From: Rahul Jain
Subject: Re: Lisp or Smalltalk: Suicide Mission, Part II
Date:
Message-ID: <878y6k0yxc.fsf@nyct.net>
·······@runbox.com writes:
> Someone (Avi?) once wrote that Lisp is multi-paradigm while Smalltalk
> is not, but that since the one paradigm Smalltalk uses is so powerful,
It's powerful if that's the only paradigm that your application needs.
--
Rahul Jain
·····@nyct.net
Professional Software Developer, Amateur Quantum Mechanicist
In an attempt to curtail some of the language 'skirmishes' that are
going on, I'd like to reiterate that I'm primarily interested in the
practical aspects that attend a programming project like mine at this
point (vendor support, implementations, preexisting code, thread
support, frameworks, tools, etc.), and not so much the language design
issues. I think the discussion has gotten off-track, whether due to
my own poorly worded questions, or due to people's general passion
about their favorite languages.
I realize that Lisp is a more flexible/powerful language than
Smalltalk. That was never my question. I knew that before I posted
the first question, and I don't know many Smalltalkers who would
dispute that...Smalltalk was, after all, designed in part to be a
language to teach children programming, while Lisp comes from a lambda
calculus/AI-type lineage. I don't know many children that are into
lambda calculus or genetic algorithms ;-)
So, yes, Lisp is powerful and flexible, but...
If I needed to screw in a flathead screw, I would use a flathead
screwdriver that I could purchase at a hardware store. I would not
want to fashion my own tool out of a "perfect metal alloy." Nor would
I want to use a Swiss Army knife. Now, a Swiss Army knife is a more
powerful and flexible tool, because it has multiple screwdrivers and
knives, a file, scissors, and a leather punch,etc. It is a greater
feat of design and engineering. But it is NOT THE BEST TOOL FOR THE
JOB. [Have you ever tried to open a tool on a Swiss Army knife, only
to have to open three others before you find the one you actually
wanted? Have you ever had a Swiss Army knife screwdriver be too short
to reach a given task?]
I marvel at the "programmable programming language," but I was trying
to find out which of the two languages (Lisp and Smalltalk, not
Ruby/Python/PHP/Perl/Java/etc.), was better suited for a programming
team of mostly amateurs in my given problem domain. And which would
provide more support from the community, in the form of preexisting
code and solid tools/vendors/implementations/frameworks.
I now know (thanks to pointers to the JP Morgan and LingoMotors
projects), that Smalltalk is particularly well-suited, like Lisp, to my
problem domain. That is, moreso that most other languages. I also know
that VisualWorks support on OS X is not ideal, but improving. I was
privately emailed about basic threading design using VisualWorks. I've
also learned much about Lisp, why to learn Lisp, and how to begin
learning Lisp. But I've learned less about the practical aspects of
using Lisp than I have about the language design features.
I'd mentioned that I was interested in LispWorks, SBCL, and UnCommon
Web, for example. LispWorks is now probably our only Lisp
implementation option (Windows support is mandatory for one of our
customers. And, yes, I'm aware of Franz and Clisp.) I only just now
realized that Uncommon Web is not currently supported on LispWorks.
THAT is the type of information I am still seeking...but, even moreso,
I am seeking information about specific things you know that might be
an issue for me that I will not find on an public website or without
wasting many hours/months finding out...
Not a total brain dump, mind you, but things that might be pertinent to
someone who doesn't have a Unix background and is beginning a web
project that will have NLP and XML components. How to approach the
problem, what tools you would avoid,etc. What Marc Battyani wrote about
modeling objects as CLOS classes in his framework is an example of
pointing out how they approach a problem. That helps. Pascal Costanza
alerted me to the fact that the LW editor can map to Mac or Emacs key
bindings and is extensible via CL. That helps.
Thanks to everyone who is taking time to help me. I greatly appreciate
it.
- Sergei
P.S.
I'm still cross-posting only because I still want feedback from both
lists, I apologize if any one message is too focused on only one
language or too verbose for those who are uninterested. I'm getting
lots of nice (and some crazy) private messages and posts. Everyone is
helping.
From: Wade Humeniuk
Subject: Re: Lisp or Smalltalk: Suicide Mission, Part II
Date:
Message-ID: <i4%Id.867$Ob.719@edtnps84>
·······@runbox.com wrote:
>
> If I needed to screw in a flathead screw, I would use a flathead
> screwdriver that I could purchase at a hardware store. I would not
> want to fashion my own tool out of a "perfect metal alloy." Nor would
> I want to use a Swiss Army knife. Now, a Swiss Army knife is a more
> powerful and flexible tool, because it has multiple screwdrivers and
> knives, a file, scissors, and a leather punch,etc. It is a greater
> feat of design and engineering. But it is NOT THE BEST TOOL FOR THE
> JOB. [Have you ever tried to open a tool on a Swiss Army knife, only
> to have to open three others before you find the one you actually
> wanted? Have you ever had a Swiss Army knife screwdriver be too short
> to reach a given task?]
>
Sergei, in all honesty, you DO NOT KNOW what the job is. Also equating
Common Lisp to a toy like a Swiss Army knife is pretty insulting and
ignorant.
Wade
Wade,
Please do not get so offended simply because I'm bad with analogies.
English is my third language, so I sometimes err. Perhaps a better
analogy would be that I do not wish to pay extra for a specialist
physician who has six board certifications when I only need my tonsils
removed. Paying extra as in needless complexity and poor tool support.
Lisp can seem overwhelming in its splendor, so I am concerned about
the learning curve and tool support.
It seems like Lisp is a wonderful building material and blueprint, but
that everyone in the Lisp community wants to cut down the trees and
make their own mortar to build a house, instead of using drywall and
pre-cut lumber or bricks...this makes sense sometimes, but not all the
time. I am simply trying to figure out why this is. I am not making
criticisms as an outsider, but as someone who wants to learn. If it
makes sense to make my own bricks, then I want to learn the best method
;-)
You and many others in the Lisp community have been very helpful, and I
have gotten many private emails from both sides of the fence that have
been very helpful...but I have also gotten seven private emails -
exclusively from Lispers - that are abusive and tell me that I am
doomed to fail and that they will ENJOY seeing me fail, except that
they wish I would fail with another language. (?!) It seems as if the
authors take great joy in being arrogant and condescending. Yet it is
obvious that most have not read much of what I've written about my
willingness to learn and the fact that there WILL be professional
programmers brought in on the project to help us. Most of the
Smalltalkers that have emailed me, on the other hand, have said that my
goal is attainable and have given me pointers on how to achieve it.
Why are Lisp users so DEFEATIST when their language is essentially a
superset of Smalltalk? I had thought smug lisp wienies were a myth,
but they are apparently not? Cultural issues are important to me, so I
ask these questions seriously, not to inflame....
- Sergei
·······@runbox.com wrote:
> You and many others in the Lisp community have been very helpful,
> and I have gotten many private emails from both sides of the fence
> that have been very helpful...but I have also gotten seven private
> emails - exclusively from Lispers - that are abusive and tell me
> that I am doomed to fail and that they will ENJOY seeing me fail,
> except that they wish I would fail with another language. (?!) It
> seems as if the authors take great joy in being arrogant and
> condescending. Yet it is obvious that most have not read much of
> what I've written about my willingness to learn and the fact that
> there WILL be professionalprogrammers brought in on the project to
> help us. Most of theSmalltalkers that have emailed me, on the other
> hand, have said that mygoal is attainable and have given me pointers
> on how to achieve it.Why are Lisp users so DEFEATIST when their
> language is essentially asuperset of Smalltalk? I had thought smug
> lisp wienies were a myth,but they are apparently not? Cultural
> issues are important to me, so Iask these questions seriously, not
> to inflame....
Holy crap! I was actually impressed when you weren't discouraged by
that guy (non-Lisp user?) who claimed programming is a profession.
Because I dislike the idea that "programmers" are supposed to wear
fancy hats and have some pseudo-standard of professionality; we are
just people who use tools. And I thought Lisp rewarded people who come
to the table with few preconceptions.
But apparently these mean-spirited guys from the 1999-2002 period are
still around. So sure, take that as a negative spot against this tool,
because you'll want to enjoy using it.
I've thought about this phenomenon a little, and I think during that
period, Lisp was wiped away from both CS and industry. So people who've
invested a lot in that skill were suddenly in a world of hurt. And
also, the more friendly types sometimes left computing altogether and
became bartenders. (Literally, in one case.) So while wonderful people
remained, there also grew a cesspool of poisonous ones.
What is a mistake but an external thing? It is an interesting
correlation that children are allowed to make the most "mistakes" and
also tend to learn the fastest.
MfG,
Tayssir
Tayssir John Gabbour wrote:
> ·······@runbox.com wrote:
>
>>You and many others in the Lisp community have been very helpful,
>>and I have gotten many private emails from both sides of the fence
>>that have been very helpful...but I have also gotten seven private
>>emails - exclusively from Lispers - that are abusive and tell me
>>that I am doomed to fail and that they will ENJOY seeing me fail,....
>
>
> Holy crap!
So you believe him?
kt
--
Cells? Cello? Cells-Gtkk?: http://www.common-lisp.net/project/cells/
Why Lisp? http://lisp.tech.coop/RtL%20Highlight%20Film
Land o' Kenny? http://www.tilton-technology.com/index.html
Obligatory quote to make me seem cool:
"Doctor, I wrestled with reality for forty years, and I am happy to
state that I finally won out over it." -- Elwood P. Dowd
From: Wade Humeniuk
Subject: Re: Lisp or Smalltalk: Suicide Mission, Part II
Date:
Message-ID: <BuaJd.21171$Qb.10038@edtnps89>
Kenny Tilton wrote:
>
>
> Tayssir John Gabbour wrote:
>
>> ·······@runbox.com wrote:
>>
>>> You and many others in the Lisp community have been very helpful,
>>> and I have gotten many private emails from both sides of the fence
>>> that have been very helpful...but I have also gotten seven private
>>> emails - exclusively from Lispers - that are abusive and tell me
>>> that I am doomed to fail and that they will ENJOY seeing me fail,....
>>
>>
>>
>> Holy crap!
>
>
> So you believe him?
>
All he has to do is post these abusive emails he is talking about.
I would like to see them. I just ignore claims with no
evidence from an anonymous source.
Wade
Kenny Tilton wrote:
> Tayssir John Gabbour wrote:
> > ·······@runbox.com wrote:
> >>You and many others in the Lisp community have been very helpful,
> >>and I have gotten many private emails from both sides of the fence
> >>that have been very helpful...but I have also gotten seven private
> >>emails - exclusively from Lispers - that are abusive and tell me
> >>that I am doomed to fail and that they will ENJOY seeing me
fail,....
> >
> > Holy crap!
>
> So you believe him?
Who cares. It's still entirely plausible that 1-7 smuglispweenies
flamed him privately. So even if he were the Master of Trolls, it's
because he's saying something believable to anyone with half a brain
and comp.lang.lisp experience.
What I don't believe is how some would prefer to waste an interesting
opportunity to know more about the Smalltalk culture, since their goals
tend to be harmonious with Lisp's.
And how would one would go about 'insulting' a tool? How about: "Lisp
is so old, those aren't parentheses, those are wrinkles."
"Lisp programmers write so much trash, no wonder why they developed
garbage collection."
"Lisp is so ugly, most computer geeks wouldn't touch her for free."
I mean, The Daily Show doesn't need the Bush Administration to find
ridiculous new material; they only need to read Usenet. "The first rule
of the ALU is -- you do not talk about the ALU. The second rule of the
ALU is -- you DO NOT talk about the ALU."
But I'm sure Lisp might fail for another 50 years because of some
troll, right? ;) Always some external boogieman bringin' us down, like
Microsoft or Alan Kay, despite the zillions pumped into Lisp/AI during
the Cold War. It's the ESTABLISHMENT, man! The Man is keepin' us down
and all the brainwashed drones don't get it!
But I make fun because I love... Lisp is fun for being insane. The evil
mastermind Alan Kay claimed that a better perspective is worth 80 IQ
points; and a few smug lisp weenies provide us all a lot of
entertainment because they don't realize this IQ boost doesn't apply
all around.
...
MfG,
Tayssir
From: Wade Humeniuk
Subject: Re: Lisp or Smalltalk: Suicide Mission, Part II
Date:
Message-ID: <VV4Jd.48924$06.46135@clgrps12>
·······@runbox.com wrote:
> It seems like Lisp is a wonderful building material and blueprint, but
> that everyone in the Lisp community wants to cut down the trees and
> make their own mortar to build a house, instead of using drywall and
> pre-cut lumber or bricks...this makes sense sometimes, but not all the
> time. I am simply trying to figure out why this is. I am not making
> criticisms as an outsider, but as someone who wants to learn. If it
> makes sense to make my own bricks, then I want to learn the best method
> ;-)
>
Programming languages and especially Lisp are neither materials nor
blueprints. They belong to a rare class of tools which I suppose
you can call "mental tools". Other things that belong in there are
things like the scientific method, written language and logic.
However programming languages are slightly different in that they
allow external execution of mental constructs. To program
one has to constrain one's thoughts to those which are executable,
you have to become rigorous. If you think your thoughts are
already expressed by some current set of pre-created libraries
(with some orgranizational glue), then you can use whatever you wish,
the problem is for all-intensive purposes solved.
If one applies a mental tool like the scientific method to curing
cancer and the process fails then one does not blame the scientific
method or even the researcher. But already you are showing signs
of blaming Lisp as a mental tool (if you get lost and confused
and you do not get the results you desire). It is not Lisp that will
confuse you, it is your own mental constructs.
From my standpoint you have become too focused on the programming
language and programming tools. One needs to take personal
responsibility instead.
Wade
From: Rahul Jain
Subject: Re: Lisp or Smalltalk: Suicide Mission, Part II
Date:
Message-ID: <87hdl618jh.fsf@nyct.net>
········@runbox.com" <·······@runbox.com> writes:
> Please do not get so offended simply because I'm bad with analogies.
> English is my third language, so I sometimes err. Perhaps a better
> analogy would be that I do not wish to pay extra for a specialist
> physician who has six board certifications when I only need my tonsils
> removed. Paying extra as in needless complexity and poor tool support.
> Lisp can seem overwhelming in its splendor, so I am concerned about
> the learning curve and tool support.
So why not learn just the things that you need? No one is forcing you to
learn logical pathnames, and you don't need them for your task, so don't
feel compelled to learn them just because they're in the standard. I
thought one of the points of your question was to find out what parts of
the language you needed to know to do what you needed to do.
--
Rahul Jain
·····@nyct.net
Professional Software Developer, Amateur Quantum Mechanicist
Rahul Jain wrote:
"I thought one of the points of your question was to find out what
parts of the language you needed to know to do what you needed to do."
This is very true. And before I can understand that I must find out
what parts of the language exist to start with, and maybe find out how
they interoperate, and maybe find out the benefits of each approach for
a given task, or even find out how each implementation might interpret
the standard slightly differently, etc. etc.
There is a great deal of learning overhead in figuring out how the map
is written, before I can even begin to find my way to my destination.
Lisp is a strange map to me, like a Russian learning Chinese.
Smalltalk's simpler abstraction and terminology is not quite as strange
to me. It is like a Spaniard learning Portuguese. Part of my original
question was an attempt to find out if this initial Lisp learning curve
would bring greater benefits to my short-term (2-3 year) future in
regards to my current projects, or if I should concentrate on Smalltalk
only at this point in time. I tend to focus very intently on things I
am trying to learn, and there are only so many hours in a day. I knew
all along I wanted to learn both languages eventually - but I thought I
could get more of a "tiebreaker" for which language to start with on
this project by getting a better feel for the landscape. And that IS
happening. Both languages are definitely capable of handling anything
I need to do, I think, so (as someone else noted) it shouldn't make a
huge difference.
There is also another aspect to the language decision which I have
avoided discussing so far because of secrecy issues and because I
thought it would cloud the discussion. I have more /specific/
questions (and responses) that I will post later...I have a big day
today meeting with our potential venture investors, so I am rushed for
time now.
Thank you,
- Sergei
From: Rahul Jain
Subject: Re: Lisp or Smalltalk: Suicide Mission, Part II
Date:
Message-ID: <878y6fzg4p.fsf@nyct.net>
········@runbox.com" <·······@runbox.com> writes:
> Rahul Jain wrote:
> "I thought one of the points of your question was to find out what
> parts of the language you needed to know to do what you needed to do."
>
> This is very true. And before I can understand that I must find out
> what parts of the language exist to start with, and maybe find out how
> they interoperate, and maybe find out the benefits of each approach for
> a given task, or even find out how each implementation might interpret
> the standard slightly differently, etc. etc.
In CL, most of that happens in the corner cases of the pathname
manipulation functions. You probably won't find them to be all that
useful for you anyway, particularly not the corner cases. :)
> There is a great deal of learning overhead in figuring out how the map
> is written, before I can even begin to find my way to my destination.
> Lisp is a strange map to me, like a Russian learning Chinese.
> Smalltalk's simpler abstraction and terminology is not quite as strange
> to me. It is like a Spaniard learning Portuguese.
I don't know. You can use the parts of lisp that correspond to the
things you already know and file away the stuff others say is useful in
your project for later investigation. E.g., learn the basic defining and
iterative constructs, list manipulation functions, basic macro usage,
symbols and packages, CLOS, basic arrays (maybe not fill-pointers,
adjustability, or displaced arrays yet), hash tables, and some basic
character stream functionality, maybe. Mostly, going through Peter's
book and skipping stuff that seems to be a bit irrelevant to your uses
will teach you everything you need to get started.
When you're looking to make the code more robust, learn the condition
system. When you're looking to make the build and deployment process
more manageable, learn asdf. As you see more opportunities for
abstraction, learn more advanced macrology, maybe from _On_Lisp_.
Another recommendation: for your development box, it's a lot easier if
you use Debian. It's got dozens of packages for lisp compilers,
libraries, and even a few complete applications.
--
Rahul Jain
·····@nyct.net
Professional Software Developer, Amateur Quantum Mechanicist
Wade Humeniuk <··················@telus.net> writes:
> Sergei, in all honesty, you DO NOT KNOW what the job is. Also equating
> Common Lisp to a toy like a Swiss Army knife is pretty insulting and
> ignorant.
This reminded me of a more than 5 year old posting by Stig Hemmer
(http://groups-beta.google.com/group/comp.lang.lisp/msg/467f30b898a87f58),
where he points out that Lisp is not merely a tool, but a tool-making
tool. A nice anology, worth a re-read!
--
(espen)
Thanks, Espen. That is a nice analogy...descriptive just like
"programmable programming language." My analogy was apparently not
only poor, but offensive. I hope I have clarified my position a bit
better...or at least not made it worse ;-)
By the way, I have a bookshelf of Java, Python, and REBOL books that
I'm interested in trading for more Lisp and Smalltalk books. I would
especially like another copy of PAIP and Touretzky, and a first copy of
SICP and Kent Beck's Smalltalk Best Practice Patterns. If anyone has
any extras and wants to learn Java/Python/REBOL (for amusement or
employment), please email me directly.
Thank you,
-Sergei
········@runbox.com" <·······@runbox.com> writes:
> Thanks, Espen. That is a nice analogy...descriptive just like
> "programmable programming language." My analogy was apparently not
> only poor, but offensive. I hope I have clarified my position a bit
> better...or at least not made it worse ;-)
>
> By the way, I have a bookshelf of Java, Python, and REBOL books that
> I'm interested in trading for more Lisp and Smalltalk books. I would
> especially like another copy of PAIP and Touretzky, and a first copy of
> SICP and Kent Beck's Smalltalk Best Practice Patterns. If anyone has
> any extras and wants to learn Java/Python/REBOL (for amusement or
> employment), please email me directly.
I would not hold my breath on the possibilities of trading Java books
for Lisp books. You'd better try to pass down these books to other
unsuspecting programmers, and buy Lisp books (or try to exchange them
against a Lisp Machine ;-).
--
__Pascal Bourguignon__ http://www.informatimago.com/
The rule for today:
Touch my tail, I shred your hand.
New rule tomorrow.
Pascal:
"I would not hold my breath on the possibilities of trading Java books
for Lisp books. You'd better try to pass down these books to other
unsuspecting programmers, and buy Lisp books (or try to exchange them
against a Lisp Machine ;-)"
True, but I thought it worth a try, since many must slave in the Java
mines to pay the mortgage :-(
"Espen Vestre" <·····@*do-not-spam-me*.vestre.net> wrote in message
···················@merced.netfonds.no...
> Wade Humeniuk <··················@telus.net> writes:
>
> > Sergei, in all honesty, you DO NOT KNOW what the job is. Also equating
> > Common Lisp to a toy like a Swiss Army knife is pretty insulting and
> > ignorant.
>
> This reminded me of a more than 5 year old posting by Stig Hemmer
> (http://groups-beta.google.com/group/comp.lang.lisp/msg/467f30b898a87f58),
> where he points out that Lisp is not merely a tool, but a tool-making
> tool. A nice anology, worth a re-read!
However, if you are not a tool maker, not interested in tool making, nor
inclined to master the skills necessary to become a tool maker, or your
timeline doesn't have space for you to learn and master those skills, then a
hardware store with perhaps less than perfect but adaptable tools is a
better option.
Regards,
Will Hartung
(·····@msoft.com)
From: Rahul Jain
Subject: Re: Lisp or Smalltalk: Suicide Mission, Part II
Date:
Message-ID: <87d5vu18b5.fsf@nyct.net>
"Will Hartung" <·····@msoft.com> writes:
> However, if you are not a tool maker, not interested in tool making, nor
> inclined to master the skills necessary to become a tool maker, or your
> timeline doesn't have space for you to learn and master those skills, then a
> hardware store with perhaps less than perfect but adaptable tools is a
> better option.
Unfortunately, he's looking to create one of the most ambitious bits of
software created to date. I'm not sure that simply using off-the-shelf
components will get him where he wants to be. You can use all the
standard lumber and nails you want, but you're not going to be able to
launch a rocket to the moon made of wood, as easy as it is to hammer the
pieces together.
--
Rahul Jain
·····@nyct.net
Professional Software Developer, Amateur Quantum Mechanicist
·······@runbox.com wrote:
> I only just now
> realized that Uncommon Web is not currently supported on LispWorks.
> THAT is the type of information I am still seeking...but, even moreso,
> I am seeking information about specific things you know that might be
> an issue for me that I will not find on an public website or without
> wasting many hours/months finding out...
[...]
> I'm still cross-posting only because I still want feedback from both
> lists
For the more practical stuff, cross-posting is probably not appropriate. I
don't want to wast the Lispers time with detailed discussions of what bits and
bobs work with which Smalltalk vendors' products on which OSes, and I'd hope
that they'd not want to waste my time with similar concerns about Lisp
implementations.
-- chris