
engine type, but the hatch will be covered by the aircraft carrier
package that is installed by default...''. :P
> --
> (concatenate 'string "cbbrowne" ·@ntlug.org")
> http://www.ntlug.org/~cbbrowne/unix.html
> "Robot: Mechanical apparatus designed to do the work of a man."
> -- Encyclopedia Galactica
--
BPT <···@tunes.org> /"\ ASCII Ribbon Campaign
backronym for Linux: \ / No HTML or RTF in mail
Linux Is Not Unix X No MS-Word in mail
Meme plague ;) ---------> / \ Respect Open Standards
Brian P Templeton <···@tunes.org> writes:
> C-L-C currently works well only on Debian, IIANM. Add these lines to
> /etc/apt/sources.list:
>
> deb http://cclan.sourceforge.net/debian cclan lisp
> deb-src http://cclan.sourceforge.net/debian cclan lisp
Actually, it's in Potato now.
--
/|_ .-----------------------.
,' .\ / | No to Imperialist war |
,--' _,' | Wage class war! |
/ / `-----------------------'
( -. |
| ) |
(`-. '--.)
`. )----'
From: Christophe Rhodes
Subject: [OT] Debian distribution pedantry (was Re: Free Lisp with GUI)
Date:
Message-ID: <sqg04upzlq.fsf_-_@cam.ac.uk>
···@famine.OCF.Berkeley.EDU (Thomas F. Burdick) writes:
> Brian P Templeton <···@tunes.org> writes:
>
> > C-L-C currently works well only on Debian, IIANM. Add these lines to
> > /etc/apt/sources.list:
> >
> > deb http://cclan.sourceforge.net/debian cclan lisp
> > deb-src http://cclan.sourceforge.net/debian cclan lisp
>
> Actually, it's in Potato now.
No, it's not.
It *is* in the current "testing" distribution, codenamed "woody". It
is not in the current "stable" distribution, codenamed "potato". Even
if it were in potato, it wouldn't work terribly well, as the lisp
environments have to be common-lisp-controller--aware (essentially,
they have to believe in mk-defsystem and have a shell script written
that takes certain arguments such as "rebuild", "remove", and so on).
Christophe
--
Jesus College, Cambridge, CB5 8BL +44 1223 510 299
http://www-jcsu.jesus.cam.ac.uk/~csr21/ (defun pling-dollar
(str schar arg) (first (last +))) (make-dispatch-macro-character #\! t)
(set-dispatch-macro-character #\! #\$ #'pling-dollar)
Christophe Rhodes <·····@cam.ac.uk> writes:
> ···@famine.OCF.Berkeley.EDU (Thomas F. Burdick) writes:
>
> > Brian P Templeton <···@tunes.org> writes:
> >
> > > C-L-C currently works well only on Debian, IIANM. Add these lines to
> > > /etc/apt/sources.list:
> > >
> > > deb http://cclan.sourceforge.net/debian cclan lisp
> > > deb-src http://cclan.sourceforge.net/debian cclan lisp
> >
> > Actually, it's in Potato now.
>
> No, it's not.
*sigh* I shouldn't even bother with the names. I just point
everything to "testing" and once in a blue moon it changes names and I
don't realize...
--
/|_ .-----------------------.
,' .\ / | No to Imperialist war |
,--' _,' | Wage class war! |
/ / `-----------------------'
( -. |
| ) |
(`-. '--.)
`. )----'
>>>>> On 17 Jan 2002 19:00:50 -0800, Joel Ray Holveck ("Joel") writes:
Joel> To be fair, most Lisp apps are not very easy for somebody to install
Joel> if they don't know what they're doing. Lisp delivery of small apps
Joel> still is something of a problem.
The last time I wanted to deliver a small Lisp app to someone (on Windows),
I simply emailed them a zip file containing the exectuable. (The zip file
also included a source directory, for reference, but they didn't have any
Lisp compiler available to them or anything.) To run the app, they clicked
on the email attachment, ran Winzip on it, and then clicked on app.EXE file.
What's the problem?
Christopher Stacy <······@swingandcircle.com> writes:
> The last time I wanted to deliver a small Lisp app to someone (on Windows),
> I simply emailed them a zip file containing the exectuable. (The zip file
> also included a source directory, for reference, but they didn't have any
> Lisp compiler available to them or anything.) To run the app, they clicked
> on the email attachment, ran Winzip on it, and then clicked on app.EXE file.
>
> What's the problem?
The free CL implementations do not support creating small executables,
I think.
······@interaccess.com (Thaddeus L Olczyk) writes:
>
> Remember this is stuff that the person would like to give away ( it
> was clipped but I did say open source/free ). He wants it on both
> Windows and Linux ( and possibly other UNICES ) . That's $1800
> out of his pocket. Compare that to perl, python, ruby,gcc etc where
> you pay nothing.
Well he could use CMUCL couldn't he or CLISP or. Well he want to use a
cross-platform GUI-Toolkit. Well there is not much available for the
free Lisps, but it's part of the free verson of LispWorks. So he could
use the free version and distribute his code for free. If people want
to have it than they have to install LispWorks on their own. I can't
see why this should be more tedious than to install either python,
perl, ruby, gcc on Windows. Where's the difference? Why is it ok to
install everything else but a Lisp?
>
> The upshot of this is that a person who wants to write something and
> make it open source ( say he wants to write somethng that finds and
> indexs all his mp3s, since he goes to that trouble to write it for
> himself, he would like to make it open source ) is better off not
> using Lisp.
Nonsense, as pointed out below. You can distribute as much
code written by you as you like -- from any langauge. Well you own
VCC++ (fine you can write your code with it). Well you have to pay for
VCC++ too but obviously it is unimportant for most of the people, the
buy that stuff on Windows and are not using GCC. Well again why
shouldn't they do the same with Lisp?
Friedrich
······@interaccess.com (Thaddeus L Olczyk) writes:
>
> Remember this is stuff that the person would like to give away ( it
> was clipped but I did say open source/free ). He wants it on both
> Windows and Linux ( and possibly other UNICES ) . That's $1800
> out of his pocket. Compare that to perl, python, ruby,gcc etc where
> you pay nothing.
Well he could use CMUCL couldn't he? Or CLISP or. Well he want to use a
cross-platform GUI-Toolkit. Well there is not much available for the
free Lisps, but it's part of the free verson of LispWorks. So he could
use the free version and distribute his code for free. If people want
to have it than they have to install LispWorks on their own. I can't
see why this should be more tedious than to install either python,
perl, ruby, gcc on Windows. Where's the difference? Why is it ok to
install everything else but a Lisp?
>
> The upshot of this is that a person who wants to write something and
> make it open source ( say he wants to write somethng that finds and
> indexs all his mp3s, since he goes to that trouble to write it for
> himself, he would like to make it open source ) is better off not
> using Lisp.
Well I really dislike such comments You can distribute as much
code written by you as you like -- from any language. If you own
VCC++, you have paid for it but obviously it is unimportant for most
of the people, they
buy that stuff on Windows and are not using GCC. Well again why
shouldn't they do the same with Lisp? Why is installing all
stuff. According to you it's ok to install Python, Ruby. But why not
the personal edition from LispWorks? I guess CAPI is at least as good
as any of the cross-platform toolkits for Python. But hey what a deal
it's not free..
Friedrich
* Thaddeus L Olczyk wrote:
> Remember this is stuff that the person would like to give away ( it
> was clipped but I did say open source/free ). He wants it on both
> Windows and Linux ( and possibly other UNICES ) . That's $1800
> out of his pocket. Compare that to perl, python, ruby,gcc etc where
> you pay nothing.
I was just correcting the error in the post.
Tim Bradshaw wrote:
> * Thaddeus L Olczyk wrote:
>
> > Remember this is stuff that the person would like to give away ( it
> > was clipped but I did say open source/free ). He wants it on both
> > Windows and Linux ( and possibly other UNICES ) . That's $1800
> > out of his pocket. Compare that to perl, python, ruby,gcc etc where
> > you pay nothing.
>
> I was just correcting the error in the post.
Does anybody know what Allegro Common Lisp actually costs (single user
license)? On slashdot I read something like: 60.000.- USD*. The poster
then has been corrected to 6.000.- USD. But even this sounds a little bit
way too high? Maybe this are the prices for a server-license?
I think the LispWorks prices are okay.
S. Gonzi
[I do not have got the ID number of the post; but I red it a few days
ago]
Siegfried Gonzi <···············@kfunigraz.ac.at> wrote in message news:<·················@kfunigraz.ac.at>...
> Does anybody know what Allegro Common Lisp actually costs (single user
When comparing prices of brands of Lisp, I would look at other factors
besides just the up-front price. I would especially want to know
which of the following would violate the terms of the license:
1. Buy it for my home computer and install it on my laptop too.
2. Release shareware and/or freeware executables.
3. Do demonstration projects on my own time to try to convince my
employer to use Lisp for future projects. (Could that be construed
as sharing the license illegally with my employer?)
Does anyone know the answers to the above for the major commercial
brands of Lisp such as Lispworks and Allegro? Or could someone post
the license text so it could be analyzed in the forum?
I think lack of knowledge of such issues might be one factor in why
Lisp is not used more than it is. People probably assume the license
terms are draconian, because the major Lisp vendors have a reputation
of not being very grassroots friendly. That reputation might actually
be the single biggest factor in keeping Lisp relatively obscure,
regardless of how closely the reputation matches reality.
··········@mailandnews.com (Software Scavenger) writes:
> Siegfried Gonzi <···············@kfunigraz.ac.at> wrote in message news:<·················@kfunigraz.ac.at>...
>
> > Does anybody know what Allegro Common Lisp actually costs (single user
>
> When comparing prices of brands of Lisp, I would look at other factors
> besides just the up-front price. I would especially want to know
> which of the following would violate the terms of the license:
>
> 1. Buy it for my home computer and install it on my laptop too.
> 2. Release shareware and/or freeware executables.
> 3. Do demonstration projects on my own time to try to convince my
> employer to use Lisp for future projects. (Could that be construed
> as sharing the license illegally with my employer?)
>
> Does anyone know the answers to the above for the major commercial
> brands of Lisp such as Lispworks and Allegro? Or could someone post
> the license text so it could be analyzed in the forum?
>
You can get the license text for Xanalys Lispworks from the Personal
Edition. I'm not a lawyer and I do not speak for Xanalys but these are
quotes from the license text.
1. You may use the Software only on a single computer at a time.
2. Distribution of Runtimes. You may distribute Runtimes solely to
end-users as part of an application developed using the Software
("Application"), except that you may not distribute any part of the
Software as a general purpose Lisp development tool. Any Runtimes
distributed as part of the Application will continue to be subject to
the terms and conditions of this Agreement. You agree to license the
Application to your customers under a written license agreement
containing terms and conditions with regard to the Software and Runtimes
that are at least as restrictive as those contained herein.
For (3) I think you're OK as long as you abide by the rules stipulated
in 2.
--
Lieven Marchand <···@wyrd.be>
She says, "Honey, you're a Bastard of great proportion."
He says, "Darling, I plead guilty to that sin."
Cowboy Junkies -- A few simple words
Lieven Marchand <···@wyrd.be> writes:
> You can get the license text for Xanalys Lispworks from the Personal
> Edition. I'm not a lawyer and I do not speak for Xanalys but these are
> quotes from the license text.
>
> 2. Distribution of Runtimes. You may distribute Runtimes solely to
And note further that as of 4.2, runtimes no longer exist for the PC
editions of Lispworks (i.e. on Linux and Windows). They still apply
on "real" (ahem) unices.
--
It would be difficult to construe Larry Wall, in article
this as a feature. <·····················@netlabs.com>
In article <····························@posting.google.com>, ··········@mailandnews.com (Software Scavenger) wrote:
> Siegfried Gonzi <···············@kfunigraz.ac.at> wrote in message news:<·················@kfunigraz.ac.at>...
> People probably assume the license
> terms are draconian, because the major Lisp vendors have a reputation
> of not being very grassroots friendly. That reputation might actually
> be the single biggest factor in keeping Lisp relatively obscure,
> regardless of how closely the reputation matches reality.
Hi-
You bet!
And in c-l-l you get to read Lispers who are very confortable with the outrageous prices commercial lisps charge!
Price is a __key__ factor! For the student, for the mathematician in a developing nation, for the newbie (who one day may turn out to be a guru).
Seems to me a lot of lispers here don't even grasp the meaning and importance of "GPL".
And, with such a great language, born in 1958, it's no wonder it hasn't beaten the likes of C/C++...There doesn't seem to be a community that's really interested in open-source politics.
Regs,
synthespian
synthespian <···········@uol.com.br> writes:
>
> Hi-
>
> You bet!
> And in c-l-l you get to read Lispers who are very confortable with the outrageous prices commercial lisps charge!
> Price is a __key__ factor! For the student, for the mathematician in a developing nation, for the newbie (who one day may turn out to be a guru).
> Seems to me a lot of lispers here don't even grasp the meaning and importance of "GPL".
> And, with such a great language, born in 1958, it's no wonder it hasn't beaten the likes of C/C++...There doesn't seem to be a community that's really interested in open-source politics.
> Regs,
> synthespian
>
Bom dia,
You must be new to this newsgroup if you think there's no CL community
interested in open-source politics. Either that, or you haven't been
reading very closely. There are quite a few people interested in open
source CL, and you can get (at least) two fine implementations that
are open source: CMUCL and CLISP.
What you won't see in this newsgroup is consistent politics. Some
people here are radically pro-GPL, some are radically
pro-commercial-software. Most are somewhere in the middle, and they
have well thought-out reasons for their opinions.
If you're interested in open-source CL resources, though, you should
check out CLiki: http://ww.telent.net/cliki/index
Thom
--
(let ((e-mail-address ····@IXG.IUS")) (loop with new-string =
(make-string (length e-mail-address)) for count from 0 to (1- (length
e-mail-address)) for char-code = (char-code (aref e-mail-address
count)) for new-char-code = (if (and (> char-code 64) (< char-code
123)) (+ (mod (+ 13 char-code) 52) 65) char-code) do (setf (aref
new-string count) (code-char new-char-code)) finally (return
new-string)))
From: Brian P Templeton
Subject: Re: Cost of Lisp (was Re: Free Lisp with GUI)
Date:
Message-ID: <87r8oing7g.fsf@tunes.org>
Thom Goodsell <·········@DESPAM.cra.com> writes:
> synthespian <···········@uol.com.br> writes:
>>
>> Hi-
>>
>> You bet!
>> And in c-l-l you get to read Lispers who are very confortable with the outrageous prices commercial lisps charge!
>> Price is a __key__ factor! For the student, for the mathematician in a developing nation, for the newbie (who one day may turn out to be a guru).
>> Seems to me a lot of lispers here don't even grasp the meaning and importance of "GPL".
>> And, with such a great language, born in 1958, it's no wonder it hasn't beaten the likes of C/C++...There doesn't seem to be a community that's really interested in open-source politics.
>> Regs,
>> synthespian
>>
>
> Bom dia,
>
> You must be new to this newsgroup if you think there's no CL community
> interested in open-source politics. Either that, or you haven't been
> reading very closely. There are quite a few people interested in open
> source CL, and you can get (at least) two fine implementations that
> are open source: CMUCL and CLISP.
>
Don't forget SBCL :)
> What you won't see in this newsgroup is consistent politics. Some
> people here are radically pro-GPL, some are radically
> pro-commercial-software. Most are somewhere in the middle, and they
> have well thought-out reasons for their opinions.
>
I think some of the pro-free-software and pro-*proprietary*-software
(which is quite different from commercial software - commercial
software can be either free or proprietary) also have well-thought-out
reasons.
(when (reader-advocates-free-software-p)
(format t "~&...particularly the pro-free-software people.~%"))
;; :)
> If you're interested in open-source CL resources, though, you should
> check out CLiki: http://ww.telent.net/cliki/index
>
> Thom
>
> --
> (let ((e-mail-address ····@IXG.IUS")) (loop with new-string =
> (make-string (length e-mail-address)) for count from 0 to (1- (length
> e-mail-address)) for char-code = (char-code (aref e-mail-address
> count)) for new-char-code = (if (and (> char-code 64) (< char-code
> 123)) (+ (mod (+ 13 char-code) 52) 65) char-code) do (setf (aref
> new-string count) (code-char new-char-code)) finally (return
> new-string)))
--
BPT <···@tunes.org> /"\ ASCII Ribbon Campaign
backronym for Linux: \ / No HTML or RTF in mail
Linux Is Not Unix X No MS-Word in mail
Meme plague ;) ---------> / \ Respect Open Standards
Brian P Templeton <···@tunes.org> writes:
> >
> > You must be new to this newsgroup if you think there's no CL community
> > interested in open-source politics. Either that, or you haven't been
> > reading very closely. There are quite a few people interested in open
> > source CL, and you can get (at least) two fine implementations that
> > are open source: CMUCL and CLISP.
> >
> Don't forget SBCL :)
touche
> > What you won't see in this newsgroup is consistent politics. Some
> > people here are radically pro-GPL, some are radically
> > pro-commercial-software. Most are somewhere in the middle, and they
> > have well thought-out reasons for their opinions.
> >
> I think some of the pro-free-software and pro-*proprietary*-software
Yeah, I agonized over this choice. The thing is, I don't think anyone
in this group is really pro-*proprietary*-software, per se. I think
that people are pro-*commercial*-software and believe that, for some
classes of software, that prohibits releasing source code. As evidence
I'd cite the open source projects worked on by people such as Kent,
who are often seen as the "enemy" by the pro-free-software crowd.
> [both extremes] also have well-thought-out
> reasons.
Yes. I didn't mean to suggest otherwise, though it's my personal
belief that on most issues the best course lies somewhere in the
middle.
Thom
--
(let ((e-mail-address ····@IXG.IUS")) (loop with new-string =
(make-string (length e-mail-address)) for count from 0 to (1- (length
e-mail-address)) for char-code = (char-code (aref e-mail-address
count)) for new-char-code = (if (and (> char-code 64) (< char-code
123)) (+ (mod (+ 13 char-code) 52) 65) char-code) do (setf (aref
new-string count) (code-char new-char-code)) finally (return
new-string)))
From: Tim Bradshaw
Subject: Re: Cost of Lisp (was Re: Free Lisp with GUI)
Date:
Message-ID: <ey3vgdtyt3i.fsf@cley.com>
* Thom Goodsell wrote:
> Yeah, I agonized over this choice. The thing is, I don't think anyone
> in this group is really pro-*proprietary*-software, per se. I think
> that people are pro-*commercial*-software and believe that, for some
> classes of software, that prohibits releasing source code. As evidence
> I'd cite the open source projects worked on by people such as Kent,
> who are often seen as the "enemy" by the pro-free-software crowd.
I wouldn't mention this other than some of the more abusive articles
in this & related threads seemed to be aimed at me.
I don't really have an attitude on free (any sense) vs non-free (any
sense) software. I use both, and I've written both. I personally
enjoy contributing to open source software because I like seeing
people use and improve stuff I've made, and because I think of
programming as a kind of communication between people, not just a way
of ordering a dumb machine about. In fact at the moment I have a
reasonable chunk of Lisp stuff which I'm gradually working towards
making available under some OSS license. It isn't available yet
because I have to write a manual and generally tidy it up enough that
I wouldn't be ashamed of it, and finding time to do that is hard.
What I *do* have a very strong attitude towards is two things. (1) I
need to be able to eat, pay my bills and have a life, and other people
do as well. This means that, for instance, I simply don't at present
have an option to do a huge amount of OSS work, because I can't find a
way of getting paid to do it, and I'm old enough that I want to spend
at least some of my `free' time doing stuff that isn't sitting in
front of a terminal. (2) I want, ultimately, to escape from the
situation where I am paid for my time: I want to do something where I
can get paid repeatedly for a single finite amount of work. I want
this because, if I can achieve it, I can do a finite amount of work
which then frees up essentially unbounded time. During this time I
could work on OSS or do some of the many other things I want to do not
involving computers: basically I will have a private income and I can
do what I want.
(1) is empirically incompatible with OSS unless I can find someone who
will pay me to produce OSS. Anyone who wants to do so knows where to
contact me... At least one approach to (2) is commercial software as
well: write something and then get paid several times for it. There
are other approaches, like spotting a stockmarket bubble, starting a
company to exploit it and then selling the stock in time: I'm not
smart enough as well as too cynical to do that.
(1) I think is a reasonable goal for everyone. (2) probably brands me
as some rabid capitalist. Well, sorry, but I want to have time to
make beautiful photographs and gardens before I die more than I care
about that.
--tim
"Tim Bradshaw" <···@cley.com> wrote in message
····················@cley.com...
> (1) is empirically incompatible with OSS unless I can find someone who
> will pay me to produce OSS. Anyone who wants to do so knows where to
> contact me... At least one approach to (2) is commercial software as
> well: write something and then get paid several times for it. There
> are other approaches, like spotting a stockmarket bubble, starting a
> company to exploit it and then selling the stock in time: I'm not
> smart enough as well as too cynical to do that.
>
> (1) I think is a reasonable goal for everyone. (2) probably brands me
> as some rabid capitalist. Well, sorry, but I want to have time to
> make beautiful photographs and gardens before I die more than I care
> about that.
Writing high quality software takes loads of concentrated time. To free up
a person to do that I am perfectly willing to pay. Hence my support of the
LispWorks product by paying for it. I believe that a great implementation
of CL will help everyone and as part of that community, my part is to
support professional Lisp tool developers. Professional as in getting paid
to spend their undivided work time on getting the job done. It nice that
people are donating their time and effort for something like CMUCL, but
really I think that they are sacrificing themselves for those using it. For
me that is not right and I do not want other people to sacrifice themselves
for my benefit or even for some "higher goal" like CL or the GPL. If CL is
important enough it is worth recognizing those people as making a valuable
contribution to society and by paying them for their work and letting them
have balanced lives.
Wade
On Wed, 23 Jan 2002, Wade Humeniuk wrote:
[snip]
> Writing high quality software takes loads of concentrated time. To free up
> a person to do that I am perfectly willing to pay. Hence my support of the
> LispWorks product by paying for it. I believe that a great implementation
> of CL will help everyone and as part of that community, my part is to
> support professional Lisp tool developers.
Why are you sacrificing yourself for this goal! :)
But, if you're going to do it, let me suggest the following course of
action: Buying full licences of LispWorks for everyone who asks, starting
with me.
Look at the advantages:
1) Support the people who work on lispworks.
2) Supply the community with *fantastic* tools.
3) Get complaints from people that you should buy *them* Allegro.
It's a winner *all* the way around!
(Hmm. Buying folks Corman Lisp would be a cheaper test run of this idea
:))
> Professional as in getting paid
> to spend their undivided work time on getting the job done. It nice that
> people are donating their time and effort for something like CMUCL, but
> really I think that they are sacrificing themselves for those using it.
Er...first, sometimes they are the *very ones* who use it. I can't believe
you've been following this thread and didn't know this.
> For
> me that is not right and I do not want other people to sacrifice themselves
> for my benefit or even for some "higher goal" like CL or the GPL. If CL is
> important enough it is worth recognizing those people as making a valuable
> contribution to society and by paying them for their work and letting them
> have balanced lives.
What IS this? This is *just* as bad as the other tack.
Some people enjoy writing Lisp systems, or working on them, or
contributing to them. Not all of these folks want to be a commercial
developer of Lisp systems, or to work for one.
Sheesh, and what exactly is *wrong* with that.
This must be sarcasm on your part, or something.
Futhermore, if one is seriously committed to FSF style free software, that
motivation can be reasonably self regarding and compatible with a balanced
life. Ah, foo.
Look, people can screw up and exploit *any* system. I'm sure free software
situations have screwed people over. I'm sure commercial software
situtations have screwed people over. So?
The Common Lisp community has 4 or more propriety vendors, and 2 or more
Free implementations with reasonably active development. Why, exactly, is
this a *problem* from the point of view of a *lisper*? I can see why, if
you're primarily pro-Free software, why it's a problem (well, sorta). I
can can see why, if you're primarily an anti-Free software, why it's a
problem. But the brute fact of having this diversity seems an unalloyed
GOOD for lispers. (Ok, pace theories like Kent's, that too much free
software will hurt the community, or vice versa.)
Gah. Personally, I take diversity of implementations and the terms,
prices, licences, etc. attaching to them to be, in this day and age, to be
a *real* strength of a language and its community.
Cheers,
Bijan Parsia.
"Bijan Parsia" <·······@email.unc.edu> wrote in message
···············································@login8.isis.unc.edu...
> > For
> > me that is not right and I do not want other people to sacrifice
themselves
> > for my benefit or even for some "higher goal" like CL or the GPL. If CL
is
> > important enough it is worth recognizing those people as making a
valuable
> > contribution to society and by paying them for their work and letting
them
> > have balanced lives.
>
> What IS this? This is *just* as bad as the other tack.
>
> Some people enjoy writing Lisp systems, or working on them, or
> contributing to them. Not all of these folks want to be a commercial
> developer of Lisp systems, or to work for one.
>
> Sheesh, and what exactly is *wrong* with that.
>
There is nothing wrong, it is your choice, my comments were directed at
synthespian. Suggesting that we dump things like LispWorks for some higher
purpose (when synthepian is probably suggesting that people sacrifice their
lives for him and his ideas) is what I disagree with.
My comments are not meant to be applied to the general Lisp community but to
the developers of the LispWorks, Corman and Allegro implementions. They
deserve financial support for producing high quality products. To do that
they need blocks of time (undisturbed time) and the money so their attention
is undivided. It like having a Doctor moonlighting to make ends meet, since
he gives his services away to the sick and needy. Who wants that?
Wade
From: Brian P Templeton
Subject: Re: Cost of Lisp (was Re: Free Lisp with GUI)
Date:
Message-ID: <87vgdhetvx.fsf@tunes.org>
Tim Bradshaw <···@cley.com> writes:
> * Thom Goodsell wrote:
>> Yeah, I agonized over this choice. The thing is, I don't think anyone
>> in this group is really pro-*proprietary*-software, per se. I think
>> that people are pro-*commercial*-software and believe that, for some
>> classes of software, that prohibits releasing source code. As evidence
>> I'd cite the open source projects worked on by people such as Kent,
>> who are often seen as the "enemy" by the pro-free-software crowd.
>
> I wouldn't mention this other than some of the more abusive articles
> in this & related threads seemed to be aimed at me.
>
This is *not* aimed at you. (I don't think you meant that, but I'd
just like to make that clear.)
> I don't really have an attitude on free (any sense) vs non-free (any
> sense) software. I use both, and I've written both. I personally
> enjoy contributing to open source software because I like seeing
> people use and improve stuff I've made, and because I think of
> programming as a kind of communication between people, not just a way
> of ordering a dumb machine about. In fact at the moment I have a
> reasonable chunk of Lisp stuff which I'm gradually working towards
> making available under some OSS license. It isn't available yet
> because I have to write a manual and generally tidy it up enough that
> I wouldn't be ashamed of it, and finding time to do that is hard.
>
> What I *do* have a very strong attitude towards is two things. (1) I
> need to be able to eat, pay my bills and have a life, and other people
> do as well. This means that, for instance, I simply don't at present
> have an option to do a huge amount of OSS work, because I can't find a
> way of getting paid to do it, and I'm old enough that I want to spend
> at least some of my `free' time doing stuff that isn't sitting in
> front of a terminal. (2) I want, ultimately, to escape from the
> situation where I am paid for my time: I want to do something where I
> can get paid repeatedly for a single finite amount of work. I want
> this because, if I can achieve it, I can do a finite amount of work
> which then frees up essentially unbounded time. During this time I
> could work on OSS or do some of the many other things I want to do not
> involving computers: basically I will have a private income and I can
> do what I want.
>
> (1) is empirically incompatible with OSS unless I can find someone who
> will pay me to produce OSS. Anyone who wants to do so knows where to
> contact me...
Have you heard of `consulting'? It's what many free software
developers, including RMS, do.
> At least one approach to (2) is commercial software as well: write
> something and then get paid several times for it. There are other
> approaches, like spotting a stockmarket bubble, starting a company
> to exploit it and then selling the stock in time: I'm not smart
> enough as well as too cynical to do that.
>
Actually, (2) is perfectly compatible with free software, again under
the model presently called `consulting' by most people (well, some
forms of it, anyway). I like Andreas Bogk's analogy with bakeries:
Andreas Bogk, in <··············@teonanacatl.andreas.org>:
[...]
| In ordinary craftsmanship (a carpenter, a baker, or whatever), the
| rules are easy: work for a day, get paid for a day. Some people
| obviously believe they are so much more valuable than a baker that
| they expect to get paid for their one day of work for the rest of
| their lives. I strongly oppose this idea.
[...]
> (1) I think is a reasonable goal for everyone. (2) probably brands me
> as some rabid capitalist. Well, sorry, but I want to have time to
> make beautiful photographs and gardens before I die more than I care
> about that.
>
What? (2) sounds un-capitalistic to me...
> --tim
Best regards,
--
BPT <···@tunes.org> /"\ ASCII Ribbon Campaign
backronym for Linux: \ / No HTML or RTF in mail
Linux Is Not Unix X No MS-Word in mail
Meme plague ;) ---------> / \ Respect Open Standards
+ Thom Goodsell <·········@DESPAM.cra.com>:
| (let ((e-mail-address ····@IXG.IUS")) (loop with new-string =
| (make-string (length e-mail-address)) for count from 0 to (1- (length
| e-mail-address)) for char-code = (char-code (aref e-mail-address
| count)) for new-char-code = (if (and (> char-code 64) (< char-code
| 123)) (+ (mod (+ 13 char-code) 52) 65) char-code) do (setf (aref
| new-string count) (code-char new-char-code)) finally (return
| new-string)))
Since you had such a Lispy signature I could not help myself - I just
have to suggest this shorter and IMO much more easily understood
version:
(map 'string
#'(lambda (char)
(let ((char-code (char-code char)))
(code-char (if (and (> char-code 64)
(< char-code 123))
(+ (mod (+ 13 char-code) 52) 65)
char-code))))
····@IXG.IUS")
which formats into a nice three line brick for your .signature:
(map 'string #'(lambda (char) (let ((char-code (char-code char)))
(code-char (if (and (> char-code 64) (< char-code 123)) (+ (mod (+ 13
char-code) 52) 65) char-code)))) ····@IXG.IUS")
Anyway, I too used constructs like
(loop for count from 0 to (1- (length something)) ...)
until I discovered BELOW:
(loop for count from 0 below (length something) ...)
--
* Harald Hanche-Olsen <URL:http://www.math.ntnu.no/~hanche/>
- Yes it works in practice - but does it work in theory?
In article <···············@thoth.math.ntnu.no>, Harald Hanche-Olsen wrote:
> + Thom Goodsell <·········@DESPAM.cra.com>:
>
>| (let ((e-mail-address ····@IXG.IUS")) (loop with new-string =
>| (make-string (length e-mail-address)) for count from 0 to (1- (length
>| e-mail-address)) for char-code = (char-code (aref e-mail-address
>| count)) for new-char-code = (if (and (> char-code 64) (< char-code
>| 123)) (+ (mod (+ 13 char-code) 52) 65) char-code) do (setf (aref
>| new-string count) (code-char new-char-code)) finally (return
>| new-string)))
>
> Since you had such a Lispy signature I could not help myself - I just
> have to suggest this shorter and IMO much more easily understood
> version:
>
> (map 'string
> #'(lambda (char)
> (let ((char-code (char-code char)))
> (code-char (if (and (> char-code 64)
> (< char-code 123))
May I suggest (< 64 char-code 123) here?
> (+ (mod (+ 13 char-code) 52) 65)
> char-code))))
> ····@IXG.IUS")
>
> which formats into a nice three line brick for your .signature:
>
> (map 'string #'(lambda (char) (let ((char-code (char-code char)))
> (code-char (if (and (> char-code 64) (< char-code 123)) (+ (mod (+ 13
> char-code) 52) 65) char-code)))) ····@IXG.IUS")
>
> Anyway, I too used constructs like
>
> (loop for count from 0 to (1- (length something)) ...)
>
> until I discovered BELOW:
>
> (loop for count from 0 below (length something) ...)
Or even (loop for count below (length something) ...).
Regards,
--
Nils Goesche
"Don't ask for whom the <CTRL-G> tolls."
PGP key ID 0x42B32FC9
Nils Goesche <······@cartan.de> writes:
> In article <···············@thoth.math.ntnu.no>, Harald Hanche-Olsen wrote:
> >
> > Since you had such a Lispy signature I could not help myself - I just
> > have to suggest this shorter and IMO much more easily understood
> > version:
> >
> > (map 'string
> > #'(lambda (char)
> > (let ((char-code (char-code char)))
> > (code-char (if (and (> char-code 64)
> > (< char-code 123))
>
> May I suggest (< 64 char-code 123) here?
>
>
> > (+ (mod (+ 13 char-code) 52) 65)
> > char-code))))
> > ····@IXG.IUS")
Well, that's the last time I cut and paste from somebody else's
signature. The new, improved version can be found below. Thanks for
all the advice. I'm much happier with the length of this one.
Thom
--
(map 'string #'(lambda (char) (let ((char-code (char-code char)))
(code-char (if (< 64 char-code 123) (+ (mod (+ 13 char-code) 52) 65)
char-code)))) ····@IXG.IUS")
+ Nils Goesche <······@cartan.de>:
| > (code-char (if (and (> char-code 64)
| > (< char-code 123))
|
| May I suggest (< 64 char-code 123) here?
Indeed. Then, reuse a variable which is no more needed (replace a let
binding by a setf) and shorten the variable name to a single letter:
(map 'string
#'(lambda (c)
(setf c (char-code c))
(code-char (if (< 64 c 123)
(+ (mod (+ 13 c) 52) 65)
c)))
····@IXG.IUS")
and we're down to a nice two-liner:
(map 'string #'(lambda (c) (setf c (char-code c)) (code-char
(if (< 64 c 123) (+ (mod (+ 13 c) 52) 65) c))) ····@IXG.IUS")
or even
(map'string #'(lambda(c)(setf c(char-code c))(code-char
(if(< 64 c 123)(+(mod(+ 13 c)52)65)c)))····@IXG.IUS")
Oh, but this is getting silly really fast...
--
* Harald Hanche-Olsen <URL:http://www.math.ntnu.no/~hanche/>
- Yes it works in practice - but does it work in theory?
On Mon, 21 Jan 2002 13:49:47 -0300, synthespian <···········@uol.com.br>
wrote:
>In article <····························@posting.google.com>, ··········@mailandnews.com (Software Scavenger) wrote:
>> Siegfried Gonzi <···············@kfunigraz.ac.at> wrote in message news:<·················@kfunigraz.ac.at>...
>> People probably assume the license
>> terms are draconian, because the major Lisp vendors have a reputation
>> of not being very grassroots friendly. That reputation might actually
>> be the single biggest factor in keeping Lisp relatively obscure,
>> regardless of how closely the reputation matches reality.
>
> Hi-
>
> You bet!
> And in c-l-l you get to read Lispers who are very confortable with the outrageous prices commercial lisps charge!
Please go to Xanalys's site and tell me what's outrageous about their prices
or draconian about their runtime-free license. Same thing for MCL and Corman
Lisp.
> Price is a __key__ factor!
Certainly.
> Seems to me a lot of lispers here don't even grasp the meaning and importance of "GPL".
Ever heard of CMUCL, Clisp or SBCL?
There's only 1 vendor which has 'peculiar' prices and licenses. All other are
not only competitive, but unexpensive.
--
Fernando Rodr�guez
frr at wanadoo dot es
--
On Mon, 21 Jan 2002, synthespian wrote:
[snip]
> You bet!
> And in c-l-l you get to read Lispers who are very confortable
> with the outrageous prices commercial lisps charge!
Common Lisp prices are competitive with loads of other
systems. Proprietary Lisps range from free of charge to $200 to $900 to
who knows. Sounds standard. There are several free beer and Free
implementations.
> Price is a __key__ factor! For the student, for the
> mathematician in a developing nation, for the newbie (who one day may
> turn out to be a guru).
And for all of these, all of the Commercial implementations offer
options. None of these require a commercial license.
> Seems to me a lot of lispers here don't even grasp the meaning
> and importance of "GPL".
That doesn't seem true. You certainly don't grasp it *at all*, given your
emphasis on free beerness.
> And, with such a great language, born in 1958, it's no wonder it
> hasn't beaten the likes of C/C++...There doesn't seem to be a
> community that's really interested in open-source politics.
As opposed to that bastion of C/C++ and of freedom...Microsoft.
Don't think so.
I think the community is "interested", but, in contrast with you and your
experience, they are also *knowledgeable* and *thoughtful* and
*experienced*, thus there is a diversity of opinion and argument. I make
think that the positions and arguments I disagree with are *daft*, but
clearly, merely free beering everything isn't going to win any "wars"...as
if that's what mattered.
Cheers,
Bijan Parsia.
From: Henry Lebowzki
Subject: Re: Cost of Lisp (was Re: Free Lisp with GUI)
Date:
Message-ID: <3C4E4454.3020307@uol.com.br>
> I think the community is "interested", but, in contrast with you and your
> experience, they are also *knowledgeable* and *thoughtful* and
> *experienced*, thus there is a diversity of opinion and argument. I make
> think that the positions and arguments I disagree with are *daft*, but
> clearly, merely free beering everything isn't going to win any "wars"...as
> if that's what mattered.
>
> Cheers,
> Bijan Parsia.
Hi-
It's not about free beer.
More like feeding the hungry.
Teach them to catch their fish, but don't charge a thousand bucks for
the fishing rod.
Do you get it now?
Cheers,
h
Henry Lebowzki <···@uol.com.br> writes:
> > I think the community is "interested", but, in contrast with you and your
> > experience, they are also *knowledgeable* and *thoughtful* and
> > *experienced*, thus there is a diversity of opinion and argument. I make
> > think that the positions and arguments I disagree with are *daft*, but
> > clearly, merely free beering everything isn't going to win any "wars"...as
> > if that's what mattered.
>
> It's not about free beer.
> More like feeding the hungry.
> Teach them to catch their fish, but don't charge a thousand bucks for
> the fishing rod.
> Do you get it now?
>
Sure--we've gotten it all along, but no one is going to give out
thousand dollar fishing rods for free. After all, if they sell for a
thousand dollars, they probably cost $600 to make and $300 to distribute.
Besides, you can get a free/Free Lisp today, right now; if you want a
free fishing rod it's going to be a stick with a string tied to it.
Thom
--
(let ((e-mail-address ····@IXG.IUS")) (loop with new-string =
(make-string (length e-mail-address)) for count from 0 to (1- (length
e-mail-address)) for char-code = (char-code (aref e-mail-address
count)) for new-char-code = (if (and (> char-code 64) (< char-code
123)) (+ (mod (+ 13 char-code) 52) 65) char-code) do (setf (aref
new-string count) (code-char new-char-code)) finally (return
new-string)))
From: Brian P Templeton
Subject: Re: Cost of Lisp (was Re: Free Lisp with GUI)
Date:
Message-ID: <87hep78acj.fsf@tunes.org>
Thom Goodsell <·········@DESPAM.cra.com> writes:
> Henry Lebowzki <···@uol.com.br> writes:
>> > I think the community is "interested", but, in contrast with you and your
>> > experience, they are also *knowledgeable* and *thoughtful* and
>> > *experienced*, thus there is a diversity of opinion and argument. I make
>> > think that the positions and arguments I disagree with are *daft*, but
>> > clearly, merely free beering everything isn't going to win any "wars"...as
>> > if that's what mattered.
>>
>> It's not about free beer.
>> More like feeding the hungry.
>> Teach them to catch their fish, but don't charge a thousand bucks for
>> the fishing rod.
>> Do you get it now?
>>
>
> Sure--we've gotten it all along, but no one is going to give out
> thousand dollar fishing rods for free. After all, if they sell for a
> thousand dollars, they probably cost $600 to make and $300 to distribute.
>
> Besides, you can get a free/Free Lisp today, right now; if you want a
> free fishing rod it's going to be a stick with a string tied to it.
>
SBCL is a *fishing rod*?! [amazed] And all this time I'd been thinking
that it was some sort of *helicopter*!
:D
> Thom
>
> --
> (let ((e-mail-address ····@IXG.IUS")) (loop with new-string =
> (make-string (length e-mail-address)) for count from 0 to (1- (length
> e-mail-address)) for char-code = (char-code (aref e-mail-address
> count)) for new-char-code = (if (and (> char-code 64) (< char-code
> 123)) (+ (mod (+ 13 char-code) 52) 65) char-code) do (setf (aref
> new-string count) (code-char new-char-code)) finally (return
> new-string)))
--
BPT <···@tunes.org> /"\ ASCII Ribbon Campaign
backronym for Linux: \ / No HTML or RTF in mail
Linux Is Not Unix X No MS-Word in mail
Meme plague ;) ---------> / \ Respect Open Standards
Brian P Templeton <···@tunes.org> writes:
> SBCL is a *fishing rod*?! [amazed] And all this time I'd been thinking
> that it was some sort of *helicopter*!
Oh crap, I thought it was a bank! No wonder I don't have any money...
--
/|_ .-----------------------.
,' .\ / | No to Imperialist war |
,--' _,' | Wage class war! |
/ / `-----------------------'
( -. |
| ) |
(`-. '--.)
`. )----'
From: Brian P Templeton
Subject: Re: Cost of Lisp (was Re: Free Lisp with GUI)
Date:
Message-ID: <87d6zolf0u.fsf@tunes.org>
···@tornado.OCF.Berkeley.EDU (Thomas F. Burdick) writes:
> Brian P Templeton <···@tunes.org> writes:
>
>> SBCL is a *fishing rod*?! [amazed] And all this time I'd been thinking
>> that it was some sort of *helicopter*!
>
> Oh crap, I thought it was a bank! No wonder I don't have any money...
>
Now that I read the comments in the source as well as the manual, I
have discovered that is *is* a bank, too... board the helicopter and
go to the eleventy-first floor, corridor �-A-�-F-G (you'll need the
Unicode support, or the corridor will be hidden), room four. Solve the
maze (the undocumented SB-EXT:MAZE-SOLVER function is useful here),
and you'll be at the bank. The stack of manuals in one of the rooms on
floor 42 document the APPLY generic function's method for LOAN
objects. (Be sure to mixin the LOW-INTEREST class with the LOAN class
first; just don't mispell `LOAN' there, or you might accidentally make
yourself or another CL hacker, um, apathetic. It's not a security
flaw, it's flexiblity.) You may also be interested in the steel mine
on floor 174, but don't steel, steal^W^Wsteal, steal^W^Wsteal steel.
Unlike CMUCL, SBCL comes with some Jet Boots already in the source
distribution, so pulling it up from your bootstraps is quite easy, and
you don't need to be a wizard (though being a wizard is still helpful,
especially since you're justified in quickly becoming angry at the
SBCL developers for no perceptible reason).
:D
> --
> /|_ .-----------------------.
> ,' .\ / | No to Imperialist war |
> ,--' _,' | Wage class war! |
> / / `-----------------------'
> ( -. |
> | ) |
> (`-. '--.)
> `. )----'
--
BPT <···@tunes.org> /"\ ASCII Ribbon Campaign
backronym for Linux: \ / No HTML or RTF in mail
Linux Is Not Unix X No MS-Word in mail
Meme plague ;) ---------> / \ Respect Open Standards
In article <················@uol.com.br>, Henry Lebowzki wrote:
> It's not about free beer. More like feeding the hungry. Teach them
> to catch their fish, but don't charge a thousand bucks for the
> fishing rod. Do you get it now?
How patronizing. If they want a commercial Lisp so hard, they
obviously already know how to fish ;-)
Regards,
--
Nils Goesche
"Don't ask for whom the <CTRL-G> tolls."
PGP key ID 0x42B32FC9
Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
>
> Does anybody know what Allegro Common Lisp actually costs (single user
> license)? On slashdot I read something like: 60.000.- USD*. The poster
> then has been corrected to 6.000.- USD. But even this sounds a little bit
> way too high? Maybe this are the prices for a server-license?
Why don't you ask Franz directly. They know better than anyoine else.
>
> I think the LispWorks prices are okay.
Well than buy LispWorks ;-)
Regards
Friedrich
On Sat, 19 Jan 2002 10:14:47 +0100, Siegfried Gonzi
<···············@kfunigraz.ac.at> wrote:
>
>Does anybody know what Allegro Common Lisp actually costs (single user
>license)? On slashdot I read something like: 60.000.- USD*. The poster
>then has been corrected to 6.000.- USD. But even this sounds a little bit
>way too high? Maybe this are the prices for a server-license?
I could not find prices on the web page. To me this is extraordinary
dumb. ( The reason that I don't use QNX is that they don't display
prices. ) It indicates that the prices are so onerous that they don't
dare put them on the site, and that they want you to engage ( like
two bucks engage in battle and wind up with there horns all tangled
together, hard to split up ) a salesperson who will try to convince
you against your better judgement to shell out big bucks, or charge
what the market will bear.
I don't see how the situation helps them at all.
······@interaccess.com (Thaddeus L Olczyk) writes:
> On Sat, 19 Jan 2002 10:14:47 +0100, Siegfried Gonzi
> <···············@kfunigraz.ac.at> wrote:
> >Does anybody know what Allegro Common Lisp actually costs (single
> >user license)? On slashdot I read something like: 60.000.-
> >USD*. The poster then has been corrected to 6.000.- USD. But even
> >this sounds a little bit way too high? Maybe this are the prices
> >for a server-license?
> I could not find prices on the web page. To me this is extraordinary
> dumb. ( The reason that I don't use QNX is that they don't display
> prices. ) It indicates that the prices are so onerous that they
> don't dare put them on the site, and that they want you to engage (
> like two bucks engage in battle and wind up with there horns all
> tangled together, hard to split up ) a salesperson who will try to
> convince you against your better judgement to shell out big bucks,
> or charge what the market will bear.
> I don't see how the situation helps them at all.
It's a somewhat painful dilemma that is all too common.
-> If they publicize too-high prices on the web, that'll scare people
away that they might ultimately negotiate better pricing for.
(After all, while 1 copy might appropriately cost a bundle, an
organization buying 50 copies is worth dickering with.)
-> Does Franz want to deal with people not willing to pay a fairly
big price up front? Possibly not...
I seriously doubt that they are interested in the sort of dilution
of revenues that would result from pushing product through third
party channels. (Consider: It would be rather surprising if Red
Hat Software gets significantly more than $15 when a box with a $50
pricetag gets sold at CompUSA...)
Consider:
"Unless you're on the Forbes' richest 100 list, you're not a market,
just another photon in the rainbow."
-- Monty Brandenberg <······@ne.mediaone.net>
This certainly doesn't apply _directly_ to companies like Franz, but
it still is suggestive of the sound principle: If you're not dropping
a reasonably important chunk of change into their lap, you're not of
economic importance.
From some perspectives, $6K USD isn't very much. I believe that the
guys at Imperial Software, of "Motif GUI Builder" fame used to charge
around $50K _per user_ for their software. (No royalty fees after
that, mind you... THERE would be insult to add to injury!)
But in short it's not obvious that cutting prices would be in Franz's
interests. It's not obvious at all.
- If they have customers ready to balk at high prices, and go
elsewhere, _THAT_ would be a good reason to cut prices.
- If they figured that by cutting the price in 6, that they would
gain 6x as many customers, _THAT_ would be a good reason to cut
prices to 1/6 the present levels. (If the number is $6k, that would
drop it to $1K, which would _still_ be daunting to anybody sitting
at the low end of the "price preference" scale.) But it is not at
all obvious that this would happen.
I might _wish_ that I could buy ACL for $200; I might even hold
tenaciously to a refusal to pay more than that. For the time being,
that's liable to lead to them not selling me a copy of ACL, and my not
paying them $200 for it. :-).
--
(concatenate 'string "aa454" ·@freenet.carleton.ca")
http://www3.sympatico.ca/cbbrowne/spreadsheets.html
/"\
\ / ASCII RIBBON CAMPAIGN
X AGAINST HTML MAIL
/ \
>
> -> Does Franz want to deal with people not willing to pay a fairly
> big price up front? Possibly not...
>
> I seriously doubt that they are interested in the sort of dilution
> of revenues that would result from pushing product through third
> party channels. (Consider: It would be rather surprising if Red
> Hat Software gets significantly more than $15 when a box with a $50
> pricetag gets sold at CompUSA...)
>
When I sat down to buy a Linux based commercial Lisp, I tried Franz
first; mostly because I recognized the name from ages ago. I
contacted their sales people, who gave me a local distributor to
contact, which I did. I never heard from Franz, or the distributor
again.
So, after a couple weeks, I got annoyed and looked again at Lispworks.
Xanalys has prices on their webpage and we were able to just go ahead
and buy a copy of Lispworks with no fuss or game playing. Just
recently we purchased the 4.2 upgrade and if I ever get a budget
again, we'll buy a copy of Lispworks Enterprise.
Perhaps Franz has a business arrangement where dealing with
individuals isn't profitable and they get their money out of some few
huge customers, but I wonder how many sales they end up losing just
from making it hard for people to give them money.
Gregm
In article <··············@europa.pienet>, Greg Menke wrote:
>
>>
>> -> Does Franz want to deal with people not willing to pay a fairly
>> big price up front? Possibly not...
>>
>> I seriously doubt that they are interested in the sort of dilution
>> of revenues that would result from pushing product through third
>> party channels. (Consider: It would be rather surprising if Red
>> Hat Software gets significantly more than $15 when a box with a $50
>> pricetag gets sold at CompUSA...)
>>
>
>
> When I sat down to buy a Linux based commercial Lisp, I tried Franz
> first; mostly because I recognized the name from ages ago. I
> contacted their sales people, who gave me a local distributor to
> contact, which I did. I never heard from Franz, or the distributor
> again.
>
> So, after a couple weeks, I got annoyed and looked again at Lispworks.
> Xanalys has prices on their webpage and we were able to just go ahead
> and buy a copy of Lispworks with no fuss or game playing. Just
> recently we purchased the 4.2 upgrade and if I ever get a budget
> again, we'll buy a copy of Lispworks Enterprise.
>
> Perhaps Franz has a business arrangement where dealing with
> individuals isn't profitable and they get their money out of some few
> huge customers, but I wonder how many sales they end up losing just
> from making it hard for people to give them money.
>
> Gregm
Well I also had some oddities with Franz's pricing. I had a project to
do that I wanted to use to, among other things learn CL with. I
called Franz and asked them how much a developer license and about a
dozen runtimes would cost. I never got a quote and when I spoke to
them on the phone, it was a presales engineer if I remember correctly,
I got the distinct impression that Franz was in the "wife business"
and not the "whore business", perhaps that was a poor choice of words
let me explain:
Wife business = I am in all of your business and I get half of your
stuff if we break off our arrangement.
Whore business = I pay for service, you deliver the service and you
leave. If I want/need more service I pay for it. Basically a
professional relationship.
And there was no way I would get anything that looked like a wife
through my boss to go to the lawyers.
I have no issue with Franz, it is there product and they can do
what they want with it and that includes pricing.
marc
* Greg Menke wrote:
> Perhaps Franz has a business arrangement where dealing with
> individuals isn't profitable and they get their money out of some few
> huge customers, but I wonder how many sales they end up losing just
> from making it hard for people to give them money.
In Franz's defense (I am not a Franz customer), they did, some years
ago, sell a cheap (~ $1000 or maybe much less) Lisp for PCs (this was
pre linux being a major platform). I presume they stopped selling it
because it wasn't profitable. There are lots of models for selling
software, and not all of them are selling large numbers of licenses at
three-digit dollar prices. Oracle got rich selling licenses at probably
two orders of magnitude more than that (MS got even richer of
course selling enormous numbers of 3-digit licenses, which just goes
to show there are lots of places to live...)
--tim
> > Perhaps Franz has a business arrangement where dealing with
> > individuals isn't profitable and they get their money out of some few
> > huge customers, but I wonder how many sales they end up losing just
> > from making it hard for people to give them money.
>
> In Franz's defense (I am not a Franz customer), they did, some years
> ago, sell a cheap (~ $1000 or maybe much less) Lisp for PCs (this was
> pre linux being a major platform). I presume they stopped selling it
> because it wasn't profitable. There are lots of models for selling
> software, and not all of them are selling large numbers of licenses at
> three-digit dollar prices. Oracle got rich selling licenses at probably
> two orders of magnitude more than that (MS got even richer of
> course selling enormous numbers of 3-digit licenses, which just goes
> to show there are lots of places to live...)
I've heard on a couple occasions that the Franz licenses are ~$3000 US
or so- its a lot of money, but I would have been willing to plunk it
down for a good tool. Their business practices are certainly their
affair of course...
Gregm
"Greg Menke" <··········@mindspring.com> wrote in message
···················@europa.pienet...
>
> I've heard on a couple occasions that the Franz licenses are ~$3000 US
> or so- its a lot of money, but I would have been willing to plunk it
> down for a good tool. Their business practices are certainly their
> affair of course...
>
What makes Franz difficult to justify is their runtime fee policy. They wanted
a substantial percentage when my company made all the enquiries. It was way
off the mark for a development environment and application delivery.
--
Coby
(remove #\space "coby . beck @ opentechgroup . com")
On Sun, 20 Jan 2002, Coby Beck wrote:
> "Greg Menke" <··········@mindspring.com> wrote in message
> ···················@europa.pienet...
> >
> > I've heard on a couple occasions that the Franz licenses are ~$3000 US
> > or so- its a lot of money, but I would have been willing to plunk it
> > down for a good tool. Their business practices are certainly their
> > affair of course...
>
> What makes Franz difficult to justify is their runtime fee policy.
> They wanted a substantial percentage when my company made all the
> enquiries. It was way off the mark for a development environment and
> application delivery.
FWIW, this is true for VisualWorks, Cincom's flagship Smalltalk. They also
want to audit your books and other, to me, insane things :)
However, they did save VisualWorks and have a very nice Non commercial
version (download and go, no licence key or whatnot), and contribute lots
to the community (source code, conference sponsership, etc. etc.)
I'd rather them have their weird (to me) pricing and do all this other
stuff than there be no VisualWorks.
Cheers,
Bijan Parsia.
Bijan Parsia <·······@email.unc.edu> wrote in message news:<··········································@login8.isis.unc.edu>...
> FWIW, this is true for VisualWorks, Cincom's flagship Smalltalk. They also
> want to audit your books and other, to me, insane things :)
That's their business. Meanwhile Lispworks is a great product with
very good terms. Lisp is better than Smalltalk, so there's no real
reason to bother with Smalltalk at all. If Lispworks is not quite as
good as Franz Allegro technically, it's close enough that it doesn't
make any practical difference, especially if Franz's terms put Allegro
out of reach.
My advice to everyone facing such conflicts is to just get Lispworks
and get to work. If we want to use Lisp to reinvent the future, we
have to spend more time doing it and less time beating around the
bushes trying to decide what to do.
The real, important, fundamental, profound questions to ask yourself
are these:
1. Do you have Lispworks yet?
2. Have you used it yet, to reinvent the future?
3. If not, why not?
"Software Scavenger" <··········@mailandnews.com> wrote in message
·································@posting.google.com...
> Bijan Parsia <·······@email.unc.edu> wrote in message
news:<··········································@login8.isis.unc.edu>...
> My advice to everyone facing such conflicts is to just get Lispworks
> and get to work. If we want to use Lisp to reinvent the future, we
> have to spend more time doing it and less time beating around the
> bushes trying to decide what to do.
>
> The real, important, fundamental, profound questions to ask yourself
> are these:
>
> 1. Do you have Lispworks yet?
> 2. Have you used it yet, to reinvent the future?
> 3. If not, why not?
I agree. I have had LWW Professional for about 4 years. It is a perfectly
capable system. It has some short-comings, but what development environment
doesn't? I just accept its limitations and move on from there. It will get
there, but people have to use it.
Wade
"Wade Humeniuk" <········@cadvision.com> wrote in message ·················@news3.cadvision.com...
> I agree. I have had LWW Professional for about 4 years. It is a perfectly
> capable system. It has some short-comings, but what development environment
> doesn't? I just accept its limitations and move on from there. It will get
> there, but people have to use it.
>
> Wade
There is a "thread" running on the Xanalys "lisp-hug" email list where the Xanalys
people are asking for requests for the next release; they seem to be at the planning
stage with ears open. If you haven't made your suggestions known to them then
now might be a good time, and even if you have, now could be an opportune time
to remind them.
- Julian.
··········@mailandnews.com (Software Scavenger) writes:
> If Lispworks is not quite as good as Franz Allegro technically
my impression is that YMMV, that it's difficult to rank the two
products. They're both _very_ good systems, and both companies give
excellent support.
--
(espen)
··········@mailandnews.com (Software Scavenger) writes:
> Bijan Parsia <·······@email.unc.edu> wrote in message news:<··········································@login8.isis.unc.edu>...
>
> > FWIW, this is true for VisualWorks, Cincom's flagship Smalltalk. They also
> > want to audit your books and other, to me, insane things :)
>
> That's their business. Meanwhile Lispworks is a great product with
> very good terms. Lisp is better than Smalltalk, so there's no real
> reason to bother with Smalltalk at all.
Where on earth do you get the gall to say that? I certainly prefer
Lisp over SmallTalk, but if someone who has real experience with both
prefers to use ST, I'm not going to think they're foolish; I'm going
to think they want somewhat different things out of a language.
And for all I know, ST might work better than CL for projects larger
than what I've used ST for. Particularly because there are so very
many dialects. I still think that CL is a better fit for me, but I
don't think I typify the world.
--
/|_ .-----------------------.
,' .\ / | No to Imperialist war |
,--' _,' | Wage class war! |
/ / `-----------------------'
( -. |
| ) |
(`-. '--.)
`. )----'
On 21 Jan 2002, Software Scavenger wrote:
> Bijan Parsia <·······@email.unc.edu> wrote in message news:<··········································@login8.isis.unc.edu>...
>
> > FWIW, this is true for VisualWorks, Cincom's flagship Smalltalk. They also
> > want to audit your books and other, to me, insane things :)
>
> That's their business.
Sure, I'm just pointing out that Fraz isn't alone in this strategy.
> Meanwhile Lispworks is a great product with
> very good terms.
Sure, no one said otherwise.
> Lisp is better than Smalltalk, so there's no real
> reason to bother with Smalltalk at all.
Nice that you can turn this into a Stupid Language Flame. Thanks.
> If Lispworks is not quite as
> good as Franz Allegro technically, it's close enough that it doesn't
> make any practical difference,
Apparently it does, or LispWorks would drive Allegro out of business.
> especially if Franz's terms put Allegro
> out of reach.
Lispers are blessed with several good choices (just like
Smalltalkers). Several decent Free/free versions, etc. etc.
> My advice to everyone facing such conflicts is to just get Lispworks
> and get to work.
Why Lispworks instead of one of the cheaper alternatives? (Not a *real*
question, just pointing out that there are plenty of choices, and,
sometimes, the choices are more complex for some.)
> If we want to use Lisp to reinvent the future,
Er...how about inventing it first? :)
> we
> have to spend more time doing it and less time beating around the
> bushes trying to decide what to do.
Maybe. Learning from the experience of others seems *good*, but I forgot,
there's no reason to *look* elsewhere... :)
> The real, important, fundamental, profound questions to ask yourself
> are these:
[snip]
Sorry, they seem like silly questions to me. What are *your* answers to
them?
FWIW, as a complete side note, this type of note is a total turn off to
me. That is, I don't see any reason for *me* to support you or endeavors
you run/strongly participate in. If you had something solid to offset your
social clumsiness, that might do, but I don't see anything like
that. Pointers?
Cheers,
Bijan Parsia.
Bijan Parsia <·······@email.unc.edu> writes:
>
> > If Lispworks is not quite as
> > good as Franz Allegro technically, it's close enough that it doesn't
> > make any practical difference,
>
> Apparently it does, or LispWorks would drive Allegro out of business.
>
Well the difference is how the extensions are implemented. That's
quite a difference. If you have used heavily all fancy stuff from
Franz you probably won't find a counterpart in LispWorks. If you are
used to use one compiler you'll use conveneience features sooner or
later. Those may not be portable. And they got you later ;-)
Well there is always a balance to hold betwen things which are there
(or should be there) and things which are there but not covered by
anything like a Standard. It was pointed more than once what things
are different.
On the other hand if the price policy from Franz would be ridicolous,
people may consider changing. But this change could mean that they
change e.g to another language. It just depends thick you skin about
pricing is ;-)
Regards
Friedrich
On 22 Jan 2002, Friedrich Dominicus wrote:
> Bijan Parsia <·······@email.unc.edu> writes:
>
> >
> > > If Lispworks is not quite as
> > > good as Franz Allegro technically, it's close enough that it doesn't
> > > make any practical difference,
> >
> > Apparently it does, or LispWorks would drive Allegro out of business.
>
> Well the difference is how the extensions are implemented. That's
> quite a difference. If you have used heavily all fancy stuff from
> Franz you probably won't find a counterpart in LispWorks. If you are
> used to use one compiler you'll use conveneience features sooner or
> later. Those may not be portable. And they got you later ;-)
Sure. I was being a wee touch sarcastic :) That LispWorks hasn't taken
over the CL world indicates that inspite of the disparity of pricing,
indicates *something*.
Personally, I don't mind those with deep pockets paying from deep inside
those pockets :) As I pointed out, Cincom is a *great* Smalltalk community
citizen, in spite of their relatively bizarre pricing and terms. So *yay
them* for figuring out how to turn enough of a buck to be such great and
generous community members.
Franz does a fair bit of the same, afaict. AllegroServer comes to mind.
[snip]
> On the other hand if the price policy from Franz would be ridicolous,
> people may consider changing. But this change could mean that they
> change e.g to another language. It just depends thick you skin about
> pricing is ;-)
But I really *don't* understand the grip about Franz pricing *as a generic
Common Lisp user* (and, FWIW, I just became one, as the recent cl-xml
stuff inspired me to actually hack around on LispWorks :)). I don't have
to pay anything to use a variety of *great* common lisps for a *slew* of
different purposes. If I was a mathematics professor in a third world
country or (in the US) fourth rate university, and I bumped into the
limits of the personal edition of LispWorks, I would just ask them if
they'd donate a copy :) They may well. If not, and I couldn't scrape up
the cash, I'd go with one of the free ones (or, if I needed good windows
support, with Corman Lisp; if I couldn't afford the IDE, the compiler is
still free, after all).
Ideally, if I went with a free one, I'd have stuff to contribute back,
etc. etc.
So, I fail to see how some people being willing to deal with Franz as is
means that the "little people" are screwed :)
Not that *you* saw that either :)
Cheers,
Bijan Parsia.
Tim Bradshaw <···@cley.com> writes:
> Oracle got rich selling licenses at probably two orders of magnitude
> more than that
Yep, but you can download a _full_ version of Oracle to play with and
use it for development purposes without any restrictions as far as I
know. You'll only have to pay big bucks if you want to deploy your
application.
I'm not in a position to critize Franz or any other commercial Lisp
vendor, but I'm sure I'd love to have something similar from
them... :)
Edi.
···@agharta.de (Dr. Edmund Weitz) wrote in message news:<············@agharta.de>...
> Tim Bradshaw <···@cley.com> writes:
>
> > Oracle got rich selling licenses at probably two orders of magnitude
> > more than that
>
> Yep, but you can download a _full_ version of Oracle to play with and
> use it for development purposes without any restrictions as far as I
> know. You'll only have to pay big bucks if you want to deploy your
> application.
>
This is how, as a matter of fact, we got into Oracle Developer. We
were using something else before, but when we saw the writing on the
wall on that product, we downloaded the Oracle RDBMS and Oracle/2000.
That wouldn't have happened if we had to pay the big bucks upfront.
Oracle Developer then became our standard.
I think that the FULL Lisps commercial development systems should be
freely downloadable, but disable deployment. It's true both Franz and
XAnalys will send you the full enterprise versions for evaluation, if
you take that extra step, but that's an extra step. Lisp isn't always
the first alternative choice for many to want to take that extra
negotiation to *officially* evaluate the product. I feel obligated to
write-up reasons for rejection in some professional looking format
(many programmers don't want to do that kind of writing). So then the
next big version of the product comes along and I think maybe this is
it, but how can I now face the same salesperson and ask for the
enterprise edition?
rmc
On 21 Jan 2002 09:13:03 -0800, ·····@epcor.ca (Resty Cena) wrote:
> I think that the FULL Lisps commercial development systems should be
> freely downloadable, but disable deployment. It's true both Franz and
I seem to remember that Franz did something very similar to that for its
free ACL Linux versions up to 5.0 (the only significant missing feature was
CORBA support), but someone abused the license.
Paolo
--
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://web.mclink.it/amoroso/ency/README
[http://cvs2.cons.org:8000/cmucl/doc/EncyCMUCLopedia/]
Paolo Amoroso <·······@mclink.it> wrote in message news:<····························@4ax.com>...
> On 21 Jan 2002 09:13:03 -0800, ·····@epcor.ca (Resty Cena) wrote:
>
> > I think that the FULL Lisps commercial development systems should be
> > freely downloadable, but disable deployment. It's true both Franz and
>
> I seem to remember that Franz did something very similar to that for its
> free ACL Linux versions up to 5.0 (the only significant missing feature was
> CORBA support), but someone abused the license.
> Paolo
If the ability to deploy executables was disabled, I don't see how
this person can abuse the license -- other than hacking the code and
re-enabling compilation into an image file, in which case, what's
needed is a better way to secure the program.
·····@epcor.ca (Resty Cena) writes:
> Paolo Amoroso <·······@mclink.it> wrote in message news:<····························@4ax.com>...
> > On 21 Jan 2002 09:13:03 -0800, ·····@epcor.ca (Resty Cena) wrote:
> >
> > > I think that the FULL Lisps commercial development systems should be
> > > freely downloadable, but disable deployment. It's true both Franz and
> >
> > I seem to remember that Franz did something very similar to that for its
> > free ACL Linux versions up to 5.0 (the only significant missing feature was
> > CORBA support), but someone abused the license.
> > Paolo
>
> If the ability to deploy executables was disabled, I don't see how
> this person can abuse the license -- other than hacking the code and
Easily, and without any hacking, in lots of different ways. Note that
the image only disabled the ability to dump new images, it did include
the full compiler, including file compilation. So you can e.g.
a) Use ACL to develop your application, taking full advantage of all
the nice development tools, but deploy on, let's say, CMU CL.
b) Not "deploying" at all, since the application is a server
application, used on few hosts, which doesn't need the ability to
dump a new image.
c) And even in other cases you can simply load the compiled files at
start-up (using a little shell-script magic), instead of dumping a
new image.
...
[ Legal Disclaimer: I'm not inviting anyone to use any of the
outlined approaches to circumvent entered licensing agreements, or
violate applicaple copyright laws, nor would I condone any such
behaviour! Abide by the law, and all valid and applicable contracts
you entered into! ]
So there was plenty of room for abuse. Later editions, which included
heap limits did much to prevent the more glaring abuses, but of course
they also limited perfectly valid uses...
Regs, Pierre.
--
Pierre R. Mai <····@acm.org> http://www.pmsf.de/pmai/
The most likely way for the world to be destroyed, most experts agree,
is by accident. That's where we come in; we're computer professionals.
We cause accidents. -- Nathaniel Borenstein
"Pierre R. Mai" <····@acm.org> writes:
> ·····@epcor.ca (Resty Cena) writes:
>
> > Paolo Amoroso <·······@mclink.it> wrote in message news:<····························@4ax.com>...
> > > On 21 Jan 2002 09:13:03 -0800, ·····@epcor.ca (Resty Cena) wrote:
> > >
> > > > I think that the FULL Lisps commercial development systems should be
> > > > freely downloadable, but disable deployment. It's true both Franz and
> > >
> > > I seem to remember that Franz did something very similar to that for its
> > > free ACL Linux versions up to 5.0 (the only significant missing feature was
> > > CORBA support), but someone abused the license.
> > > Paolo
> >
> > If the ability to deploy executables was disabled, I don't see how
> > this person can abuse the license -- other than hacking the code and
>
> Easily, and without any hacking, in lots of different ways. Note that
> the image only disabled the ability to dump new images, it did include
> the full compiler, including file compilation. So you can e.g.
I think you may be remembering the Windows version. The original
Linux version we put out was fully operational, including the compiler
and the ability to dump new images. As for the lack of Corba (and CLIM)
availability, it was simply a case of us not having gotten to it yet.
It is possible that we would have charged for a CLIM, but that issue was
moot, since the state of the Motif toolkits (I think the only ones available
at the time were Lesstif and Moteeth) was not high enough quality to
port CLIM. And Orblink and the CL/Corba binding weren't available until
1999 (seems so long ago, doesn't it?) so that issue is also moot - we
wouldn't have provided a product on our free lisp that wasn't even yet
on our product list...
--
Duane Rettig Franz Inc. http://www.franz.com/ (www)
1995 University Ave Suite 275 Berkeley, CA 94704
Phone: (510) 548-3600; FAX: (510) 548-8253 ·····@Franz.COM (internet)
On Sun, 20 Jan 2002 18:11:00 +0100, ···@agharta.de (Dr. Edmund Weitz)
wrote:
>Tim Bradshaw <···@cley.com> writes:
>
>> Oracle got rich selling licenses at probably two orders of magnitude
>> more than that
>
>Yep, but you can download a _full_ version of Oracle to play with and
>use it for development purposes without any restrictions as far as I
>know. You'll only have to pay big bucks if you want to deploy your
>application.
>
>I'm not in a position to critize Franz or any other commercial Lisp
>vendor, but I'm sure I'd love to have something similar from
>them... :)
>
>Edi.
And even though you can download a version of Oracle, you need it less
then a version of Lisp. Every DB programmer knows the fundamentals of
Oracle ie SQL. It's just a matter of learning the platform specific
stuff, but many can get by just by learning how to pass queries, so
that's only one thing. ( On top of that most use ODBC, so they don't
have to learn much Oracle anyway. )
Consider a small program I just wrote. It descends a directory and
extracts all the filename extensions. Now it used car, cdr mapc and
a few other basic "lispisms". But it also uses regular expression
matching, string maniplation, file system examination. The funiest
was that I had no idea if there was a function that told me if
something was already in a list. ( I settled on using pushnew. )
The point is that someone who read Touretsky or Winston and Horn
might say that it looks like a Lisp program, but they would never have
been able to write it. For some stupid reason, the writers of Lisp
books do not believe Lisp programmers need to be able to write "real
applications". Why do I think this? Because they don't teach it.
So, Take someone who learned from these books. Further he's a
profesional programmer from other languages. Compare him to some guy
who has wants to learn another language. Being a proffesional
programmer, he feals reading the book puts him a long way toward
knowing the language. Especially since "he's been there before".
So now the programmer gets laid off and is looking for a new job.
Programmer1 says, "Even though I would like to get a job where I
can use Lisp. It would basically be entry level, and I would have to
get paid significantly less." Programmer2 says, "I would like to get
a job doing Language X. I've read a lot about it. Have played with it
at home a little bit. And now I'm ready to take it on in a job. Yes,
I'm not as profficient in it as I am in TheLanguageIUsedBefore, but
mmy general programming skills should carry the day. I won't have to
take a large paycut."
Paycut aside this is also the perspective of managers. When C ruled
and C++ started to take over, managers told prospective employees
who were profficient at C but not at C++, "Don't worry, if you know
C, then C++ is easy." When C++ ruled and Java took over managers said,
"Don't worry, if you know C++, then Java is easy."
They don't say things like this about Lisp. So before a person can
apply for a Lisp job, they are going to have to invest time in really
learning Lisp. And if the Lisp community wants people to acquire
skills in Lisp, they are going to have to supply a
"free-for-non-commercial-use ( including delivering open source
executables ) Lisp that runs on both Windows and Linux ( at least,
a generic Unix wold be better ) and comes with an industrial strength
GUI ( please don't say TK, but GTK 2.0 might be doable when it finally
comes out )."
I expect many here to say that I am out of touch with reality, that
commercial companies can't do this. ( BTW that is what they used to
say about Chicago Rapid Transit. They don't have the money to clean
it up and keep fares cheap. Eventually they manged to find the money.
Now a decaying system has been restored. ) But here is the reality.
1) Without such an implementation the base of Lisp programmers will
not expand very fast.
2) Without an expanding base of Lisp programmers, managers will be
hesitant to use Lisp.
3) Without an expanding base of usage, managers will worry about
vendors going out of business. So that mangers will be even less
likely to use it.
4) Without an expanding base of jobs, programmers will be less likely
to want to learn Lisp.
5) If this cycle keeps going then Lisp will die.
Oh yes. One more fact.
6) Supporters of other programming languages somehow manage to
survive without all the problems that Lisp interpretations have.
>>>>> On Mon, 21 Jan 2002 20:52:44 GMT, Thaddeus L Olczyk ("Thaddeus") writes:
Thaddeus> So before a person can apply for a Lisp job, they are going
Thaddeus> to have to invest time in really learning Lisp. And if the
Thaddeus> Lisp community wants people to acquire skills in Lisp, they
Thaddeus> are going to have to supply a "free-for-non-commercial-use
Thaddeus> ( including delivering open source executables ) Lisp that
Thaddeus> runs on both Windows and Linux ( at least, a generic Unix
Thaddeus> wold be better ) and comes with an industrial strength GUI
Thaddeus> ( please don't say TK, but GTK 2.0 might be doable when it
Thaddeus> finally comes out )."
As is repeated about four or five messages on this newsgroup,
and documented all over the place: you can get numerous free
implementations of Lisp, including full-blown development
environments that run on Windows and Unix with several GUI kits.
On Mon, 21 Jan 2002 21:47:02 GMT, Christopher Stacy
<······@swingandcircle.com> wrote:
>>>>>> On Mon, 21 Jan 2002 20:52:44 GMT, Thaddeus L Olczyk ("Thaddeus") writes:
>
> Thaddeus> So before a person can apply for a Lisp job, they are going
> Thaddeus> to have to invest time in really learning Lisp. And if the
> Thaddeus> Lisp community wants people to acquire skills in Lisp, they
> Thaddeus> are going to have to supply a "free-for-non-commercial-use
> Thaddeus> ( including delivering open source executables ) Lisp that
> Thaddeus> runs on both Windows and Linux ( at least, a generic Unix
> Thaddeus> wold be better ) and comes with an industrial strength GUI
> Thaddeus> ( please don't say TK, but GTK 2.0 might be doable when it
> Thaddeus> finally comes out )."
>
>As is repeated about four or five messages on this newsgroup,
>and documented all over the place: you can get numerous free
>implementations of Lisp, including full-blown development
>environments that run on Windows and Unix with several GUI kits.
There are no "free-for-non-commercial-use" ( including delivering open
source executables ) Lisp that run on both Windows and Linux and come
with an industrial strength GUI. So either get a clue or stop lie
about it which ever is the case.
The closest is CLisp but on windows the only viable GUI is TK which
I've already said I don't think of as a viable GUI. Also Clisp doesn't
support threads ( necessary because fork doesn't exist in Windows ).
A simple brief look at other languages shows that they do indeed have
such compilers/interpreters. C/C++ has gcc, with many many toolkits.
Java has Suns version (btw I know that JBuilder is free, I believe
Visual Age Forte and a few others are free too ). Smalltalk has
squeak. Even closer to home is DrScheme (allthough I can't judge the
GUI ). If you do include TK there is Tkklos.
In article <··················@nntp.interaccess.com>, Thaddeus L Olczyk wrote:
[WTF is wrong with you changing the subject line like that? Are you
nuts? Trolling?]
> On Mon, 21 Jan 2002 21:47:02 GMT, Christopher Stacy
><······@swingandcircle.com> wrote:
>
>>As is repeated about four or five messages on this newsgroup,
>>and documented all over the place: you can get numerous free
>>implementations of Lisp, including full-blown development
>>environments that run on Windows and Unix with several GUI kits.
> There are no "free-for-non-commercial-use" ( including delivering open
> source executables ) Lisp that run on both Windows and Linux and come
> with an industrial strength GUI. So either get a clue or stop lie
> about it which ever is the case.
Which C implementation that runs on both Windows and Linux comes
with an ``industrial strength GUI'', then?
> The closest is CLisp but on windows the only viable GUI is TK which
> I've already said I don't think of as a viable GUI.
That's your opinion. A friend of mine is working for a company
that ships an extremely expensive software product with a Tk GUI.
> Also Clisp doesn't support threads ( necessary because fork doesn't
> exist in Windows ).
Again, just arbitrary opinion. A former collegue once said ``Threads
are only for people who are too dumb to write state machines.'' The
product he is responsible for doesn't use threads at all. It is
a very large distributed system, runs on Windows, costs > 100000$ per
copy.
> A simple brief look at other languages shows that they do indeed have
> such compilers/interpreters. C/C++ has gcc, with many many toolkits.
You would seriously write a GUI application with gcc that is supposed
to run on Windows and Linux?
> Java has Suns version (btw I know that JBuilder is free, I believe
> Visual Age Forte and a few others are free too ).
``Write once, debug everywhere'' comes to mind ;-)
Regards,
--
Nils Goesche
"Don't ask for whom the <CTRL-G> tolls."
PGP key ID 0x42B32FC9
Nils Goesche <······@cartan.de> writes:
> > Also Clisp doesn't support threads ( necessary because fork doesn't
> > exist in Windows ).
[Okay, someone doesn't understand threads]
> Again, just arbitrary opinion. A former collegue once said ``Threads
> are only for people who are too dumb to write state machines.'' The
> product he is responsible for doesn't use threads at all. It is
> a very large distributed system, runs on Windows, costs > 100000$ per
> copy.
That's wonderful! I disagree, but it's still wonderful. Threads are
a great, portable abstraction to build *other* abstractions on top of.
Cooperative threads included. State machines, too. I wouldn't give
any of em up, but having read crazy wanna-be-state-machine code done
with pthreads, I know exactly where that comment's coming from.
On the CLISP tip, I'd like it if it had threads. But I'd like a lot
of things. My cake and eating it too, included. If I need
portability, CLISP is absolutely the best environment I know, no
exceptions. There are other fields where it can win or lose,
depending on other requirements. But if it needs to be widely
portable, I know of no better environment than CLISP. Squeak would be
the closest second, with no runners up.
--
/|_ .-----------------------.
,' .\ / | No to Imperialist war |
,--' _,' | Wage class war! |
/ / `-----------------------'
( -. |
| ) |
(`-. '--.)
`. )----'
In article <···············@conquest.OCF.Berkeley.EDU>, Thomas F. Burdick wrote:
> Threads are a great, portable abstraction to build *other*
> abstractions on top of. Cooperative threads included. State
> machines, too. I wouldn't give any of em up, but having read crazy
> wanna-be-state-machine code done with pthreads, I know exactly where
> that comment's coming from.
I guess. I am slowly getting back to threads, trying to think of
good uses for them, but it's hard. Do you have any hint for a book
or something where I could find examples for how to make /good/ use
of threads? I didn't find the examples in OReilly's pthread-book
very convincing.
Regards,
--
Nils Goesche
"Don't ask for whom the <CTRL-G> tolls."
PGP key ID 0x42B32FC9
From: Bulent Murtezaoglu
Subject: Re: changed back to Re: Free Lisp with GUI
Date:
Message-ID: <87sn8vmxh5.fsf@nkapi.internal>
>>>>> "NG" == Nils Goesche <······@cartan.de> writes:
NG> [...] I am slowly getting back to threads, trying to think
NG> of good uses for them, but it's hard.
I like threads when the computation is to be parallelized for execution
on multi-processors or the program does something nontrivial and time
consuming but still needs to be responsive to the GUI or the network and
such. In the former case there is a computational advantage, in the
latter you might need transparent preemption even on a single CPU. I
dislike using threads for state machine avoidance, but can see how they
may be appealing. For GUI work outside of wrapping heavy computations
in GUIs, event loops seem to work well in my limited experience.
NG> Do you have any hint
NG> for a book or something where I could find examples for how to
NG> make /good/ use of threads? [...]
I'd look for parallel programming textbooks for really convincing cases.
cheers,
BM
In article <··············@nkapi.internal>, Bulent Murtezaoglu wrote:
>>>>>> "NG" == Nils Goesche <······@cartan.de> writes:
>
> NG> [...] I am slowly getting back to threads, trying to think
> NG> of good uses for them, but it's hard.
>
> I like threads when the computation is to be parallelized for execution
> on multi-processors or the program does something nontrivial and time
> consuming but still needs to be responsive to the GUI or the network and
> such. In the former case there is a computational advantage,
Ok, that's obvious. I was looking for something different, Thomas
was talking about `abstractions' (I like abstractions :-). And I
only have one processor machines ;-)
> in the latter you might need transparent preemption even on a
> single CPU.
Hm. I used to program GUIs in the past (Win32 API), but using
threads for that always seemed messy to me, except for the very case
of really heavy computations you mention below. Threads, already
by their very existence, introduce a high degree of complexity,
so I always try /very/ hard before I give in to using them.
> I dislike using threads for state machine avoidance, but can see
> how they may be appealing. For GUI work outside of wrapping heavy
> computations in GUIs, event loops seem to work well in my limited
> experience.
Indeed.
Regards,
--
Nils Goesche
"Don't ask for whom the <CTRL-G> tolls."
PGP key ID 0x42B32FC9
Nils Goesche <······@cartan.de> writes:
>
> Ok, that's obvious. I was looking for something different, Thomas
> was talking about `abstractions' (I like abstractions :-). And I
> only have one processor machines ;-)
Programs which can be modeled as several components, each component
providing a service. Components request services of each other
by placing work requests on work queues. This can lead to simple,
clean code.
--
It would be difficult to construe Larry Wall, in article
this as a feature. <·····················@netlabs.com>
In article <··············@gondolin.local.net>, Alain Picard wrote:
> Nils Goesche <······@cartan.de> writes:
>
>> Ok, that's obvious. I was looking for something different, Thomas
>> was talking about `abstractions' (I like abstractions :-). And I
>> only have one processor machines ;-)
>
> Programs which can be modeled as several components, each component
> providing a service. Components request services of each other
> by placing work requests on work queues.
Interesting, I'll think about it; thanks! Although I'll get
nightmares of thousands of components doing nothing but filling
up each other's wait queues like state officials do :-)
Regards,
--
Nils Goesche
"Don't ask for whom the <CTRL-G> tolls."
PGP key ID 0x42B32FC9
In article <···············@ID-125440.news.dfncis.de>, Nils Goesche wrote:
> In article <··············@gondolin.local.net>, Alain Picard wrote:
>> Nils Goesche <······@cartan.de> writes:
>>
>>> Ok, that's obvious. I was looking for something different, Thomas
>>> was talking about `abstractions' (I like abstractions :-). And I
>>> only have one processor machines ;-)
>>
>> Programs which can be modeled as several components, each component
>> providing a service. Components request services of each other
>> by placing work requests on work queues.
>
> Interesting, I'll think about it; thanks! Although I'll get
> nightmares of thousands of components doing nothing but filling
> up each other's wait queues like state officials do :-)
Now that I've thought about it, it sounds a bit like System V Streams
drivers to me. Those run in kernel mode and have nothing to do with
threads, of course, but they're triggered by interrupts and system
calls which could also be thought of as kind of preemption. Maybe
it is possible to build an abstraction in user mode programs, multi-
threaded, which models somewhat System V Streams. Now /that/ would
be interesting!
Regards,
--
Nils Goesche
"Don't ask for whom the <CTRL-G> tolls."
PGP key ID 0x42B32FC9
Nils Goesche <······@cartan.de> writes:
> In article <···············@conquest.OCF.Berkeley.EDU>, Thomas F. Burdick wrote:
> > Threads are a great, portable abstraction to build *other*
> > abstractions on top of. Cooperative threads included. State
> > machines, too. I wouldn't give any of em up, but having read crazy
> > wanna-be-state-machine code done with pthreads, I know exactly where
> > that comment's coming from.
>
> I guess. I am slowly getting back to threads, trying to think of
> good uses for them, but it's hard. Do you have any hint for a book
> or something where I could find examples for how to make /good/ use
> of threads? I didn't find the examples in OReilly's pthread-book
> very convincing.
I actually don't think that threads are the right abstraction, but
they're close enough to get the job done. When I use threads, I
almost always really want scheduler activations. That's the first use
of them that pops to mind (probably because it's what I'm doing at the
moment). We have one thread per CPU, and have cooperative
multitasking going on within the threads. I don't really want to
discuss the specific abstractions we've built on top of this (sorry,
I'll try to come up with an example I can talk about :) -- though I'll
gladly point you towards the paper we're gonna write on it, when we
do.
One time that we *do* take advantage of the preemptive threads,
though, is for network activity, which needs to be dealt with as
quickly as possible. So to our application the world looks like two
or three CPUs that our tasks schedule eachother on, though sometimes
they get bumped for network traffic.
--
/|_ .-----------------------.
,' .\ / | No to Imperialist war |
,--' _,' | Wage class war! |
/ / `-----------------------'
( -. |
| ) |
(`-. '--.)
`. )----'
In article <···············@famine.OCF.Berkeley.EDU>, Thomas F. Burdick wrote:
> Nils Goesche <······@cartan.de> writes:
>
>> Do you have any hint for a book or something where I could find
>> examples for how to make /good/ use of threads?
>
> We have one thread per CPU, and have cooperative
> multitasking going on within the threads. I don't really want to
> discuss the specific abstractions we've built on top of this (sorry,
> I'll try to come up with an example I can talk about :) -- though I'll
> gladly point you towards the paper we're gonna write on it, when we
> do.
That would be nice, thank you.
> One time that we *do* take advantage of the preemptive threads,
> though, is for network activity, which needs to be dealt with as
> quickly as possible. So to our application the world looks like two
> or three CPUs that our tasks schedule eachother on, though sometimes
> they get bumped for network traffic.
Hm; unless you need much computing time for serving one client, I
think creating a new thread for each client is actually slower than
handling all clients in one (state machine) thread. Or do you mean
you have just one extra thread for network activity?
Regards,
--
Nils Goesche
"Don't ask for whom the <CTRL-G> tolls."
PGP key ID 0x42B32FC9
Nils Goesche <······@cartan.de> writes:
> In article <···············@famine.OCF.Berkeley.EDU>, Thomas F. Burdick wrote:
> > Nils Goesche <······@cartan.de> writes:
> >
> >> Do you have any hint for a book or something where I could find
> >> examples for how to make /good/ use of threads?
> >
> > We have one thread per CPU, and have cooperative
> > multitasking going on within the threads. I don't really want to
> > discuss the specific abstractions we've built on top of this (sorry,
> > I'll try to come up with an example I can talk about :) -- though I'll
> > gladly point you towards the paper we're gonna write on it, when we
> > do.
>
> That would be nice, thank you.
>
> > One time that we *do* take advantage of the preemptive threads,
> > though, is for network activity, which needs to be dealt with as
> > quickly as possible. So to our application the world looks like two
> > or three CPUs that our tasks schedule eachother on, though sometimes
> > they get bumped for network traffic.
>
> Hm; unless you need much computing time for serving one client, I
> think creating a new thread for each client is actually slower than
> handling all clients in one (state machine) thread. Or do you mean
> you have just one extra thread for network activity?
Yeah, we have one thread dedicated to network activity. It does
something state-machine-ish internally, and has the ability to spawn
another network thread (to use another CPU) if it starts to get
overwhelmed. Of course, we're not using POSIX threads (that would be
entirely too easy ... well, also we want to support Linux, and does
anyone trust pthreads on linux? maybe people who haven't used em...)
but something similar enough.
>
> Regards,
> --
> Nils Goesche
> "Don't ask for whom the <CTRL-G> tolls."
>
> PGP key ID 0x42B32FC9
--
/|_ .-----------------------.
,' .\ / | No to Imperialist war |
,--' _,' | Wage class war! |
/ / `-----------------------'
( -. |
| ) |
(`-. '--.)
`. )----'
······@interaccess.com (Thaddeus L Olczyk) writes:
Hey, the second entry in my killfile from c.l.l! I'm normally loathe
to killfile people (anyone who reads c.l.smalltalk knows TOPmind -- he
makes a point worth reading once in a *long* while, so I don't
killfile him), but this dude makes it. Welcome to your company with
that french dude trolling recently. I'm probably just bitter because
of Kent's and Erik's absence recently, but still...
--
/|_ .-----------------------.
,' .\ / | No to Imperialist war |
,--' _,' | Wage class war! |
/ / `-----------------------'
( -. |
| ) |
(`-. '--.)
`. )----'
From: Christopher Stacy
Subject: Re: Christopher Stacy: Liar or just plain stupid ( Re: Free Lisp with GUI
Date:
Message-ID: <u7kqa8s5m.fsf@swingandcircle.com>
I'm outta here.
I'll check back in to comp.lang.lisp in three months and see if things are better then.
[I'm following up to Christopher's article because Thaddeus' has not shown
up in my news feed, but my comments are actually related to Thaddeus'
article; apologies to Christopher - Paolo]
On Mon, 21 Jan 2002 21:47:02 GMT, Christopher Stacy
<······@swingandcircle.com> wrote:
> >>>>> On Mon, 21 Jan 2002 20:52:44 GMT, Thaddeus L Olczyk ("Thaddeus") writes:
>
> Thaddeus> So before a person can apply for a Lisp job, they are going
> Thaddeus> to have to invest time in really learning Lisp. And if the
> Thaddeus> Lisp community wants people to acquire skills in Lisp, they
> Thaddeus> are going to have to supply a "free-for-non-commercial-use
> Thaddeus> ( including delivering open source executables ) Lisp that
Why such a system has to also support application delivery in order to be
useful as a Lisp _learning_ environment? Incidentally, there are
open-source Lisp implementations, most notably CLISP and CMU CL, that can
do that even if this feature is often not explicitly mentioned in the
documentation (but the CLISP implementation notes do provide information on
this).
Paolo
--
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://web.mclink.it/amoroso/ency/README
[http://cvs2.cons.org:8000/cmucl/doc/EncyCMUCLopedia/]
* Paolo Amoroso wrote:
> Why such a system has to also support application delivery in order to be
> useful as a Lisp _learning_ environment?
This is easy: the desired system must have a set of features that no
existing system does, in order to correctly support SUW (that's
Standard Usenet Whining). New requirements for the system will spring
spontaneously into existence as systems are demonstrated which satisfy
the previous set.
It is left as an exercise for the reader to demonstrate the infinite
nature of the system required.
--tim
Tim Bradshaw wrote:
> * Paolo Amoroso wrote:
>
>
>> Why such a system has to also support application delivery in order to be
>> useful as a Lisp _learning_ environment?
>
>
> This is easy: the desired system must have a set of features that no
> existing system does, in order to correctly support SUW (that's
> Standard Usenet Whining). New requirements for the system will spring
> spontaneously into existence as systems are demonstrated which satisfy
> the previous set.
>
> It is left as an exercise for the reader to demonstrate the infinite
> nature of the system required.
>
> --tim
Ain't it funny how you don't see the contempt with which you guys regard
free software (and free in _all_ it's meanings - free and FULLY FUNCTIONAL).
If only the C++, Python or Perl community were like that...I'm glad they
aren't.
I'm new to LISP, I am very enthusiastic about it, but c.l.l. looks like
it's full of grey-haired engineers (nothing wrong with that per se)
making big bucks on some software company, who couldn't care less if the
free version has the same cool GUI then the proprietary LISP.
Regs,
H
Henry Lebowzki <···@uol.com.br> writes:
> Ain't it funny how you don't see the contempt with which you guys
> regard free software (and free in _all_ it's meanings - free and FULLY
> FUNCTIONAL).
>
> If only the C++, Python or Perl community were like that...I'm glad
> they aren't.
They ARE.
> I'm new to LISP, I am very enthusiastic about it, but c.l.l. looks
> like it's full of grey-haired engineers (nothing wrong with that per
> se) making big bucks on some software company, who couldn't care less
> if the free version has the same cool GUI then the proprietary LISP.
Feel free to tell us about the C++, Python, and Perl GUI builder
systems that are freely available and which run on every conceivable
platform.
Of course REALITY is otherwise.
Visual Drek costs quite big dollars (less than Franz Lisp; sometimes
more than LW), and only works with Windows. And there are other C++
compilers from various vendors at various kinds of pricing.
GCC is free, but (remarkably similar to CLISP and CMU/CL) it doesn't
have an integrated GUI builder.
Python and Perl are free, just like CLISP and CMU/CL, and there are
similar sorts of limited offerings of GUI builders for them all that
tend to be pretty platform-specific.
Somehow, when you try to point at the things you _want_ to criticize
about the Lisp community, the result is pretty incriminating against
the languages you suggest comparing to...
--
(reverse (concatenate 'string ········@" "enworbbc"))
http://www.ntlug.org/~cbbrowne/emacs.html
"Normally, we don't do people's homework around here, but Venice is a
very beautiful city, so I'll make a small exception."
--- Robert Redelmeier compromises his principles
Christopher Browne <········@acm.org> writes:
> Henry Lebowzki <···@uol.com.br> writes:
>> Ain't it funny how you don't see the contempt with which you guys
>> regard free software (and free in _all_ it's meanings - free and FULLY
>> FUNCTIONAL).
>>
>> If only the C++, Python or Perl community were like that...I'm glad
>> they aren't.
>
> They ARE.
>
>> I'm new to LISP, I am very enthusiastic about it, but c.l.l. looks
>> like it's full of grey-haired engineers (nothing wrong with that per
>> se) making big bucks on some software company, who couldn't care less
>> if the free version has the same cool GUI then the proprietary LISP.
>
> Feel free to tell us about the C++, Python, and Perl GUI builder
> systems that are freely available and which run on every conceivable
> platform.
>
> Of course REALITY is otherwise.
>
> Visual Drek costs quite big dollars (less than Franz Lisp; sometimes
> more than LW), and only works with Windows. And there are other C++
> compilers from various vendors at various kinds of pricing.
>
> GCC is free, but (remarkably similar to CLISP and CMU/CL) it doesn't
> have an integrated GUI builder.
>
Not ``integrated'' in the M$ sense of just being the same program,
which is a *stupid* and *brain-dead* *mis*use of the concept; it does,
however, have integrated (in the more sensible Unix sense of the term)
GUI builders, e.g., Glade, whatever KDE has, etc. I think CLM (a Motif
library for CL) also comes with a GUI builder of some sort (or maybe
that's Garnet, or maybe I'm misremembering).
> Python and Perl are free, just like CLISP and CMU/CL, and there are
> similar sorts of limited offerings of GUI builders for them all that
> tend to be pretty platform-specific.
>
Tk and GTK+ are not platform-specific (nor is QT, for that matter).
CLIM is a *lot* more platform-independent, just as Common Lisp is more
platform-independant than Perl or Python (the scripting language, not
the CL compiler) - because it is a *standard*, not a single
implementation of an unwritten standard (incidentally, an unwritten
standard that fluctuates greatly depending on the phase of the moon
and what the main developer of the single implementation ate for
breakfast).
> Somehow, when you try to point at the things you _want_ to criticize
> about the Lisp community, the result is pretty incriminating against
> the languages you suggest comparing to...
> --
> (reverse (concatenate 'string ········@" "enworbbc"))
> http://www.ntlug.org/~cbbrowne/emacs.html
> "Normally, we don't do people's homework around here, but Venice is a
> very beautiful city, so I'll make a small exception."
> --- Robert Redelmeier compromises his principles
--
BPT <···@tunes.org> /"\ ASCII Ribbon Campaign
backronym for Linux: \ / No HTML or RTF in mail
Linux Is Not Unix X No MS-Word in mail
Meme plague ;) ---------> / \ Respect Open Standards
* Henry Lebowzki wrote:
> Ain't it funny how you don't see the contempt with which you guys
> regard free software (and free in _all_ it's meanings - free and FULLY
> FUNCTIONAL).
The free CLs *are* fully functional, that's the point. They are
high-quality systems fully capable of supporting serious applications.
I use them every day.
What I'm objecting to is the continual redefinition of the requirement
that a free Lisp system must meet, which is the classic behaviour of
someone who prefers whining to actually bloody *doing something*. Why
not argue about gcc, which also is missing features that commercial
systems have? Do you think I despise gcc because of that? Do you, in
fact, think at all?
--tim
Henry Lebowzki wrote:
> Ain't it funny how you don't see the contempt with which you guys regard
> free software (and free in _all_ it's meanings - free and FULLY FUNCTIONAL).
> If only the C++, Python or Perl community were like that...I'm glad they
> aren't.
> I'm new to LISP, I am very enthusiastic about it, but c.l.l. looks like
> it's full of grey-haired engineers (nothing wrong with that per se)
> making big bucks on some software company, who couldn't care less if the
> free version has the same cool GUI then the proprietary LISP.
______
/ \
.' PLEASE `.
| DO NOT | _____
| FEED THE | ,'.....`.
`. TROLLS ,' ,'........ )
\_ _/ |........ ,'
| | `. .... _/
| | ,'.,'-'
| | /../
| | ,'.,'
| | /../
. | | /..'
.\_\| |/_/,
___ | | ___
. `--' .
. .
--
Gareth McCaughan ················@pobox.com
.sig under construc
* Thaddeus L Olczyk wrote:
> Consider a small program I just wrote. It descends a directory and
> extracts all the filename extensions. Now it used car, cdr mapc and
> a few other basic "lispisms". But it also uses regular expression
> matching, string maniplation, file system examination. The funiest
> was that I had no idea if there was a function that told me if
> something was already in a list. ( I settled on using pushnew. )
Really? Can you say why it does all this hair?
I'd say the following. Given two functions:
DIRECTORY-DIRECTORIES d -> subdirs of d
DIRECTORY-FILES d -> files in d
(and making mild assumptions about the tree structure of the
filesystem, which probably come down to ignoring symlinks), you can
write this algorithm without any of this weird stuff you seem to need.
(defun unique-extensions (root)
(let ((extensions '())) ;might want a hashtable or something
(labels ((do-one-dir (d)
(dolist (file (directory-files d))
(let ((type (pathname-type file)))
(when type
(pushnew type extensions :test #'string=))))
(dolist (sub (directory-directories d))
(do-one-dir sub))))
(do-one-dir root)
extensions)))
the two DIRECTORY-* functions are not completely portable, true. But
their implementation is not likely to involve huge complexity with
regexps or anything.
Indeed I'm in the process of writing a system which does really a lot
of filename bashing in CL and these two functions, as well as a couple
of others to canonicalise user input and output are almost all I need.
--tim
From: Thaddeus L Olczyk
Subject: Lisp books don't teach you how write real programs ( Re: Free Lisp with GUI )
Date:
Message-ID: <3c531b3a.347750671@nntp.interaccess.com>
On 21 Jan 2002 23:27:01 +0000, Tim Bradshaw <···@cley.com> wrote:
>* Thaddeus L Olczyk wrote:
>
>> Consider a small program I just wrote. It descends a directory and
>> extracts all the filename extensions. Now it used car, cdr mapc and
>> a few other basic "lispisms". But it also uses regular expression
>> matching, string maniplation, file system examination. The funiest
>> was that I had no idea if there was a function that told me if
>> something was already in a list. ( I settled on using pushnew. )
>
>Really? Can you say why it does all this hair?
>
>I'd say the following. Given two functions:
>
> DIRECTORY-DIRECTORIES d -> subdirs of d
> DIRECTORY-FILES d -> files in d
>
>(and making mild assumptions about the tree structure of the
>filesystem, which probably come down to ignoring symlinks), you can
>write this algorithm without any of this weird stuff you seem to need.
>
>(defun unique-extensions (root)
> (let ((extensions '())) ;might want a hashtable or something
> (labels ((do-one-dir (d)
> (dolist (file (directory-files d))
> (let ((type (pathname-type file)))
> (when type
> (pushnew type extensions :test #'string=))))
> (dolist (sub (directory-directories d))
> (do-one-dir sub))))
> (do-one-dir root)
> extensions)))
>
>the two DIRECTORY-* functions are not completely portable, true. But
>their implementation is not likely to involve huge complexity with
>regexps or anything.
>
So is this some sort of "show what a hot shot developer I am so I can
compensate for other short comings" post?
Some Lisp programmers always seem to want to take a path where they
lose site of the broad sense in exchange for the details.
The statement was: "An experienced developer reading books about Lisp
is not adequately prepared to take anything more than an entry level
job in Lisp."
Another variant was: "A manager is not likely to hire an experienced
nonLisp developer to do Lisp unless he shows significant independent
work in Lisp." (Entry level jobs excluded.)
To show a reason why these two assertions were true, I picked a simple
program ( less than an hour, fairly good considering I'm not that
experienced with Lisp ) and pulled out things that most book taught
programmers with no experience writing would not be that familiar
with.
So what do you do? In an effort to show you are not that "small" a
person, you post code showing how brilliant you are. However in
posting your code, you demonstrate precisely what I was saying.
You eschew regex's to use pathname-type. ( Kind of dumb to gloat that
you don't need to need to use regex's. What do you think pathname-type
does? What id you want to pick out a "dual" extension like foo.ps.tgz
or even a triple: foo.ps.tar.gz?) Yet pathname-type is not mentioned
in Touretzky or Winston and Horn. So already there is something the
book-learned Lisp programmer is missing. Then you hide stuff behind
directory-files and directory-directories pretending that it was not
something not found in books. ( Again the two sources don't mention
this stuff, in particular the directories command. )
The fact is this. In every project there is a lot of stuff that is
ordinary programming. ( Meaning Lisp doesn't take particular advantage
over other languages. ) Some would say slough that stuff onto C but
sometimes it isn't possible. Therefore a Lisp programmer must be able
to write ordinary programs in Lisp.
But if you ask a programmer with passing knowledge of Lisp, he will
say that you can't write such a program ( to find all file extensions
) in Lisp. Even many prorammers who know lisp will say that you
can't write such a program in Lisp. If asked to do so they will admit
they couldn't do so.
That's why you need free implementations. So that programmers can go
out and give them a test drive.
From: Raymond Toy
Subject: Re: Lisp books don't teach you how write real programs ( Re: Free Lisp with GUI )
Date:
Message-ID: <4ny9iqgttj.fsf@rtp.ericsson.se>
>>>>> "Thaddeus" == Thaddeus L Olczyk <······@interaccess.com> writes:
Thaddeus> or even a triple: foo.ps.tar.gz?) Yet pathname-type is not mentioned
Thaddeus> in Touretzky or Winston and Horn. So already there is something the
Thaddeus> book-learned Lisp programmer is missing. Then you hide stuff behind
Thaddeus> directory-files and directory-directories pretending that it was not
Thaddeus> something not found in books. ( Again the two sources don't mention
Thaddeus> this stuff, in particular the directories command. )
Why do you think a textbook must cover everything about a subject?
Did you complain that your Physics I text book didn't talk about
special relativity or general relativity or quantum mechanics? Can
you solve real problems on general relativity from your Physics I
text?
Ray
From: Pierre R. Mai
Subject: Re: Lisp books don't teach you how write real programs ( Re: Free Lisp with GUI )
Date:
Message-ID: <87hepe8r11.fsf@orion.bln.pmsf.de>
······@interaccess.com (Thaddeus L Olczyk) writes:
> You eschew regex's to use pathname-type. ( Kind of dumb to gloat that
> you don't need to need to use regex's. What do you think pathname-type
> does? What id you want to pick out a "dual" extension like foo.ps.tgz
I'd be heavily astonished if an implementation of pathname-type
resorted to using regular expressions (in the POSIX sense of the
word), where a simple character search sufficed. This introduces
needless hair and complexity, where a simple, robust and fast
alternative exists, and would lead me to question the mental acuity of
the implementor.
Obsession with RegExp's is IMHO a growing problem with professional
developers, that probably should be recognized as a work-related
illness, especially in Unix-related jobs.
> or even a triple: foo.ps.tar.gz?) Yet pathname-type is not mentioned
Again, I would never use POSIX-style regular expressions for this kind
of job. Simple character searches suffice for this kind of job. I
don't use a hairy graph-traversal algorithm where a simple tree-walk
suffices, so why should I use regular expressions when a simple
character search suffices?
> in Touretzky or Winston and Horn. So already there is something the
> book-learned Lisp programmer is missing. Then you hide stuff behind
But it is "mentioned" (i.e. specified) in the ANSI CL standard, and
I'd expect every book-learned Lisp programmer to have read relevant
portions of the standard, which is available both as a PDF file for
15$ from ANSI, and more importantly in the form of the HyperSpec, from
www.xanalys.com for free.
Anyone who doesn't regularly look at the relevant standards for any
programming language he is using/learning, is in my book neither a
book-learned nor any other kind of programmer, at all.
> That's why you need free implementations. So that programmers can go
> out and give them a test drive.
And they are abundantly available. At the same site you got your copy
of the HyperSpec, you will find personal editions of LispWorks, both
for Windows and for Linux available for free download. They include a
serious GUI IDE and GUI toolkit, lots of documentation, and plenty of
of opportunities to test drive them. The only restrictions are that
you can't redistribute applications, there is a heap limit, and a time
limit of 5 hours per session. None of those would in any way have
restricted your ability to e.g. develop the application we have been
talking about here.
Franz offers similar trial editions for free download, as does
Digitool, and Corman, too.
And then there are the open source implementations, like CMU CL, CLISP
or ECLS, which would have been totally sufficient for the purposes
mentioned here, too.
Regs, Pierre.
--
Pierre R. Mai <····@acm.org> http://www.pmsf.de/pmai/
The most likely way for the world to be destroyed, most experts agree,
is by accident. That's where we come in; we're computer professionals.
We cause accidents. -- Nathaniel Borenstein
From: Tim Bradshaw
Subject: Re: Lisp books don't teach you how write real programs ( Re: Free Lisp with GUI )
Date:
Message-ID: <ey3u1te1hcf.fsf@cley.com>
* Thaddeus L Olczyk wrote:
> So is this some sort of "show what a hot shot developer I am so I can
> compensate for other short comings" post?
Not as far as I'm aware. I was and am genuinely astonished that all
these things (regexps &c) seemed to be necessary. I certainly don't
regard myself as a brilliant programmer in CL or any other language.
I wish I was...
> You eschew regex's to use pathname-type. ( Kind of dumb to gloat that
> you don't need to need to use regex's. What do you think pathname-type
> does?
It gets the pathname type. In a filesystem with direct support for
types it will do it by asking the FS, otherwise it will do it by some
code. If I found an implementation that used regexps to do this I'd
probably consider that a serious incentive to not use it.
> What id you want to pick out a "dual" extension like foo.ps.tgz
> or even a triple: foo.ps.tar.gz?)
I think this kind of stretches the notion of the type of a file.
However if I needed to do this I'd find out what my lisp
implementation did and then either implement my own or use the
provided tools to solve the problem. I checked the implementations I
have to hand (2 free, one commercial) and they all parse filenames the
same way so you'd need to parse the name to get the additional
extensions. A hacky, probably slow, but kind of wonderful way of
doing it is to keep parsing the name component until nothing more
comes out:
(defun all-extensions (pathname)
;; Return the root name of a pathname and a list of all the extensions.
;; "foo.tar.gz" -> "foo" and ("tar" "gz").
;;
;; This depends on the system parsing a.b.c as a.b and c. This needs
;; to be tested per implementation.
(loop with types
for p = pathname then (parse-namestring (pathname-name p))
for type = (pathname-type p)
while type
do (push type types)
finally (return (values (pathname-name p) types))))
A better solution would probably be to tokenize the thing yourself,
especially if your system is large enough that you have one or more of
the split-sequence functions already.
> Yet pathname-type is not mentioned in Touretzky or Winston and
> Horn. So already there is something the book-learned Lisp programmer
> is missing.
I guess we disagree what `book-learned' implies. I assumed, obviously
incorrectly, that such a person would at least be aware of the chapter
titles in the standard, or if not that have looked at CLtL2.
> Then you hide stuff behind
> directory-files and directory-directories pretending that it was not
> something not found in books. ( Again the two sources don't mention
> this stuff, in particular the directories command. )
No, I didn't pretend it was not found in books. DIRECTORY is found in
the spec, and it is specified that it may take unspecified arguments,
presumably for this kind of thing. How to get just the subdirs or
files is not specified, and is probably mildly implementation
dependent. For LW for instance:
(defun directory-subdirectories (dir)
;; Return only the subdirectories in DIR: not . or .. if the
;; implementation does not filter them anyway (LW does).
;; The documentation implies we (might) need to give CHECK-FOR-SUBS
;; which is not actually supported in 4.2
;;
;; LWW seems a bit weird here. components that are :UNSPECIFIC
;; prevent things matching, and stuff that comes back from DIRECTORY
;; have :UNSPECIFIC components. Specifying :DIRECTORIES seems to
;; `cure' this (if it's a bug).
(directory dir
:directories t
:test #'(lambda (p)
(null (pathname-name p)))))
Is what you need (this is actually part of the system I'm working on,
in the (small) impdep part.)
> The fact is this. In every project there is a lot of stuff that is
> ordinary programming. ( Meaning Lisp doesn't take particular advantage
> over other languages. ) Some would say slough that stuff onto C but
> sometimes it isn't possible. Therefore a Lisp programmer must be able
> to write ordinary programs in Lisp.
Yes, and they do, or I do, anyway.
> That's why you need free implementations. So that programmers can go
> out and give them a test drive.
Shall I tell you a secret? There are several hundred percent more
free CL implementations than there are free perl implementations, and
infinitely more fully commercial implementations.
--tim
From: Daniel Barlow
Subject: Re: Lisp books don't teach you how write real programs ( Re: Free Lisp with GUI )
Date:
Message-ID: <874rleuoyf.fsf@noetbook.telent.net>
······@interaccess.com (Thaddeus L Olczyk) writes:
> You eschew regex's to use pathname-type. ( Kind of dumb to gloat that
> you don't need to need to use regex's. What do you think pathname-type
> does? What id you want to pick out a "dual" extension like foo.ps.tgz
I think pathname-type is an accessor for the "type" property of a
structured pathname object. I doubt very much that its implementation
involves regexes - regexes on _what_ exactly?
(Even if the pathname _were_ stored internally as a string, returning
the type is simply a matter of finding the rightmost #\. and returning
everything to the right of it. Only a Perl programmer would think
that a regex is the most suitable tool for doing this)
> or even a triple: foo.ps.tar.gz?) Yet pathname-type is not mentioned
> in Touretzky or Winston and Horn. So already there is something the
> book-learned Lisp programmer is missing. Then you hide stuff behind
We're fortunate, then, that the ANSI standard is so clear and readable
and in the form of the HyperSpec freely (lowercase f) acessible to anyone
who wants it. I don't think it's unreasonable that the "book-learned"
cl programmer also know how to look things up in the language reference
> That's why you need free implementations. So that programmers can go
> out and give them a test drive.
I happen to believe that we need free implementations, which is why I
work on them. I don't see how that follows from the existence or
nonexistence of documentation describing pathname-type, though.
Oh, I will add: "extension" is not a well-defined concept in unix
filesystem terms anyway (what's the extension of ".cshrc"? of
"/etc/rc.d/"?), and this uncertainty is reflected in the mapping of
the common lisp pathname system onto unix filesystems. Unless you
carefully define what you actually want in the ambiguous cases, you
will be surprised by at least something you get. Productive use of
the CL pathname system is something I actually do think an article
could usefully be written about. But I'm not volunteering, my plate
is full with SBCL already.
-dan
--
http://ww.telent.net/cliki/ - Link farm for free CL-on-Unix resources
> ... Yet pathname-type is not mentioned
> in Touretzky or Winston and Horn. So already there is something the
> book-learned Lisp programmer is missing. Then you hide stuff behind
> directory-files and directory-directories pretending that it was not
> something not found in books. ( Again the two sources don't mention
> this stuff, in particular the directories command. )
>
> The fact is this. In every project there is a lot of stuff that is
> ordinary programming...
This brings up a very good point--it would be nice if there was a CL
Cookbook (in the spirit of the Perl Cookbook) that shows how to do
day-to-day tasks in CL.
Come to think of it, file access is all but missing from all lisp
books I've read. Paul Graham's ANSI CL covers touches on it, but
that's about it.
From: Dr. Edmund Weitz
Subject: CL cookbook (Was: Lisp books don't teach you how write real programs)
Date:
Message-ID: <m3elkhcslr.fsf_-_@bird.agharta.de>
·············@yahoo.com (dj special ed) writes:
> This brings up a very good point--it would be nice if there was a CL
> Cookbook (in the spirit of the Perl Cookbook) that shows how to do
> day-to-day tasks in CL.
I've thought the same thing recently. There are a couple of wonderful
books about CL - including Norvig's and Graham's - but most of them
deal with high-level stuff that you couldn't do that easily in other
languages. Unfortunately, the stuff that's _easy_ in other languages
is mostly missing - partly due to the fact that it's either OS
specific or requires extensions (sockets, Internet protocols, GUI,
database access, you name it) that aren't part of the ANSI standard.
I think a good start (especially to attract newbies) would be a
cookbook that implements the recipes from the various Perl, Python,
and Java cookbooks for the free CL implementations on *nix systems
plus CLOCC.
I'd actually volunteer for such a project but I'm afraid my Lisp ain't
good enough and there's lots of stuff that I have to learn - currently
my attempts would be an insult to the community.
Edi.
From: Erik Winkels
Subject: Re: CL cookbook (Was: Lisp books don't teach you how write real programs)
Date:
Message-ID: <87u1td9sm8.fsf@xs4all.nl>
···@agharta.de (Dr. Edmund Weitz) writes:
>
> I'd actually volunteer for such a project but I'm afraid my Lisp
> ain't good enough and there's lots of stuff that I have to learn
Same here, both on the volunteering and on the inexperience.
Frankly, I don't think that our limited experience is really a
problem. What we write might help other newcomers along and there's
bound to be someone with more experience to make improvements to our
code if they feel like it (or if they just can't contain themselves :)
Erik.
From: Thom Goodsell
Subject: Re: CL cookbook (Was: Lisp books don't teach you how write real programs)
Date:
Message-ID: <7vd701f7b2.fsf@shalott.cra.com>
Erik Winkels <·······@xs4all.nl> writes:
> ···@agharta.de (Dr. Edmund Weitz) writes:
> >
> > I'd actually volunteer for such a project but I'm afraid my Lisp
> > ain't good enough and there's lots of stuff that I have to learn
>
> Same here, both on the volunteering and on the inexperience.
>
> Frankly, I don't think that our limited experience is really a
> problem. What we write might help other newcomers along and there's
> bound to be someone with more experience to make improvements to our
> code if they feel like it (or if they just can't contain themselves :)
>
Why doesn't someone throw up a page on CLiki for this? (I was going
to, but decided I'd leave it to someone who knows how CLiki works.) If
there's a page there, then anyone can post a "recipe" they'd like
converted, others can post a Lisp version, and still others can refine
the programs. The page should have some guidelines (like, "this is not
the place to show off obscure optimizations" and "don't post homework
questions; don't answer things that look like homework questions"),
but I think it would work out fine. I'd certainly be happy to spend 20
minutes throwing together a few recipes on my coffee breaks.
Once there's enough content, someone can format it up nicely. I
suppose that if it actually comes to publishing it there could be some
copyright issue, but if we just PDF it and make it freely available
then at least no one will be fighting over the royalties.
Thom
--
(let ((e-mail-address ····@IXG.IUS")) (loop with new-string =
(make-string (length e-mail-address)) for count from 0 to (1- (length
e-mail-address)) for char-code = (char-code (aref e-mail-address
count)) for new-char-code = (if (and (> char-code 64) (< char-code
123)) (+ (mod (+ 13 char-code) 52) 65) char-code) do (setf (aref
new-string count) (code-char new-char-code)) finally (return
new-string)))
In article <·················@bird.agharta.de>, Dr. Edmund Weitz wrote:
> ·············@yahoo.com (dj special ed) writes:
>
>> This brings up a very good point--it would be nice if there was a CL
>> Cookbook (in the spirit of the Perl Cookbook) that shows how to do
>> day-to-day tasks in CL.
>
> I've thought the same thing recently. There are a couple of wonderful
> books about CL - including Norvig's and Graham's - but most of them
> deal with high-level stuff that you couldn't do that easily in other
> languages. Unfortunately, the stuff that's _easy_ in other languages
> is mostly missing - partly due to the fact that it's either OS
> specific or requires extensions (sockets, Internet protocols, GUI,
> database access, you name it) that aren't part of the ANSI standard.
IIRC, Kent Pitman is working on a book about Lisp. I think he wanted
to address precisely this issue.
Regards,
--
Nils Goesche
"Don't ask for whom the <CTRL-G> tolls."
PGP key ID 0x42B32FC9
From: Thaddeus L Olczyk
Subject: Re: CL cookbook (Was: Lisp books don't teach you how write real programs)
Date:
Message-ID: <3c4ef5e0.48908765@nntp.interaccess.com>
On 23 Jan 2002 10:41:04 +0100, ···@agharta.de (Dr. Edmund Weitz)
wrote:
>·············@yahoo.com (dj special ed) writes:
>
>> This brings up a very good point--it would be nice if there was a CL
>> Cookbook (in the spirit of the Perl Cookbook) that shows how to do
>> day-to-day tasks in CL.
>
>I've thought the same thing recently. There are a couple of wonderful
>books about CL - including Norvig's and Graham's - but most of them
>deal with high-level stuff that you couldn't do that easily in other
>languages. Unfortunately, the stuff that's _easy_ in other languages
>is mostly missing - partly due to the fact that it's either OS
>specific or requires extensions (sockets, Internet protocols, GUI,
>database access, you name it) that aren't part of the ANSI standard.
>
Not just high-level stuff but middle level stuff.
The problem is that withoput the low level stuff it's all rather weak.
>I think a good start (especially to attract newbies) would be a
>cookbook that implements the recipes from the various Perl, Python,
>and Java cookbooks for the free CL implementations on *nix systems
>plus CLOCC.
>
Use CLisp woth CLOCC and don't limit yourself to UNIX.
Look at the popularity of emacs when the NT version became viable.
Ask the mainainers and they will tell you usage went through the roof.
>I'd actually volunteer for such a project but I'm afraid my Lisp ain't
>good enough and there's lots of stuff that I have to learn - currently
>my attempts would be an insult to the community.
Actually that would be the best time. What seems hard to beginers is
easy to experts, so they don't bother to explain it. Wereas a beginer
would feel compeled.
Perhaps you might rate each entry on difficulty too.
From: Dr. Edmund Weitz
Subject: Re: CL cookbook (Was: Lisp books don't teach you how write real programs)
Date:
Message-ID: <m3pu406by0.fsf@bird.agharta.de>
······@interaccess.com (Thaddeus L Olczyk) writes:
> Use CLisp woth CLOCC and don't limit yourself to UNIX. Look at the
> popularity of emacs when the NT version became viable. Ask the
> mainainers and they will tell you usage went through the roof.
I don't want to rule out Windows, I just think it'll be easier to cope
with stuff like sockets, FFI, database, etc. if you limit yourself to
*nix platforms in the beginning.
Look at the popular (and rightly so, IMHO) Perl cookbook and you'll
see that a lot of stuff there is *nix-specific, too.
Edi.
From: Duane Rettig
Subject: Re: CL cookbook (Was: Lisp books don't teach you how write real programs)
Date:
Message-ID: <4hepchi9j.fsf@beta.franz.com>
···@agharta.de (Dr. Edmund Weitz) writes:
> ······@interaccess.com (Thaddeus L Olczyk) writes:
>
> > Use CLisp woth CLOCC and don't limit yourself to UNIX. Look at the
> > popularity of emacs when the NT version became viable. Ask the
> > mainainers and they will tell you usage went through the roof.
>
> I don't want to rule out Windows, I just think it'll be easier to cope
> with stuff like sockets, FFI, database, etc. if you limit yourself to
> *nix platforms in the beginning.
I prefer unixen as well, but to be fair, our product supports all of the
above on Windows (all versions, not just NT).
> Look at the popular (and rightly so, IMHO) Perl cookbook and you'll
> see that a lot of stuff there is *nix-specific, too.
I think that being anything-specific is a negative for a general-purpose
tool.
--
Duane Rettig Franz Inc. http://www.franz.com/ (www)
1995 University Ave Suite 275 Berkeley, CA 94704
Phone: (510) 548-3600; FAX: (510) 548-8253 ·····@Franz.COM (internet)
From: Dr. Edmund Weitz
Subject: Re: CL cookbook (Was: Lisp books don't teach you how write real programs)
Date:
Message-ID: <m3pu404msn.fsf@bird.agharta.de>
Duane Rettig <·····@franz.com> writes:
> ···@agharta.de (Dr. Edmund Weitz) writes:
> >
> > I don't want to rule out Windows, I just think it'll be easier to
> > cope with stuff like sockets, FFI, database, etc. if you limit
> > yourself to *nix platforms in the beginning.
>
> I prefer unixen as well, but to be fair, our product supports all of
> the above on Windows (all versions, not just NT).
>
> > Look at the popular (and rightly so, IMHO) Perl cookbook and
> > you'll see that a lot of stuff there is *nix-specific, too.
>
> I think that being anything-specific is a negative for a
> general-purpose tool.
You're right of course. I've thought about it on my way home from a
customer before I read your posting and came to the conclusion that if
we're going to succeed in a community effort like a CL cookbook it
would be dumb to rule out specific platforms, especially because we
have at least three implementations (Allegro CL, CLISP and - to a
lesser degree - LispWorks) that are almost platform-agnostic. A CL
cookbook should provide answers to (seemingly) simple questions
without being tied to an implementation or an operating system.
I'll try to come up with a proposal for such a cookbook in the next
days. (Don't hold your breath - I have a family to feed first... :)
Edi.
From: Marc Spitzer
Subject: Re: Lisp books don't teach you how write real programs ( Re: Free Lisp with GUI )
Date:
Message-ID: <slrna4tq18.1crl.marc@oscar.eng.cv.net>
In article <····························@posting.google.com>,
dj special ed wrote:
>> ... Yet pathname-type is not mentioned
>> in Touretzky or Winston and Horn. So already there is something the
>> book-learned Lisp programmer is missing. Then you hide stuff behind
>> directory-files and directory-directories pretending that it was not
>> something not found in books. ( Again the two sources don't mention
>> this stuff, in particular the directories command. )
>>
>> The fact is this. In every project there is a lot of stuff that is
>> ordinary programming...
>
> This brings up a very good point--it would be nice if there was a CL
> Cookbook (in the spirit of the Perl Cookbook) that shows how to do
> day-to-day tasks in CL.
>
> Come to think of it, file access is all but missing from all lisp
> books I've read. Paul Graham's ANSI CL covers touches on it, but
> that's about it.
Object Oriented Common Lisp has a chapter on string handling and
streams, with basic file operations: create, open, close, write,
read.
I think that this is covered in the hyperspec and in vendor docs means
most lispers(not me yet) will be able to figure it out with out much
work after they are reasonably experienced with the language.
lisp does evil things with mediocre programmers, everything is a list
now go build me a btree. It will work but boy is it slow.
marc
From: Marc Spitzer
Subject: Re: Lisp books don't teach you how write real programs ( Re: Free Lisp with GUI )
Date:
Message-ID: <slrna4tstc.1d0p.marc@oscar.eng.cv.net>
In article <····················@oscar.eng.cv.net>, Marc Spitzer wrote:
> In article <····························@posting.google.com>,
> dj special ed wrote:
>>> ... Yet pathname-type is not mentioned
>>> in Touretzky or Winston and Horn. So already there is something the
>>> book-learned Lisp programmer is missing. Then you hide stuff behind
>>> directory-files and directory-directories pretending that it was not
>>> something not found in books. ( Again the two sources don't mention
>>> this stuff, in particular the directories command. )
>>>
>>> The fact is this. In every project there is a lot of stuff that is
>>> ordinary programming...
>>
>> This brings up a very good point--it would be nice if there was a CL
>> Cookbook (in the spirit of the Perl Cookbook) that shows how to do
>> day-to-day tasks in CL.
>>
>> Come to think of it, file access is all but missing from all lisp
>> books I've read. Paul Graham's ANSI CL covers touches on it, but
>> that's about it.
>
> Object Oriented Common Lisp has a chapter on string handling and
> streams, with basic file operations: create, open, close, write,
> read.
>
> I think that this is covered in the hyperspec and in vendor docs means
> most lispers(not me yet) will be able to figure it out with out much
> work after they are reasonably experienced with the language.
>
> lisp does evil things with mediocre programmers, everything is a list
> now go build me a btree. It will work but boy is it slow.
This one did not make much sence, please ignore.
>
> marc
>
>
From: Brad Knotwell
Subject: Re: Lisp books don't teach you how write real programs ( Re: Free Lisp with GUI )
Date:
Message-ID: <86adv4l34y.fsf@localhost.my.domain>
·············@yahoo.com (dj special ed) writes:
> This brings up a very good point--it would be nice if there was a CL
> Cookbook (in the spirit of the Perl Cookbook) that shows how to do
> day-to-day tasks in CL.
For this sort of thing, the following URL might be helpful:
http://www.bagley.org/~doug/shootout
It's Doug Bagley's attempt to compare programming languages on a
fairly wide variety of simple tasks. From my perspective, it does
a pretty good job of achieving its stated goal. However, I'd say
it does an even better job at its unstated goal--providing small
examples of code for a variety of common tasks and, maybe more importantly,
_methodologies_ on how to "use" specific implementations to
organize/build/execute/deliver code.
I include methodologies as being important because it's not always
clear how to use an implementation or language to build and deliver a
standalone application (standalone in a broad sense, not just a simple
executable). In other words, Ocaml (at least for Unix) is easier to
learn than SML/NJ because its build and delivery process are parallel to
C's--make and standalone (in a narrow sense) executables (NOTE: it's also
easier to learn because many of their library calls mimic standard C and
Unix library calls; depending on your perspective, this is a bug or feature
and, correspondingly, it's to Ocaml's credit/detriment to be a
mono-implementation without a published specification). OTOH, SML/NJ has
an unquestionably cooler build mechanism (CM), but it's proven wholly
daunting to this mere mortal. Likewise, without instruction, someone coming
from a more traditional delivery environment will be more likely find the
build and delivery process for Common Lisp applications confusing and/or
kludgy.
--Brad
From: Bulent Murtezaoglu
Subject: Re: Lisp books don't teach you how write real programs ( Re: Free Lisp with GUI )
Date:
Message-ID: <874rlcnvom.fsf@nkapi.internal>
>>>>> "BK" == Brad Knotwell <········@ix.netcom.com> writes:
[...]
BK> For this sort of thing, the following URL might be helpful:
BK> http://www.bagley.org/~doug/shootout
BK> It's Doug Bagley's attempt to compare programming languages on
BK> a fairly wide variety of simple tasks. From my perspective,
BK> it does a pretty good job of achieving its stated goal.
BK> However, I'd say it does an even better job at its unstated
BK> goal--providing small examples of code for a variety of common
BK> tasks and, maybe more importantly, _methodologies_ on how to
BK> "use" specific implementations to
BK> organize/build/execute/deliver code. [...]
I agree with the sentiment, but some of the CL code there are hardly
examples of what we'd do w/o the microbenchmark pressure. I think you
contributed, as did I and Jochen Schmidt and probably others whose code
I didn't go through so I can't remember. What we should have done -- and
probably still could -- is to create notes about how we did what and why.
I understand your point (esp the parts I deleted below) and agree, but
I also think the optimization for Bagley's page and the similar
optimization/refinement threads that occasionally appear here are very
worthwhile things to take notes on, polish/edit and save somewhere. Maybe
a section in cliki?
cheers,
BM
From: Jochen Schmidt
Subject: Re: Lisp books don't teach you how write real programs ( Re: Free Lisp with GUI )
Date:
Message-ID: <3C4FF637.8080504@dataheaven.de>
Bulent Murtezaoglu wrote:
>>>>>>"BK" == Brad Knotwell <········@ix.netcom.com> writes:
>>>>>>
> [...]
> BK> For this sort of thing, the following URL might be helpful:
>
> BK> http://www.bagley.org/~doug/shootout
>
> BK> It's Doug Bagley's attempt to compare programming languages on
> BK> a fairly wide variety of simple tasks. From my perspective,
> BK> it does a pretty good job of achieving its stated goal.
> BK> However, I'd say it does an even better job at its unstated
> BK> goal--providing small examples of code for a variety of common
> BK> tasks and, maybe more importantly, _methodologies_ on how to
> BK> "use" specific implementations to
> BK> organize/build/execute/deliver code. [...]
>
> I agree with the sentiment, but some of the CL code there are hardly
> examples of what we'd do w/o the microbenchmark pressure. I think you
> contributed, as did I and Jochen Schmidt and probably others whose code
> I didn't go through so I can't remember. What we should have done -- and
> probably still could -- is to create notes about how we did what and why.
> I understand your point (esp the parts I deleted below) and agree, but
> I also think the optimization for Bagley's page and the similar
> optimization/refinement threads that occasionally appear here are very
> worthwhile things to take notes on, polish/edit and save somewhere. Maybe
> a section in cliki?
Since Bagley's shootout is no longer maintained I thought a while that
it might be a good idea to setup another page were people actually can
work on. We could use the available programs of Bagley's shootout for a
start but I would recommend to have an emphasis on how the things are
_realistically_ solved in the language. So it should not get enforced to
use READ-LINE in lisp when reusing a buffer greatly gains in
performance. I began to like Bagley's shootout because it sometimes
showed very unoptimized parts of the language implementations. The
benchmarks are mostly silly yes - but if there is a huge difference
between CMUCL and C/C++ then in most times you can do better.
Even if we do not maintain further the solutions to all other languages
we should at least try to have C/C++ and probably a common
scripting-language like Perl or Python on the page. Adding an
interpreted CL like CLISP is certainly another interesting idea.
ciao,
Jochen
Brad Knotwell <········@ix.netcom.com> writes:
> ·············@yahoo.com (dj special ed) writes:
> > This brings up a very good point--it would be nice if there was a CL
> > Cookbook (in the spirit of the Perl Cookbook) that shows how to do
> > day-to-day tasks in CL.
>
> For this sort of thing, the following URL might be helpful:
>
> http://www.bagley.org/~doug/shootout
Do you know http://pleac.sourceforge.net/ ?
Greetings,
-- Jorgen
--
((email . ·······@mindless.com") (www . "http://forcix.cx/")
(irc . ·······@#StarWars (IRCnet)") (gpg . "1024D/028AF63C"))
From: Marco Antoniotti
Subject: Re: Lisp books don't teach you how write real programs ( Re: Free Lisp with GUI )
Date:
Message-ID: <y6cu1tbaibj.fsf@octagon.mrl.nyu.edu>
Jorgen Schaefer <······@mindless.com> writes:
> Brad Knotwell <········@ix.netcom.com> writes:
>
> > ·············@yahoo.com (dj special ed) writes:
> > > This brings up a very good point--it would be nice if there was a CL
> > > Cookbook (in the spirit of the Perl Cookbook) that shows how to do
> > > day-to-day tasks in CL.
> >
> > For this sort of thing, the following URL might be helpful:
> >
> > http://www.bagley.org/~doug/shootout
>
> Do you know http://pleac.sourceforge.net/ ?
Now... Who in the world would want to program in a language namede
`merd'? :)
Ciao
--
Marco Antoniotti ========================================================
NYU Courant Bioinformatics Group tel. +1 - 212 - 998 3488
719 Broadway 12th Floor fax +1 - 212 - 995 4122
New York, NY 10003, USA http://bioinformatics.cat.nyu.edu
"Hello New York! We'll do what we can!"
Bill Murray in `Ghostbusters'.
From: Dr. Edmund Weitz
Subject: Re: Lisp books don't teach you how write real programs ( Re: Free Lisp with GUI )
Date:
Message-ID: <m3elkgoyt4.fsf@bird.agharta.de>
Jorgen Schaefer <······@mindless.com> writes:
> Do you know http://pleac.sourceforge.net/ ?
I did not know it until I saw your posting and it looks
interesting. So I think contributing to PLEAC should have been variant
2.4 in my Cookbook proposal (see
<···················@bird.agharta.de>).
Edi.
Tim Bradshaw <···@cley.com> writes:
> * Greg Menke wrote:
>
> > Perhaps Franz has a business arrangement where dealing with
> > individuals isn't profitable and they get their money out of some few
> > huge customers, but I wonder how many sales they end up losing just
> > from making it hard for people to give them money.
>
> In Franz's defense (I am not a Franz customer), they did, some years
> ago, sell a cheap (~ $1000 or maybe much less) Lisp for PCs (this was
> pre linux being a major platform). I presume they stopped selling it
> because it wasn't profitable. There are lots of models for selling
FWIW the SRP of Allegro CL/PC was $995, and you could e.g. get a
discount of $200 with the voucher in the back of at least some
printings of Graham's On Lisp.
Regs, Pierre.
--
Pierre R. Mai <····@acm.org> http://www.pmsf.de/pmai/
The most likely way for the world to be destroyed, most experts agree,
is by accident. That's where we come in; we're computer professionals.
We cause accidents. -- Nathaniel Borenstein
In article <···············@cley.com>, Tim Bradshaw <···@cley.com> wrote:
> * Greg Menke wrote:
>
>> Perhaps Franz has a business arrangement where dealing with
>> individuals isn't profitable and they get their money out of some few
>> huge customers, but I wonder how many sales they end up losing just
>> from making it hard for people to give them money.
>
> In Franz's defense (I am not a Franz customer), they did, some years
> ago, sell a cheap (~ $1000 or maybe much less) Lisp for PCs (this was
> pre linux being a major platform).
You're out of your mind!
You're probably an American, snug in your confortable Suburbia home, out of contact with the whole f*****g planet!
Do you know just __how much__ $1000 is in Venezuela? Russia? India? It's __not__ cheap!
That f*****g alienated mentality, I guess, explains why commercial Lisp is years ahead of free-software lisp (which, as I see often here in c.l.l. is regarded as something for the sorry fellows who don't have a thousand bucks to spend).
If there were more programmers like that, we wouldn't have GNU/Linux, GCC, C, C++, Perl , Python, etc.
So you see, the problem with LISP is not so much the __language__, but lack of involvement in the community with developing free software, while they just remain seated waiting for Franz or whatever to deliver them their next fix.
Hey, you don't watch world news on TV, do you?
Regs,
(and it's __not__ personal, ok? You just represent a certain mind set, that's all...)
synthespian
> --tim
Okay, you're really starting to look like a troll, but I'm giving you
the benefit of the doubt, here.
Tim is most definitely not out of his mind, and I suspect that he's
got a good (though probably not precise) idea of how much $1000 is in
Venezuela, Russia, India, and Brazil. The problem is that developing
good software, like that sold by Franz, requires a considerable amount
of time. Now, since most developers want to be paid, that requires
money. Since Franz develops software in the U.S., where the cost of
living is relatively high, it actually requires a lot of
money. Consequently, Franz charges money for their products. Their
prices are set the way they are because it keeps them in business.
Now, if someone in India wants to develop a Common Lisp environment
at India rates and price it to be attractive in India, Russia,
Venezuela, and Pakistan, they're more than welcome to do so. If it's a
quality product it will also find buyers in North America and Europe,
since it will cost less than the U.S.-built competitors.
Similarly, anyone who already _has_ the money they want can spend time
(or money) to improve the free CLs that are available. CMUCL is a very
good product that lacks only a few things to be a truly "industrial
strength" implementation; if you know someone whose economic needs
have already been met, please have them contribute to it. Maybe you
could even start a non-profit to work on it. That would certainly seem
appropriate, since you imply that writing free software is a
humanitarian project.
At any rate, bitching on c.l.l isn't going to solve the problem. If
you don't believe me, try searching google for all the times this has
come up in the past. And good luck with CMUCL or CLISP--there really
are a lot of people quietly working on improving those
implementations.
Regards,
Thom
P.S. The next time you attack on mindset and then claim it's nothing
personal, please refrain from the ad hominem attacks.
synthespian <···········@uol.com.br> writes:
>
> You're out of your mind!
> You're probably an American, snug in your confortable Suburbia home, out of contact with the whole f*****g planet!
> Do you know just __how much__ $1000 is in Venezuela? Russia? India? It's __not__ cheap!
> That f*****g alienated mentality, I guess, explains why commercial Lisp is years ahead of free-software lisp (which, as I see often here in c.l.l. is regarded as something for the sorry fellows who don't have a thousand bucks to spend).
> If there were more programmers like that, we wouldn't have GNU/Linux, GCC, C, C++, Perl , Python, etc.
> So you see, the problem with LISP is not so much the __language__, but lack of involvement in the community with developing free software, while they just remain seated waiting for Franz or whatever to deliver them their next fix.
> Hey, you don't watch world news on TV, do you?
>
> Regs,
> (and it's __not__ personal, ok? You just represent a certain mind set, that's all...)
> synthespian
>
>
>
> > --tim
>
>
--
(let ((e-mail-address ····@IXG.IUS")) (loop with new-string =
(make-string (length e-mail-address)) for count from 0 to (1- (length
e-mail-address)) for char-code = (char-code (aref e-mail-address
count)) for new-char-code = (if (and (> char-code 64) (< char-code
123)) (+ (mod (+ 13 char-code) 52) 65) char-code) do (setf (aref
new-string count) (code-char new-char-code)) finally (return
new-string)))
synthespian <···········@uol.com.br> writes:
> In article <···············@cley.com>, Tim Bradshaw <···@cley.com> wrote:
> > * Greg Menke wrote:
> >
> >> Perhaps Franz has a business arrangement where dealing with
> >> individuals isn't profitable and they get their money out of some few
> >> huge customers, but I wonder how many sales they end up losing just
> >> from making it hard for people to give them money.
> >
> > In Franz's defense (I am not a Franz customer), they did, some years
> > ago, sell a cheap (~ $1000 or maybe much less) Lisp for PCs (this was
> > pre linux being a major platform).
>
> You're out of your mind!
> You're probably an American, snug in your confortable Suburbia home, out of contact with the whole f*****g planet!
And you use a news posting software that is not capable of folding
lines at 78 characters. That is to say... you are not using GNUS on
Emacs.
> Do you know just __how much__ $1000 is in Venezuela? Russia? India? It's __not__ cheap!
> That f*****g alienated mentality, I guess, explains why commercial Lisp is years ahead of free-software lisp (which, as I see often here in c.l.l. is regarded as something for the sorry fellows who don't have a thousand bucks to spend).
> If there were more programmers like that, we wouldn't have
> GNU/Linux, GCC, C, C++, Perl , Python, etc.
GNU/Linux is coming out of some suburb of Boston and some suburb in
Scandinavia. Not bad for a product of rich-world living conditions.
> So you see, the problem with LISP is not so much the __language__, but lack of involvement in the community with developing free software, while they just remain seated waiting for Franz or whatever to deliver them their next fix.
> Hey, you don't watch world news on TV, do you?
And you do not know how to look for resources and seem to talk like a
troll who does not know what he is talking about.
>
> Regs,
> (and it's __not__ personal, ok? You just represent a certain
> mind set, that's all...)
... and you seem to represent the mindset of the bozos who went to
Genova and set the city on fire, falling in the trap set for them by
the real bad guys, who, incidentally, control most of the world media
*you* do not seem to understand.
Go away troll, and come back when you can claim to be at my left
(living in Brazil, does not make you automatically a good
guy. Remember Spain 82!)
Cheers
--
Marco Antoniotti ========================================================
NYU Courant Bioinformatics Group tel. +1 - 212 - 998 3488
719 Broadway 12th Floor fax +1 - 212 - 995 4122
New York, NY 10003, USA http://bioinformatics.cat.nyu.edu
"Hello New York! We'll do what we can!"
Bill Murray in `Ghostbusters'.
* synthespian wrote:
> You're out of your mind!
> You're probably an American, snug in your confortable Suburbia
> home, out of contact with the whole f*****g planet!
Well, no, I'm not, sorry to disappoint you.
> Do you know just __how much__ $1000 is in Venezuela? Russia?
> India? It's __not__ cheap!
Yes, I do know. I also understand some economics unlike you,
obviously.
> So you see, the problem with LISP is not so much the
> __language__, but lack of involvement in the community with
> developing free software, while they just remain seated
> waiting for Franz or whatever to deliver them their next fix.
Right, and fools like you whining away really helps this situation
doesn't it? Perhaps you should go away and contribute some free
software, like I have?
> Hey, you don't watch world news on TV, do you?
Actually I generally listen to the wireless, my TV is kind of worn
out, and the wireless news is generally better quality anyway.
--tim
synthespian wrote:
> > In Franz's defense (I am not a Franz customer), they did, some years
> > ago, sell a cheap (~ $1000 or maybe much less) Lisp for PCs (this was
> > pre linux being a major platform).
>
> You're out of your mind!
> You're probably an American, snug in your confortable Suburbia home, out of contact with the whole f*****g planet!
> Do you know just __how much__ $1000 is in Venezuela? Russia? India? It's __not__ cheap!
This is really a very high price in this aereas (and nobody should wonder when they foster software-piracy then). And sometimes it is even much more worse, because the local dealer in this places gets an extra amount of money for distributing the
software in his country; that means (under some worse cases) USD 1000.- sums up to USD 2000.-.
I personally would not mind if the prices of a copy would only costs lets say USD 100.- in this countries; even we in the West would have to pay USD 1000.-
S. Gonzi
synthespian <···········@uol.com.br> writes:
> In article <···············@cley.com>, Tim Bradshaw <···@cley.com> wrote:
> > * Greg Menke wrote:
> >
> >> Perhaps Franz has a business arrangement where dealing with
> >> individuals isn't profitable and they get their money out of some few
> >> huge customers, but I wonder how many sales they end up losing just
> >> from making it hard for people to give them money.
> >
> > In Franz's defense (I am not a Franz customer), they did, some years
> > ago, sell a cheap (~ $1000 or maybe much less) Lisp for PCs (this was
> > pre linux being a major platform).
>
> You're out of your mind!
Well I doubt the OP is, but you are definitly.
> You're probably an American, snug in your confortable Suburbia home, out of contact with the whole f*****g planet!
> Do you know just __how much__ $1000 is in Venezuela? Russia?
>India? It's __not__ cheap!
Wow how much does a VW Beatle cost wow $10000, you have no idea on how
much that is there or there of there.
> That f*****g alienated mentality, I guess, explains why commercial
> Lisp is years ahead of free-software lisp (which, as I see often here
> in c.l.l. is regarded as something for the sorry fellows who don't
> have a thousand bucks to spend).
Well how much is Borlands software stuff for Professionals? How much
for IBMs software how much for SAP. Does that harm in anyway the
success of their products? Well of course a Lisp must be free
obviously.
> If there were more programmers like that, we wouldn't have
> GNU/Linux, GCC, C, C++, Perl , Python, etc.
Well there are more Lisp implementations available than from any other
language, there are at least CMUCL and CLISP and ECLS which try to be
ANSI-Commoin Lisp what do you want more. Well maybe there are not
fancy GUIS available well but ther is CormanLisp and LispWorks
available which are both reasonable priced software pieces. So check
the facts before starting flames.
Friedrich
Friedrich Dominicus wrote:
>
> > If there were more programmers like that, we wouldn't have
> > GNU/Linux, GCC, C, C++, Perl , Python, etc.
> Well there are more Lisp implementations available than from any other
> language, there are at least CMUCL and CLISP and ECLS which try to be
> ANSI-Commoin Lisp what do you want more.
If the case he wants to start a business based on Lisp programming? I personally would never start this business based on
CMUCL or CLISP (I do not know ECLS).
And a software developer-company in a Third-World country does not automatically make more money than the rest in this
specific country (except the main company is in the US or Europe and tries to develop software in a country where the wages
are ridiculous low).
> Well maybe there are not
> fancy GUIS available well but ther is CormanLisp and LispWorks
> available which are both reasonable priced software pieces. So check
> the facts before starting flames.
Even in Europe we would say USD 1000.- are not that cheap (assuming the average wages in a contry are about USD 1700.- netto).
What should one think when he earns maybe USD 200,. per month when he has to buy a software product which actually costs USD
1000.-.
S. Gonzi
> * In message <·················@kfunigraz.ac.at>
> * On the subject of "Re: Free Lisp with GUI"
> * Sent on Mon, 21 Jan 2002 17:58:23 +0100
> * Honorable Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
>
> I personally would never start this business based on
> CMUCL or CLISP (I do not know ECLS).
ITA Software runs on CMUCL and makes money.
Paul Graham made $50mln on CLISP (viaweb --> yahoo store)
Flashy GUI IDE's are not everything.
It might be that the (falsely perceived) lack of flashy GUI IDE, in a
certain perverse way, benefits CL by keeping away people who do not
understand that...
"... a computer language is not just a way of getting a computer to
perform operations, but rather ... it is a novel formal medium for
expressing ideas about methodology"
Abelson/Sussman "Structure and Interpretation of Computer Programs".
--
Sam Steingold (http://www.podval.org/~sds)
Keep Jerusalem united! <http://www.onejerusalem.org/Petition.asp>
Read, think and remember! <http://www.iris.org.il> <http://www.memri.org/>
"Syntactic sugar causes cancer of the semicolon." -Alan Perlis
Sam Steingold wrote:
>
> > * In message <·················@kfunigraz.ac.at>
> > * On the subject of "Re: Free Lisp with GUI"
> > * Sent on Mon, 21 Jan 2002 17:58:23 +0100
> > * Honorable Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
> >
> > I personally would never start this business based on
> > CMUCL or CLISP (I do not know ECLS).
>
> ITA Software runs on CMUCL and makes money.
CMUCL or ACL?
From Franz:
http://www.franz.com/success/customer_apps/data_mining/itastory.lhtml
"ITA Software selected Allegro CL as their development environment, and
they currently run it on several different platforms. "We use Lisp for
the high level structure, in conjunction with a variety of other
languages such as C and Java throughout the application." Explains
Wertheimer. "We've been pleased with Allegro CL's strong foreign
function interfaces, powerful compiler, and multi-platform support." He
adds."
Kenny
clinisys
Kenny Tilton <·······@nyc.rr.com> writes:
> Sam Steingold wrote:
> >
> > > * In message <·················@kfunigraz.ac.at>
> > > * On the subject of "Re: Free Lisp with GUI"
> > > * Sent on Mon, 21 Jan 2002 17:58:23 +0100
> > > * Honorable Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
> > >
> > > I personally would never start this business based on
> > > CMUCL or CLISP (I do not know ECLS).
> >
> > ITA Software runs on CMUCL and makes money.
>
> CMUCL or ACL?
From the posting on Graham's website, it seems they are using both.
Regs, Pierre.
--
Pierre R. Mai <····@acm.org> http://www.pmsf.de/pmai/
The most likely way for the world to be destroyed, most experts agree,
is by accident. That's where we come in; we're computer professionals.
We cause accidents. -- Nathaniel Borenstein
* Sam Steingold wrote:D
> Flashy GUI IDE's are not everything.
Does perl have a flashy GUI IDE? if it does do 98% of people writing
perl use it or emacs/vi and xterm?
Hmmm.
--tim
* Siegfried Gonzi wrote:
> If the case he wants to start a business based on Lisp programming?
> I personally would never start this business based on CMUCL or CLISP
> (I do not know ECLS).
viaweb used clisp.
--tim
In article <·················@kfunigraz.ac.at>, Siegfried Gonzi wrote:
> Friedrich Dominicus wrote:
>
>> Well there are more Lisp implementations available than from any other
>> language, there are at least CMUCL and CLISP and ECLS which try to be
>> ANSI-Commoin Lisp what do you want more.
>
> If the case he wants to start a business based on Lisp programming? I
> personally would never start this business based on CMUCL or CLISP
> (I do not know ECLS).
Why not? I have used CMUCL a lot and there is nothing wrong with it
AFAICT.
Regards,
--
Nils Goesche
"Don't ask for whom the <CTRL-G> tolls."
PGP key ID 0x42B32FC9
* Nils Goesche wrote:
> Why not? I have used CMUCL a lot and there is nothing wrong with it
> AFAICT.
Well, it would bring up the inconvenient truth that there are multiple
high-quality zero-cost-to-use CL implementations and thus destroy the
thesis that Lisp is dying because you can't get a cheap implementation
and people like me don't care about this because we're all arrogant
Americans (I mean, come on, with a name like Nils with a second name
that almost certainly has an umlauted-o in it, and an email address
that ends in .de you're *bound* to be an American, obviously). It
wouldn't do to let facts get in the way of a nice rant about Lisp
being dead and Lisp hackers all being American suburban morons, would
it?
Of course Lisp is dead, has been since at least the mid 80s.
--tim
Tim Bradshaw wrote:
> * Nils Goesche wrote:
>
>
>> Why not? I have used CMUCL a lot and there is nothing wrong with it
>> AFAICT.
>
>
> Well, it would bring up the inconvenient truth that there are multiple
> high-quality zero-cost-to-use CL implementations and thus destroy the
> thesis that Lisp is dying because you can't get a cheap implementation
> and people like me don't care about this because we're all arrogant
> Americans (I mean, come on, with a name like Nils with a second name
> that almost certainly has an umlauted-o in it, and an email address
> that ends in .de you're *bound* to be an American, obviously). It
> wouldn't do to let facts get in the way of a nice rant about Lisp
> being dead and Lisp hackers all being American suburban morons, would
> it?
>
> Of course Lisp is dead, has been since at least the mid 80s.
>
> --tim
Yeah, whatever.
Fact is, you can put up a whole business on free software, running
Apache, Perl, and Linux.
Now, I'm new to LISP, but I find it sad that some of you Lispers find it
OK that you have to resort to LispWorks and Franz Allegro to do your
thing, and that you __prefer__ to use those over, say, CMU CL.
__That__ is what I think shows the alienated mentality of some in this
community. Why don't you use * only * free Lisps? Hell, I know lots of
people who wouldn't use Micro$oft for anything, because they feel it's a
bad choice, and would rather use free software * only *.
Not like that here, huh? Why? Do you ever stop to __think__ about it?
Damn, LISP is so good and so old, so why didn't it beat all the other
languages?
hy
Henry Lebowzki <···@uol.com.br> wrote in message news:<················@uol.com.br>...
> Damn, LISP is so good and so old, so why didn't it beat all the other
> languages?
That's like saying Thoreau was such a good author, and asking why his
books aren't in the top five percent. The answer is simple: They are
in the top five percent.
Henry Lebowzki <···@uol.com.br> writes:
>
> Yeah, whatever.
> Fact is, you can put up a whole business on free software, running
> Apache, Perl, and Linux.
>
> Now, I'm new to LISP, but I find it sad that some of you Lispers find
> it OK that you have to resort to LispWorks and Franz Allegro to do
> your thing, and that you __prefer__ to use those over, say, CMU CL.
What does the fact that you can build you business on free software
has to do with preferring LispWorks over the free alternatives? Well I
know why I prefer LispWorks so if you don't find that acceptable than
it's fine for me that you take a free one. Why do I have to use CMU
CL?
Once in this thread it was ased for a Lisp with runs on Linux and
Windows with GUI-stuff. Well CAPI does run there. And it's part of
LispWorks. So the simple answer is LispWorks provided those things but
obviously it's not so for the free alternatives. So why shouldn't one
use LispWorks?
>
> __That__ is what I think shows the alienated mentality of some in this
> community. Why don't you use * only * free Lisps?
What are you? It's my choice if I want to use free Lisps I use them if
I do not want I don't.
> Hell, I know lots of
> people who wouldn't use Micro$oft for anything, because they feel it's
> a bad choice, and would rather use free software * only *.
Well than use free stuff. And go along with it. I like LispWorks and I
use it Dot.
>
> Not like that here, huh? Why? Do you ever stop to __think__ about
>it?
Go along as you like but don't blame others if the decide
differently. Ever heard of something like freedom?
> Damn, LISP is so good and so old, so why didn't it beat all the other
> languages?
Wel i could give a harsh answer. But I doubt it would make a
difference. So I propose you check you attitude.
Friedrich
Henry Lebowzki <···@uol.com.br> writes:
> Fact is, you can put up a whole business on free software, running
> Apache, Perl, and Linux.
There are many more choices, including the ones involving Lisp.
That's a fact, as well.
> Now, I'm new to LISP, but I find it sad that some of you Lispers
> find it OK that you have to resort to LispWorks and Franz Allegro to
> do your thing, and that you __prefer__ to use those over, say, CMU
> CL.
What's sad about that? People use what they know is best for them,
they do not resort to something.
> __That__ is what I think shows the alienated mentality of some in this
> community.
Where does this mentality come from, if there is such a thing? Or,
where is it directed to?
> Why don't you use * only * free Lisps?
Why should people use _only_ something? There are people whe have
brains, and can decide for themselves what are their needs and what
can satisfy them. And they can choos. whatever they think is best.
That's what freedom is about, isn't it?
> Hell, I know lots of people who wouldn't use Micro$oft for anything,
> because they feel it's a bad choice, and would rather use free
> software * only *.
Where does that feeling come from? Do these people even know why they
chose _that_ something, _only_? Have they looked at alternatives, any
at all? Or is the definition of software split into two categories,
"free software" and Microsoft software, that is the exhaustive
partitioning of the software term?
Or is writing software for food something ethically unacceptable to
you, too?
> Not like that here, huh? Why? Do you ever stop to __think__ about it?
Yes. And you _should_, too. You can easily find the thoughs of many
intelligent people of this grop using google. I can give you some
shortcuts (message-id's):
<··············@orion.bln.pmsf.de>
<················@naggum.net>
<···············@world.std.com>
And you should read the complete threads if you wanted to understand
the topics involved, not just find the support your current view on
the topic while dismissing all other views.
> Damn, LISP is so good and so old, so why didn't it beat all the other
> languages?
I does. You just don't see it, yet. Open your eyes, free your mind,
learn Common Lisp, and you will see it, too.
--
Janis Dzerins
Eat shit -- billions of flies can't be wrong.
* Henry Lebowzki wrote:
> Yeah, whatever.
> Fact is, you can put up a whole business on free software, running
> Apache, Perl, and Linux.
> Now, I'm new to LISP, but I find it sad that some of you Lispers find
> it OK that you have to resort to LispWorks and Franz Allegro to do
> your thing, and that you __prefer__ to use those over, say, CMU CL.
So, for instance, viaweb never existed. Right.
> Damn, LISP is so good and so old, so why didn't it beat all the other
> languages?
Worse is better. I guess you haven't read that.
Henry Lebowzki wrote:
> Now, I'm new to LISP, but I find it sad that some of you Lispers find it
> OK that you have to resort to LispWorks and Franz Allegro to do your
> thing, and that you __prefer__ to use those over, say, CMU CL.
> __That__ is what I think shows the alienated mentality of some in this
> community. Why don't you use * only * free Lisps?
[etc]
Um, because not everyone who uses CL is a free software
enthusiast. Just like not everyone who uses C++ is, and
not everyone who uses Perl is, and so on and so on.
> Not like that here, huh? Why? Do you ever stop to __think__ about it?
> Damn, LISP is so good and so old, so why didn't it beat all the other
> languages?
______
/ \
.' PLEASE `.
| DO NOT | _____
| FEED THE | ,'.....`.
`. TROLLS ,' ,'........ )
\_ _/ |........ ,'
| | `. .... _/
| | ,'.,'-'
| | /../
| | ,'.,'
| | /../
. | | /..'
.\_\| |/_/,
___ | | ___
. `--' .
. .
--
Gareth McCaughan ················@pobox.com
.sig under construc
Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
> If the case he wants to start a business based on Lisp programming?
> I personally would never start this business based on CMUCL or CLISP
> (I do not know ECLS).
I cannot argue against your personal convictions, but I can point out
that there are others who have based (parts of) their businesses on
e.g. CMU CL. As just one case in point[1], CMU CL currently is a
mainstay of our own business (large-scale industrial ERP/SCM
server-based simulations and related tools), and I have no trouble
sleeping at night.
Of course 98% of our code-base is pure ANSI CL (+ some MOP stuff),
with only a couple of thousand LoC that are really implementation
dependent, e.g. command-line processing, some parts of our secure code
updating service, some lowest level socket code, database interfaces,
etc.
In a pinch, moving to another implementation will take at most a
couple of weeks, probably less. We take great care to keep as much
code as is reasonable portable to the main implementations that are
relevant to our problem domain.
Of course CMU CL is probably not going to be your implementation of
choice when it comes to implementing GUI frontends on Windows, but for
lots of other application areas, CMU CL can be a good fit indeed.
Cudos to the legions of CMU CL/Spice Lisp developers and maintainers
over the years for producing very high quality code, which provides a
very stable base for our products[2].
Regs, Pierre.
Footnotes:
[1] For other examples, see ITA Software, and several other companies
that have posted to the CMU CL mailing-lists in the past.
[2] Disclaimer: I'm currently doing my very small bit at helping
maintain CMU CL, so obviously I'm not including myself in that
paragraph... ;)
--
Pierre R. Mai <····@acm.org> http://www.pmsf.de/pmai/
The most likely way for the world to be destroyed, most experts agree,
is by accident. That's where we come in; we're computer professionals.
We cause accidents. -- Nathaniel Borenstein
>>>>> "Pierre" == Pierre R Mai <····@acm.org> writes:
Pierre> Cudos to the legions of CMU CL/Spice Lisp developers and maintainers
Pierre> over the years for producing very high quality code, which provides a
Pierre> very stable base for our products[2].
Pierre> Regs, Pierre.
Pierre> Footnotes:
Pierre> [1] For other examples, see ITA Software, and several other companies
Pierre> that have posted to the CMU CL mailing-lists in the past.
Pierre> [2] Disclaimer: I'm currently doing my very small bit at helping
Pierre> maintain CMU CL, so obviously I'm not including myself in that
Pierre> paragraph... ;)
But you should, because you have done a good job at tracking down
bugs, fixing things, improving things. Certainly no less than any of
the other current CMUCL developers.
Ray
Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
> Friedrich Dominicus wrote:
>
> >
> > > If there were more programmers like that, we wouldn't have
> > > GNU/Linux, GCC, C, C++, Perl , Python, etc.
> > Well there are more Lisp implementations available than from any other
> > language, there are at least CMUCL and CLISP and ECLS which try to be
> > ANSI-Commoin Lisp what do you want more.
>
> If the case he wants to start a business based on Lisp programming? I personally would never start this business based on
> CMUCL or CLISP (I do not know ECLS).
That is your choice. I know quite a few companies which do not have
any problem relying GCC, Python, Perl, Apache and whatever is around.
>
> And a software developer-company in a Third-World country does not automatically make more money than the rest in this
> specific country (except the main company is in the US or Europe and tries to develop software in a country where the wages
> are ridiculous low).
So what?
Let us assume that you are in a third world country, for progamm
development you need a bit more than "just" a Lisp. you need hardware
and software. So a decent computer will always cost above $1000. Well
he obviously needs one to get into business. Well than of course he
needs a development tool. Well they won't take gcc or the like
because you can not settle a business based on free alternatives, so
you can't use gcc obviously. What will he do? Well he probably will
have to buy a development environment. And than he/she will have to
pay for it the "usual" price. It seems ok to pay for M$-tools but
again it seems not obviously to pay for other IDES. Bad luck I guess.
>
>
>
> > Well maybe there are not
> > fancy GUIS available well but ther is CormanLisp and LispWorks
> > available which are both reasonable priced software pieces. So check
> > the facts before starting flames.
>
> Even in Europe we would say USD 1000.- are not that cheap (assuming
>the average wages in a contry are about USD 1700.- netto).
Well should I cite the prices for one progarm hour. It's work for one
to two day for professional software developers in Europe. That's a
fact. I just wonder again why it's ok for a job in industry to cost >
100 000 EUR. but a working place for a normal programmer shouild not
cost more than 10 000 EUR. What are those for relations?
> What should one think when he earns maybe USD 200,. per month when he has to buy a software product which actually costs USD
> 1000.-.
And how much is the computer? Will he be able to get away without one?
No. Will he get away with cheaper software? Maybe. But hardly if one
follows your suggestions.
Friedrich
Friedrich Dominicus wrote:
> Let us assume that you are in a third world country, for progamm
> development you need a bit more than "just" a Lisp. you need hardware
> and software. So a decent computer will always cost above $1000.
All this little pieces sum up!
> Well
> he obviously needs one to get into business. Well than of course he
> needs a development tool. Well they won't take gcc or the like
> because you can not settle a business based on free alternatives, so
> you can't use gcc obviously.
This is not true. I didn't say that free software-tools are not capable of handling business-projects. But, under my naive guess,
I develop a software and want to sell it, why should I then expect that anybody will buy it when the people are accustomed to have
free software around?
One now could make a caveat and say: "But first of all, a Third-World country should use the free tools; make money; and then they
can afford the more expensive tools".
But I think it is not that easy.
>
> > Even in Europe we would say USD 1000.- are not that cheap (assuming
> >the average wages in a contry are about USD 1700.- netto).
> Well should I cite the prices for one progarm hour. It's work for one
> to two day for professional software developers in Europe.
Please cite it! What software developers expect and what companies are ready to spent ist another topic. But normally
software-developers rate themselves way too hig.
> That's a
> fact. I just wonder again why it's ok for a job in industry to cost >
> 100 000 EUR. but a working place for a normal programmer shouild not
> cost more than 10 000 EUR. What are those for relations?
I am not sure what your figures here represent; but neither of them are reliable? You can count the worker who earns Euro
100.000.- per year with your fingers.
S. Gonzi
Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
> Friedrich Dominicus wrote:
>
> > That's a fact. I just wonder again why it's ok for a job in
> > industry to cost > 100 000 EUR. but a working place for a normal
> > programmer shouild not cost more than 10 000 EUR. What are those
> > for relations?
>
> I am not sure what your figures here represent; but neither of them
> are reliable? You can count the worker who earns Euro 100.000.- per
> year with your fingers.
If an employee costs an employer 100,000 EUR per year that doesn't
mean that the employee gets all this money - you should know this.
Edi.
"Dr. Edmund Weitz" wrote:
> > I am not sure what your figures here represent; but neither of them
> > are reliable? You can count the worker who earns Euro 100.000.- per
> > year with your fingers.
>
> If an employee costs an employer 100,000 EUR per year that doesn't
> mean that the employee gets all this money - you should know this.
I meant the gross amount (but not the complete amount for rents: farms,
buildings, office-rooms,...) and even then 50.000 EUR net are not the
wages which a worker can earn (or he is his own boss); and even a
programmer (lets say with 10 years experience) does not get this amount
(or he is in managment).
Before starting my PhD I had a conversation with a consultant; he said it
is ridiculous what people often expect on wages due to the believe that
they are programmers. But I can understand this, because every day a
politcian on televison says: "We need programmers".
S. Gonzi
* Siegfried Gonzi wrote:
> I meant the gross amount (but not the complete amount for rents:
> farms, buildings, office-rooms,...) and even then 50.000 EUR net are
> not the wages which a worker can earn (or he is his own boss); and
> even a programmer (lets say with 10 years experience) does not get
> this amount (or he is in managment).
In the UK, 50,000 euros is probably slightly above the average salary
for IT people, but not extreme. *Good* people, or people with
heavily-in-demand skills (like Oracle) get 80,000 and up...
--tim
Tim Bradshaw <···@cley.com> writes:
> * Siegfried Gonzi wrote:
>
> > I meant the gross amount (but not the complete amount for rents:
> > farms, buildings, office-rooms,...) and even then 50.000 EUR net are
> > not the wages which a worker can earn (or he is his own boss); and
> > even a programmer (lets say with 10 years experience) does not get
> > this amount (or he is in managment).
>
> In the UK, 50,000 euros is probably slightly above the average salary
> for IT people, but not extreme. *Good* people, or people with
> heavily-in-demand skills (like Oracle) get 80,000 and up...
There are lots of people in the US making these salaries, but even
more, who are skilled enough to do many of the same jobs, making
*much* lower wages. There are *states* (and I live in one) in the US
that are in the same situation as the UK, but with the US being a
little bigger than the EU, this is an important point, if we're
speaking internationally...
--
/|_ .-----------------------.
,' .\ / | No to Imperialist war |
,--' _,' | Wage class war! |
/ / `-----------------------'
( -. |
| ) |
(`-. '--.)
`. )----'
Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
> I meant the gross amount (but not the complete amount for rents:
> farms, buildings, office-rooms,...) and even then 50.000 EUR net are
> not the wages which a worker can earn (or he is his own boss); and
> even a programmer (lets say with 10 years experience) does not get
> this amount (or he is in managment).
According to
<http://www.heise.de/newsticker/data/anm-20.03.01-000/default.shtml>
(text in German), the _average_ wage for employed IT professionals in
Germany is above 50,000 EUR while the maximum is at about 110,000
EUR. Note that the real numbers are probably higher as the trade union
which published them used a basis of 35 hours per week (which in my
perception is not the usual working time in this industry).
Edi.
In article <··············@bird.agharta.de>, Dr. Edmund Weitz wrote:
> Note that the real numbers are probably higher as the trade union
> which published them used a basis of 35 hours per week (which in my
> perception is not the usual working time in this industry).
In reality or in the contracts? ;-|
Regards,
--
Nils Goesche
"Don't ask for whom the <CTRL-G> tolls."
PGP key ID 0x42B32FC9
Nils Goesche <······@cartan.de> writes:
> In article <··············@bird.agharta.de>, Dr. Edmund Weitz wrote:
> > Note that the real numbers are probably higher as the trade union
> > which published them used a basis of 35 hours per week (which in
> > my perception is not the usual working time in this industry).
>
> In reality or in the contracts? ;-|
I myself am a free-lancer, but the people I know from my working
experience usually have contracts saying they should work 40 hours per
week. The reality is that they work (much) more but that they don't
get paid for this overtime.
Note that I'm describing mostly small 'start-up' companies here. The
situation in big companies which pay standard rates _is_ different.
Edi.
"Dr. Edmund Weitz" wrote:
> According to
> <http://www.heise.de/newsticker/data/anm-20.03.01-000/default.shtml>
> (text in German), the _average_ wage for employed IT professionals in
> Germany is above 50,000 EUR while the maximum is at about 110,000
> EUR. Note that the real numbers are probably higher as the trade union
> which published them used a basis of 35 hours per week (which in my
> perception is not the usual working time in this industry).
I know the heise-links. I believe this numebers a gross-numbers. I think
this is important to know.
S. Gonzi
Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
> "Dr. Edmund Weitz" wrote:
>
> > According to
> > <http://www.heise.de/newsticker/data/anm-20.03.01-000/default.shtml>
> > (text in German), the _average_ wage for employed IT professionals
> > in Germany is above 50,000 EUR while the maximum is at about
> > 110,000 EUR. Note that the real numbers are probably higher as the
> > trade union which published them used a basis of 35 hours per week
> > (which in my perception is not the usual working time in this
> > industry).
>
> I know the heise-links. I believe this numebers a gross-numbers. I
> think this is important to know.
Sure. You can only compare gross income because net income is
dependent on a couple of factors that have nothing to do with your
profession, your knowledge, or your experience.
Edi.
Siegfried Gonzi wrote:
> I meant the gross amount (but not the complete amount for rents: farms,
> buildings, office-rooms,...) and even then 50.000 EUR net are not the
> wages which a worker can earn (or he is his own boss); and even a
> programmer (lets say with 10 years experience) does not get this amount
> (or he is in managment).
How have you derived these figures? In the UK, for example: an
experienced c, c++ or java developer would on average be paid
about �30.000 gross[1] = 48.0000 euro. For more experienced
programmers or in certain areas of the country -- specifically
London -- the average will rise to about �42.000 = 68.000 Euro.
There will be a sizable percentage of programmers earning more
than this.
As the tax rate in the UK are about 35% -- including health care,
income tax, blah, blah, blan -- which then gives a net *average*
of about 30.000-45.000 Euro.
:)w
[1] The salaries survey that I borrowed this information from is
available at lifesupportal.com/cgi-bin/php.cgi/salarysurvey.php
and is "based on ... the salaries that real IT professionals have
actually accepted."
* Siegfried Gonzi wrote:
> I am not sure what your figures here represent; but neither of them
> are reliable? You can count the worker who earns Euro 100.000.- per
> year with your fingers.
Daily costs *to the employer* are likely in the region of 300 to 1000
pounds (convert to euros as you wish). at 220 days / year, that's
66,000 to 220,000 pounds / year. Of course *not* all of this, or
anywhere like all of it goes to the person employed. Anyone who has
done contracting work, not done too much tax evasion *and* done the
figures reasonably well will see that there is a fairly large
overhead (a factor of 2 or more).
--tim
Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
>
> >
> > > Even in Europe we would say USD 1000.- are not that cheap (assuming
> > >the average wages in a contry are about USD 1700.- netto).
> > Well should I cite the prices for one progarm hour. It's work for one
> > to two day for professional software developers in Europe.
>
> Please cite it! What software developers expect and what companies are ready to spent ist another topic. But normally
> software-developers rate themselves way too hig.
Well I'm talking about what is charged per hour. Now I have one
pricelist here (some years old) for implementation they will charge
240 DM around 120 EUR. Now picking a calculator gives
960 EUR/day. Well LispWorks Prof. costs 900 EUR. So it's not even one
programmers work day. Well this can't be hardly be told to be too
expensive.
>
> > That's a
> > fact. I just wonder again why it's ok for a job in industry to cost >
>> > 100 000 EUR. but a working place for a normal programmer shouild not
> > cost more than 10 000 EUR. What are those for relations?
>
> I am not sure what your figures here represent; but neither of them are reliable? You can count the worker who earns Euro
> 100.000.- per year with your fingers.
I was talking about a working place. Tha's just the things to provide
one worker with such a thing. Of course one have to add the payments
too, but I left that open for now.
Friedrich
Friedrich Dominicus wrote:
> Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
>
> > Please cite it! What software developers expect and what companies are ready to spent ist another topic. But normally
> > software-developers rate themselves way too hig.
> Well I'm talking about what is charged per hour. Now I have one
> pricelist here (some years old) for implementation they will charge
> 240 DM around 120 EUR. Now picking a calculator gives
> 960 EUR/day. Well LispWorks Prof. costs 900 EUR. So it's not even one
> programmers work day. Well this can't be hardly be told to be too
> expensive.
Only one second: where did I write that LispWorks is too expensive? Oh, too expensive for a worker with USD 200.- per month? If
you do not know it: nobody is discussing here prices in the West.
And I do also not believe that the dealers of Lisp are responsible for the Third-World. But sometimes I really feel that some
people believe the world is the same on every damn place on this earth.
Are you Bildzeitungsleser? Answer per private mail please, because comp.lang.lisp cannot solve the problems in the Third-World,
neither can it Lisp.
S. Gonzi
Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
> Friedrich Dominicus wrote:
>
> > Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
> >
> > > Please cite it! What software developers expect and what companies are ready to spent ist another topic. But normally
> > > software-developers rate themselves way too hig.
> > Well I'm talking about what is charged per hour. Now I have one
> > pricelist here (some years old) for implementation they will charge
> > 240 DM around 120 EUR. Now picking a calculator gives
> > 960 EUR/day. Well LispWorks Prof. costs 900 EUR. So it's not even one
> > programmers work day. Well this can't be hardly be told to be too
> > expensive.
>
> Only one second: where did I write that LispWorks is too expensive?
Well I just wanted to get those things right.
> Oh, too expensive for a worker with USD 200.- per month?
It depends on what one want to do with it.
> If
> you do not know it: nobody is discussing here prices in the West.
Don't we? Well than I must miss something. We were talking about
prices. And I gave some figures on how much one Lisp costs. And I
can't see that it's expensive.
>
> Are you Bildzeitungsleser?
How sensible you are.
>Answer per private mail please, because comp.lang.lisp cannot solve the problems in the Third-World,
> neither can it Lisp.
Well I won't discuss problems of the third world here. I put out some
figures for "developed" countries and this is the market for any
vendor. If you do not sell well there you hardly won't get big.
Friedrich
Siegfried Gonzi <···············@kfunigraz.ac.at> writes:
> > > Even in Europe we would say USD 1000.- are not that cheap (assuming
> > >the average wages in a contry are about USD 1700.- netto).
> > Well should I cite the prices for one progarm hour. It's work for one
> > to two day for professional software developers in Europe.
>
> Please cite it! What software developers expect and what companies
> are ready to spent ist another topic. But normally
> software-developers rate themselves way too hig.
Current studies show that on average programmers earn around 45000EUR
in Germany, with professional software developers averaging around
58000 EUR (total income, including benefits, etc.). Assuming 250 work
days, minus 6*5 days vacation, etc., this comes out to around
58000/220 =~ 264 EUR/day.
But that isn't what it costs the company to employ that person, which
is often nearly double that amount, i.e. 528 EUR/day.
Correspondingly, it isn't unheard of for companies to pay free-lancers
700-1000 EUR per day (I have been personal witness to such
transactions).
That means that 1200 EUR is really the cost of one-two days of a
professional developer.
> > That's a
> > fact. I just wonder again why it's ok for a job in industry to cost >
> > 100 000 EUR. but a working place for a normal programmer shouild not
> > cost more than 10 000 EUR. What are those for relations?
>
> I am not sure what your figures here represent; but neither of them are reliable? You can count the worker who earns Euro
> 100.000.- per year with your fingers.
No one has said anything about someone earning 100000 EUR. The cited
figure is what it costs a company to set up the working environment of
a worker in lots of tool-intensive fields of business.
But even in the case of normal office workers (i.e. software
developers), you will have to spend lots of money to set up and
maintain the workers place of work. Even using fairly cheap desks,
chairs, lamps, and computers the initial capital outlay surpasses 1000
EUR very, very quickly. Rent alone, even at the cheapest rates
available in e.g. Berlin (which is one of the lowest cost cities in
Germany when it comes to office rents) will cost you around 30 EUR per
person/month, and easily 4-10 times as much, and doesn't include
shared office space, heating, electricity, taxes, etc.
Employing people in industrial nations is a high-cost endeavour, hence
even something that makes your people less than 1% more productive is
cheap at 1000 EUR per person.
And that is the target market of most software development tool
providers. They don't target the high-school teacher creating
fantastically useful teaching software, they don't target the hobbyist
creating small share-ware utilities as a means to earn a couple of
hundred dollars extra per year (and lots of fun), they don't target
developing nations, etc. At the most they will provide low-cost
restricted versions of their normal products for those markets.
Take a look at most SmallTalk companies, and/or IBM, or Borland. Most
of their products are in the plus 1000 EUR category. E.g. Borland
Delphi 6:
Product incl. VAT excl. VAT
- Enterprise Version: 3864.00 EUR 3331.03 EUR
- Professional Version: 1277.00 EUR 1100.86 EUR
- Personal Version: 148.00 EUR 127.59 EUR
The personal version only allows non-commercial use, has lots of hairy
licencing restrictions, and doesn't include most of the useful
additional components, like DB access, HTML/HTTP, advanced editor
and debugger features, cross-platform component library, ActiveX, and
lots of other stuff...
If you compare that to e.g. Xanalys' pricing, I think you'll agree
that Xanalys compares favourably.
Several years back, Borland did have another focus, less competition
from under-priced stuff like Java, and correspondingly cheaper
prices. But times have changed, and Borland had to react. You can't
expect to compete against heavily subsidised stuff, and still make a
profit. So they targeted the higher-cost segment of the market, and
seem to be doing quite well. I'm highly sceptical that Borland could
exist in the "Turbo Pascal" segment of the market any longer, with
people getting all sorts of stuff for free in that segment.
[ BTW, this isn't intended to slam Delphi, which IMHO, despite earlier
problems in its implementational quality, is a reasonable choice
when it comes to creating GUI-heavy Windows applications. You can
even do fairly intricate stuff with this, e.g. in an earlier life, I
once created reusable components that more or less implemented
something similar to Kenny Tilton's Semaphores/Cells concept, in
truly drag and drop fashion, which eased the creation and
maintenance of financial calculations software tremendously. ]
Regs, Pierre.
--
Pierre R. Mai <····@acm.org> http://www.pmsf.de/pmai/
The most likely way for the world to be destroyed, most experts agree,
is by accident. That's where we come in; we're computer professionals.
We cause accidents. -- Nathaniel Borenstein
* Pierre R Mai wrote:
[A really good article, I think]
> And that is the target market of most software development tool
> providers. They don't target the high-school teacher creating
> fantastically useful teaching software, they don't target the hobbyist
> creating small share-ware utilities as a means to earn a couple of
> hundred dollars extra per year (and lots of fun), they don't target
> developing nations, etc. At the most they will provide low-cost
> restricted versions of their normal products for those markets.
The other significant market that software development tool vendors
target is cross-subsidizing their own products. In a slightly
different area you can look at something like Solaris: how much money
does Sun make from Solaris? Well superficially the answer should be
`nothing' because it's free. But actually they sell these nice
machines which Solaris runs on really well, and they make a good deal
of money from the machines. And no one (or very few people) would buy
a sunfire without Solaris - indeed most of the cool, important, stuff
in these machines like hot-swappable everything and partitioning and
so on is meaningless without software.
Java is an even better example. Java environments are cheap, or free.
And they have all this GUI stuff. But they're cheap because Sun are
losing money hand over fist on them to make money elsewhere. It's a
little less simple to see where, since Java is cross platform and
therefore doesn't directly sell Sun hardware. But the issue here is
that Sun see (correctly) that anything non-MS specific will benefit
them, so they're willing to fund systems which are not Sun-specific.
.NET, again. Do you think MS will make a penny on .NET? They won't
but they'll make an awful lot out of the increased market share it
gets for Windows. Go read Big Blue to see how this worked for IBM 25
years ago.
Loads of `free' software is like this too: IBM are pumping money into
Linux because it sells their hardware and their support services (and
there's another secret battle here - IBM and Sun are fighting for the
mainframe market, IBM with large machines running lots of Linux images
on top of their amazing proprietary OS (OS/390? I forget the current
name), and Sun with their large machines running Solaris). Loads of
companies employ free software authors because what they do benefits
some other business activity. They don't do it out of the goodness of
their hearts.
The Lisp vendors sell Lisp. If I buy a license to system x, and as a
result buy 10 enterprise-class machines from Sun or IBM, the Lisp
vendor doesn't get paid a penny. So they have to actually charge
enough for their product to make money.
It should not be surprising that a company that needs to make a profit
on an item has to charge more for it than a company which is making a
loss on it.
--tim
Friedrich Dominicus <·····@q-software-solutions.com> writes:
> Well there are more Lisp implementations available than from any other
> language, there are at least CMUCL and CLISP and ECLS which try to be
> ANSI-Commoin Lisp what do you want more. Well maybe there are not
> fancy GUIS available
I don't know, Garnet seems pretty fancy to me. You can build GUIs
visually with CLG and Glade. And Motif isn't particularly fancy, but
it's obviously industrial-strength.
--
/|_ .-----------------------.
,' .\ / | No to Imperialist war |
,--' _,' | Wage class war! |
/ / `-----------------------'
( -. |
| ) |
(`-. '--.)
`. )----'
On 21 Jan 2002, Friedrich Dominicus wrote:
[snip]
> > If there were more programmers like that, we wouldn't have
> > GNU/Linux, GCC, C, C++, Perl , Python, etc.
> Well there are more Lisp implementations available than from any other
> language,
Probably not, actually. Prologs are a dime a dozen. Schemes! They litter
the ground. There are several Smalltalks, even if we're just talking free
implementations.
Of course, none of these is overall as challenging to implement fully
*and* well as Common Lisp. So that's impressive enough in my book.
[snip]
> available which are both reasonable priced software pieces. So check
> the facts before starting flames.
Or dousing them :) Though, this poster was annoying enough to cause me to
forgive a multitude of sins in replies. :)
It's really interesting how these clueless trolls don't know the *massive*
amounts of code and effort sharing tha has occured in the CL community
over the years. And what a massive undertaking a full and good CL
implementation is. Etc. :)
Cheers,
Bijan Parsia.
On Mon, 21 Jan 2002 14:01:19 -0300, synthespian <···········@uol.com.br>
wrote:
> free-software lisp (which, as I see often here in c.l.l. is regarded
> as something for the sorry fellows who don't have a thousand bucks to spend).
For the record, CLISP (GPL) has been used for developing a successful
e-commerce site whose owning company was worth more than 50M$, and CMU CL
(public domain) was also used for an important commercial airline
reservation system.
Paolo
--
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://web.mclink.it/amoroso/ency/README
[http://cvs2.cons.org:8000/cmucl/doc/EncyCMUCLopedia/]
On 19 Jan 2002, Christopher Browne wrote:
[snip]
> - If they have customers ready to balk at high prices, and go
> elsewhere, _THAT_ would be a good reason to cut prices.
>
> - If they figured that by cutting the price in 6, that they would
> gain 6x as many customers, _THAT_ would be a good reason to cut
> prices to 1/6 the present levels.
Actually, probably not. 6x the customers/sales can mean much more than 6x
the overhead (support, etc.). Plus, 6x the customers might mean quite a
bit less than 6x sales.
Over on comp.lang.smalltalk, the Cincom smalltalk product manager gave a
pretty brutal breakdown of their pricing dilemmas. It was
horrid. (Cincom's pricing tactis are much like Franz's, FWIW. VisualWorks,
their flagship Smalltalk, almost died at its prior company with a more
"acceptible" pricing model.)
> (If the number is $6k, that would
> drop it to $1K, which would _still_ be daunting to anybody sitting
> at the low end of the "price preference" scale.)
So you end up with the worst of both: No money, lots of people to support,
and all the complainers still complaining :)
> But it is not at
> all obvious that this would happen.
Yes, then *really* no money, more people to support, all the old
complainers complaining, and everyone else complaining that sucky Franz
management drove the company and the system into the dirt. woo hoo :)
> I might _wish_ that I could buy ACL for $200; I might even hold
> tenaciously to a refusal to pay more than that. For the time being,
> that's liable to lead to them not selling me a copy of ACL, and my not
> paying them $200 for it. :-).
One thing I've wondered aloud at over in c.l.smalltalk was if there was a
reasonable small developer sweet spot that wouldn't be too much
trouble. CodeWarrior had education pricing limitied to freeware and
shareware products. I can see a hobbyist market being a useful thing for
many implementers. Not sure about the super high end guys though...I mean,
would it help Franz to have a bazillion little hackers peddling Allegro
bases microutilities? Maybe, maybe not :)
Aside from shareware, there are quite a few of small time developers. One
fellow over on comp.lang.smalltalk develops medical billing software for
doctors (very specialized). He may have 10 customers and not a *huge*
amount of revenue. Langauges like common lisp or smalltalk can be ideal
for such one person, but tricky jobs. They *are* marginal operations,
though.
The trick is whether the price can be made worth it for Franz without
pissing off their real customers.
Cheers,
Bijan Parsia.
Bijan Parsia wrote:
> One thing I've wondered aloud at over in c.l.smalltalk was if there was a
> reasonable small developer sweet spot that wouldn't be too much
> trouble. CodeWarrior had education pricing limitied to freeware and
> shareware products.
For the Mac one can also buy Macintosh Common Lisp for a student price of
about USD 100.- (or therelike). I think he then gets the full version but
support and printed-manuals.
Normally every software vendor sells at least a low cost version of their
products. They sell it and nothing more, because a low cost version does not
include service/hotline/support. I would say it is not too much work for the
software vendor to sell a low cost copy to a student or individual.
The chicken egg problem: if Common Lisp is really that great concerning the
productivity of a programmer or development team why then should one pay
overgenerous prices for a hotline/support which should not happen very often;
otherwise: how can one beeing sure -beforehand- that the high single license
prices are worth the money and that a call for support will be not required
too often (due to the productivity of Lisp)?
But I have to mention that I do believe that it is not the fault of the
vendors that there are not more high quality Lisp versions for the PC platform
out there.
S. Gonzi
On Sun, 20 Jan 2002, Siegfried Gonzi wrote:
> Bijan Parsia wrote:
>
> > One thing I've wondered aloud at over in c.l.smalltalk was if there was a
> > reasonable small developer sweet spot that wouldn't be too much
> > trouble. CodeWarrior had education pricing limitied to freeware and
> > shareware products.
>
> For the Mac one can also buy Macintosh Common Lisp for a student price of
> about USD 100.- (or therelike). I think he then gets the full version but
> support and printed-manuals.
They may still be selling their "newstand" and "champion" editions.
*But*, they have a relatively cheap full price. The differnce between the
uber cheap and the full price is not so great.
> Normally every software vendor sells at least a low cost version of their
> products.
Well, depends on what you mean by normal. If you're small and your main
business is selling custom contracts to large corps (for the most part),
then not having a low cost version seems standard. That's working from a
sample of 3 or so :)
>They sell it and nothing more, because a low cost version does not
> include service/hotline/support. I would say it is not too much work for the
> software vendor to sell a low cost copy to a student or individual.
Well, you have to include marketing and sales handling, which might not be
too much, but also concerns from you're other customers about why the punk
kid gets it cheap.
It really seems to me doable, but I don't have access to all the details,
after all, so am hesitent to judge.
Cheers,
Bijan Parsia.
······@interaccess.com (Thaddeus L Olczyk) writes:
> On 17 Jan 2002 01:43:31 -0800, ··········@tfeb.org (Tim Bradshaw)
> wrote:
>
> >······@interaccess.com (Thaddeus L Olczyk) wrote in message news:<·················@nntp.interaccess.com>...
> >
> >> Producing freely distributed executables with both is ( my
> >> understanding ) highly problematic.
> >
> >For LispWorks at least this is not so for Windows and Linux: you can
> >generate freely distributable executables for both these platforms
> >from the commercial product (possibly with some constraints - there
> >may be stuff like CORBA or something you can't include). I think
> >there are royalties for the Unix platforms.
> >
> >--tim
> Well we begin with the subject-- the person was asking for *free*
> lisps. But lets just stop for a moment and assume that a person was
> willing to pay something.
Actually, the OP *was* willing to pay something "reasonable". Why do
you post shit like this?
> Remember this is stuff that the person would like to give away ( it
> was clipped but I did say open source/free ).
Yeah, and on Windows, this costs money. Like everything on Windows.
This is *not* a language issue. $1800 ain't cheap, but you get a lot
for them 1800. There's also Corman Lisp, which is cheap. Oh, wait,
the OP wants this on Linux *and* Windows? Fine, use CMUCL on Linux.
But extra-language libraries is gonna be a bit of a pain. That's
always the case with Unix-Windows portability.
--
/|_ .-----------------------.
,' .\ / | No to Imperialist war |
,--' _,' | Wage class war! |
/ / `-----------------------'
( -. |
| ) |
(`-. '--.)
`. )----'
Tim Bradshaw wrote:
> ······@interaccess.com (Thaddeus L Olczyk) wrote in message news:<·················@nntp.interaccess.com>...
>
>
>>Producing freely distributed executables with both is ( my
>>understanding ) highly problematic.
>>
>
> For LispWorks at least this is not so for Windows and Linux: you can
> generate freely distributable executables for both these platforms
> from the commercial product (possibly with some constraints - there
> may be stuff like CORBA or something you can't include). I think
> there are royalties for the Unix platforms.
As far as I understand the new license-terms for LW4.2 (Windows and
Linux) you can include modules like CORBA, CommonSQL, or KnowledgeWorks
without paying royalities.
ciao,
Jochen
······@interaccess.com (Thaddeus L Olczyk) writes:
> On 16 Jan 2002 17:55:06 +0100, Friedrich Dominicus
> <·····@q-software-solutions.com> wrote:
>
> >> I think about using lisp for a new project (purely non-commericial).
> >> There is a need for a GUI (Buttons, Listboxes, scrollbars, grahical
> >> output, ...) and it should run under Win32 (NT / XP / 2000) and
> >> Linux.
> >You could try LispWorks from Xanalys. IIRC is there cross-platform
> >GUI-stuff (CAPI) part of the free Version. It works on Windows and
> >Linux and according to their homepage other Unices too.
> You should ask why they want a free lisp.
> If they want something for delivering open source and free software
> then both Franz and LispWorks are not really useful.
Well you can write you own open source with it can't you?
> You generally distribute your code as both source and executables.
> ( Do you really want the user to download Xanalys and compile your
> address book? )
Do you really want the user to download Perl to run your address book?
> Producing freely distributed executables with both is ( my
> understanding ) highly problematic.
Higly problematic, I doubt it very much. At least LispWorks is used
for there whole "intelligence" Software and that are all stand-alone
applications. So it should not be very hard to get it done.
Regards
Friedrich
Friedrich Dominicus <·····@q-software-solutions.com> writes:
> > I think about using lisp for a new project (purely non-commericial).
> > There is a need for a GUI (Buttons, Listboxes, scrollbars, grahical
> > output, ...) and it should run under Win32 (NT / XP / 2000) and
> > Linux.
> You could try LispWorks from Xanalys. IIRC is there cross-platform
> GUI-stuff (CAPI) part of the free Version. It works on Windows and
> Linux and according to their homepage other Unices too.
I think LispWorks is far from being free!
--
Frederic Brunel
Software Engineer
In-Fusio, The Mobile Fun Connection
The personal edition of LispWorks is indeed free (download from
www.xanalys.com) but is limited in its capabilities (heap size, no
stand-alone delivery, etc., but the CAPI GUIS is included).
Jim Bushnell
"Frederic Brunel" <··········@in-fusio.com> wrote in message
···················@buzz.in-fusio.com...
> Friedrich Dominicus <·····@q-software-solutions.com> writes:
>
> > > I think about using lisp for a new project (purely non-commericial).
> > > There is a need for a GUI (Buttons, Listboxes, scrollbars, grahical
> > > output, ...) and it should run under Win32 (NT / XP / 2000) and
> > > Linux.
> > You could try LispWorks from Xanalys. IIRC is there cross-platform
> > GUI-stuff (CAPI) part of the free Version. It works on Windows and
> > Linux and according to their homepage other Unices too.
>
> I think LispWorks is far from being free!
>
> --
> Frederic Brunel
> Software Engineer
> In-Fusio, The Mobile Fun Connection
········@acm.org (Johann Murauer) writes:
> I think about using lisp for a new project (purely non-commericial).
> There is a need for a GUI (Buttons, Listboxes, scrollbars, grahical
> output, ...) and it should run under Win32 (NT / XP / 2000) and Linux.
>
> My question: is there a free (or very cheap) lisp implementation which
> either has a build-in GUI or has an easy interface to Qt, GTK or
> something similar that runs under Linux and Win32.
gcl has a Tk & I think a GTk interface.
--
Harvey Stein
Bloomberg LP
·······@bloomberg.com
·······@bloomberg.com (Harvey J. Stein) writes:
> ········@acm.org (Johann Murauer) writes:
>
> gcl has a Tk & I think a GTk interface.
But again, with the lisp2wish idea I mentioned above you don't really
NEED an implementation-specific interface (besides which last time I
checked gcl-tk was broken with newer Tk versions). It's quite easy to
get used to:
(format *wish*
"button .b1 -text {Click me} -command {puts lisp-foo}~%")
which would then run the function lisp-foo when clicked. (It's
criminal how little Tcl you actually have to learn to make your GUI.)
Or to wrap the thing up with something general like:
(defun make-button (name &optional callback &key (frame ".f1"))
"Sends a button constructor to wish, callback MUST be specified if
name is more than one word."
(when (null callback)
(setq callback name))
(format *wish*
"button ~a.b~A -text {~A} -command {set timedShow 0;puts ~A}~%"
frame callback name callback)
(format *wish* "pack ~a.b~a -side left~%" frame callback))
And I'm quite the newbie, again, so that may be a crude way to do it,
but if it works even for newbies, then that's a selling point. The
bigger selling point is you can use CLISP, CMUCL, Allegro, and
probably every other implementation as well since you're just sending
text to the windowing shell via FORMAT. And Tk exists on multiple
platforms as well and I think looks quite nice.
I suppose for a major project going this way would perhaps involve
re-inventing the wheel on things that CLIM already has implemented,
but in my small experience, if all you have in mind is putting up a
simple, nice GUI for your program it's the easiest way to do it.
-JC
<http://ww.telent.net/cliki/Graphics%20Toolkit> -- Lisp2wish.lisp
<https://sourceforge.net/projects/clslideshow/> -- a demonstration
<http://www.scriptics.com./man/tcl8.4/> -Tcl/Tk attempt at a hyperspec
Jonathan Craven <········@craven.mail.mcgill.ca> writes:
> ·······@bloomberg.com (Harvey J. Stein) writes:
>
> > ········@acm.org (Johann Murauer) writes:
> >
> > gcl has a Tk & I think a GTk interface.
>
> But again, with the lisp2wish idea I mentioned above you don't really
> NEED an implementation-specific interface (besides which last time I
> checked gcl-tk was broken with newer Tk versions). It's quite easy to
> get used to:
>
> (format *wish*
> "button .b1 -text {Click me} -command {puts lisp-foo}~%")
That's a low level interface. The lisps with graphics toolkits
integrated usually have a higher level, more lispish interface. In
STk you'd have:
(define b (make <button>
:text "Click me"
:action (lambda () (do-something in lisp environment))))
etc.
--
Harvey Stein
Bloomberg LP
·······@bloomberg.com
On 17 Jan 2002 12:13:09 -0500, Jonathan Craven
<········@craven.mail.mcgill.ca> wrote:
>
>·······@bloomberg.com (Harvey J. Stein) writes:
>
>> ········@acm.org (Johann Murauer) writes:
>>
>> gcl has a Tk & I think a GTk interface.
>
>But again, with the lisp2wish idea I mentioned above you don't really
>NEED an implementation-specific interface (besides which last time I
>checked gcl-tk was broken with newer Tk versions). It's quite easy to
>get used to:
>
>(format *wish*
> "button .b1 -text {Click me} -command {puts lisp-foo}~%")
>
>which would then run the function lisp-foo when clicked. (It's
>criminal how little Tcl you actually have to learn to make your GUI.)
>Or to wrap the thing up with something general like:
>
>(defun make-button (name &optional callback &key (frame ".f1"))
> "Sends a button constructor to wish, callback MUST be specified if
> name is more than one word."
> (when (null callback)
> (setq callback name))
> (format *wish*
> "button ~a.b~A -text {~A} -command {set timedShow 0;puts ~A}~%"
> frame callback name callback)
> (format *wish* "pack ~a.b~a -side left~%" frame callback))
>
>And I'm quite the newbie, again, so that may be a crude way to do it,
>but if it works even for newbies, then that's a selling point. The
>bigger selling point is you can use CLISP, CMUCL, Allegro, and
>probably every other implementation as well since you're just sending
>text to the windowing shell via FORMAT. And Tk exists on multiple
>platforms as well and I think looks quite nice.
>
>I suppose for a major project going this way would perhaps involve
>re-inventing the wheel on things that CLIM already has implemented,
>but in my small experience, if all you have in mind is putting up a
>simple, nice GUI for your program it's the easiest way to do it.
>
>-JC
>
><http://ww.telent.net/cliki/Graphics%20Toolkit> -- Lisp2wish.lisp
><https://sourceforge.net/projects/clslideshow/> -- a demonstration
><http://www.scriptics.com./man/tcl8.4/> -Tcl/Tk attempt at a hyperspec
Hi,
(it's that basic that I think I should not pollute the list. And I am
a very bloody beginner ....)
I want to you CLISP and wish83 on Windows XP
#1: I installed CLISP2.27 --- seems to be okay
#2: I installed wish83 --- seems also akay
#3: In CLISP I loaded (load "lisp2wish.lisp") --- also okay
When I tried
(TEST-WISH)
I get
*** - EVAL: the function WISH::RUN-PROGRAM is undefined
Also
(format *wish* "button .b1 -text {click.me} -command {exit}")
does not work
I tried it in different ways: first with running wish before I
loaded lisp2wish, second I loaded lisp2wish and then started
wish --- nothing help.
BTW: what about a "pack .b1", not needed?
It would be great if you could point me to my error and give me
the most simple "Hello started tutorial".
Many thanks,
Johann Murauer
········@acm.org (Johann Murauer) writes:
> My question: is there a free (or very cheap) lisp implementation which
> either has a build-in GUI or has an easy interface to Qt, GTK or
> something similar that runs under Linux and Win32.
If Scheme counts as a LISP, then MzScheme, STk, and possibly other
implementations (bigloo? MIT-Scheme?) have portable GUIs that work
accross platforms (Linux, Win32, and perhaps also MacOS for some of them).
Yours freely,
[ Fran�ois-Ren� �VB Rideau | Reflection&Cybernethics | http://fare.tunes.org ]
[ TUNES project for a Free Reflective Computing System | http://tunes.org ]
Tradition is the matter of which civilization is made.
Anyone who rejects tradition per se should be left naked in a desert island.
Innovation is the matter with which civilization is built.
Anyone who rejects innovation per se should be left naked in a desert island.
Francois-Rene Rideau <····@tunes.org> writes:
> ········@acm.org (Johann Murauer) writes:
>> My question: is there a free (or very cheap) lisp implementation which
>> either has a build-in GUI or has an easy interface to Qt, GTK or
>> something similar that runs under Linux and Win32.
> If Scheme counts as a LISP, then MzScheme, STk, and possibly other
> implementations (bigloo? MIT-Scheme?) have portable GUIs that work
> accross platforms (Linux, Win32, and perhaps also MacOS for some of them).
>
Yes, but do they work across implementations? (which would normally
ensure portability)
> Yours freely,
>
> [ Fran�ois-Ren� �VB Rideau | Reflection&Cybernethics | http://fare.tunes.org ]
> [ TUNES project for a Free Reflective Computing System | http://tunes.org ]
> Tradition is the matter of which civilization is made.
> Anyone who rejects tradition per se should be left naked in a desert island.
> Innovation is the matter with which civilization is built.
> Anyone who rejects innovation per se should be left naked in a desert island.
--
BPT <···@tunes.org> /"\ ASCII Ribbon Campaign
backronym for Linux: \ / No HTML or RTF in mail
Linux Is Not Unix X No MS-Word in mail
Meme plague ;) ---------> / \ Respect Open Standards
Try newLISP
http://www.nuevatec.com/
http://welcome.to/newlisp
It's easy enough for ME to use!
Johann Murauer wrote:
> Hi,
>
> I think about using lisp for a new project (purely non-commericial).
> There is a need for a GUI (Buttons, Listboxes, scrollbars, grahical
> output, ...) and it should run under Win32 (NT / XP / 2000) and Linux.
>
> My question: is there a free (or very cheap) lisp implementation which
> either has a build-in GUI or has an easy interface to Qt, GTK or
> something similar that runs under Linux and Win32.
>
> (If not I have to do the programming in C++, but maybe I could run
> some subporcesses in Lisp which will do the logical operations. Is
> this possible ?)
>
> May thanks and best regards,
> Johann Murauer
> ········@acm.org
Johann Murauer schrieb:
> Hi,
> I think about using lisp for a new project
(purely non-commericial).
> There is a need for a GUI (Buttons, Listboxes, scrollbars, grahical
> output, ...) and it should
run under Win32 (NT / XP / 2000) and Linux.
> My question: is there a free (or very cheap) lisp implementation which
> either has
a build-in GUI
or has an
easy interface
to Qt, GTK or
> something similar that runs under Linux and Win32.
> (If not I have to do the programming in C++, but maybe I could
run some subporcesses in Lisp
which will do the logical operations. Is
> this possible ?)
try:
http://nuevatec.com/download/
latest version (07-01-2002) is 6.3.23 tk-76, 2.2mb.exe
more?
Multiple name spaces
OOP extensions
Automatic memory management
Small memory footprint
Shared library interface to import 'C' functions
TCP/IP functions
Matrix and advanced Math functions
Statistics library module with GNUPLOT support
Integrated web based Development Environment IDE v.1.1
Tcl/Tk GUI (similar to the old Win32 version) included
Module for MySQL relational database access
Fast loading and small memory footprint for CGI
Mandelbrot and Environment CGI examples
httpd web server script example
built in XML parser
Premade binary for Linux libc.so.6 based systems
Compiles also under freeBSD and mos UNIIXs
Compiles on Windows2000 using CYGWIN
License: GNU Public License, GPL
:)
stefan