Hi
"Reinhard Oldenburg" <··········@gmx.de> skrev i en meddelelse
·····················@gmx.de...
> Has anybody used the MathLink interface of Mathematica
> from common lisp? Any hints of examples, software or
> discussion papers?
>
> Reinhard Oldenburg
>
Don't you just call the functions in whatever context you need providing the
var's for whatever calculation ? ----- Eh ,if you want the yield of any math.
function look it up in the Common Lisp Handbook the brick you know, there are
plenty descriptions of any agreaed function.
P.C.
http://groups.yahoo.com/group/pavilions/
There is a bit more to the MathLink interface than just calling a math
function. Calling has to be possible in both directions and data
types mut be converted. My hope was that somebody had alread written
code for all that.
Reinhard Oldenburg
P.C. wrote:
> Hi
>
> "Reinhard Oldenburg" <··········@gmx.de> skrev i en meddelelse
> ·····················@gmx.de...
>
>>Has anybody used the MathLink interface of Mathematica
>>from common lisp? Any hints of examples, software or
>>discussion papers?
>>
>>Reinhard Oldenburg
>>
>>
>
> Don't you just call the functions in whatever context you need providing the
> var's for whatever calculation ? ----- Eh ,if you want the yield of any math.
> function look it up in the Common Lisp Handbook the brick you know, there are
> plenty descriptions of any agreaed function.
> P.C.
> http://groups.yahoo.com/group/pavilions/
>
>
>
Reinhard Oldenburg <··········@gmx.de> writes:
> Has anybody used the MathLink interface of Mathematica
> from common lisp? Any hints of examples, software or
> discussion papers?
>
> Reinhard Oldenburg
Just a remark: Maxima.sourceforge.net is entirely written in Lisp and its very
straightforward to call it from Lisp. Maybe you should look at it. If you
really want to use MMA, you might want to consider Richard Fatemans MockMMA,
but I don't know.
Martin
Martin Rubey wrote:
> Reinhard Oldenburg <··········@gmx.de> writes:
>
>
>>Has anybody used the MathLink interface of Mathematica
>>from common lisp? Any hints of examples, software or
>>discussion papers?
>>
>>Reinhard Oldenburg
>>
>
> Just a remark: Maxima.sourceforge.net is entirely written in Lisp and its very
> straightforward to call it from Lisp. Maybe you should look at it. If you
> really want to use MMA, you might want to consider Richard Fatemans MockMMA,
> but I don't know.
>
> Martin
>
Yes, I know of Maxima and it is an interesting option. However,
I need
- fast Groebner basis computations
- fast numerical non-linear-many-variable optimization
I suppose that mathematica beats Maxima in both of this tasks.
Yours, Reinhard Oldenburg
Reinhard Oldenburg <··········@gmx.de> writes:
> Yes, I know of Maxima and it is an interesting option. However,
> I need
> - fast Groebner basis computations
http://alamos.math.arizona.edu/%7Erychlik/MaximaGrobner.zip
> - fast numerical non-linear-many-variable optimization
Don't know about this one, but I'll ask on the maxima list.
> I suppose that mathematica beats Maxima in both of this tasks.
I wouldn't be sure about this. Maxima can be fine-tuned very well, since you
write the stuff you want to be fast in Lisp, not some CAS language. Eg, the
Gr�bner package above is in Lisp.
Martin
From: Christopher C. Stacy
Subject: Re: Mathematica interface
Date:
Message-ID: <usmrsnhmv.fsf@dtpq.com>
You should also consider Macsyma (the commercial offering
from Symbolics). It has a lot more stuff in it than Maxima,
and might have what you're looking for.
Two things:
* Macsyma is dead, it is Maxima now.
* I asked on the maxima list about
- fast Groebner basis computations
and
- fast numerical non-linear-many-variable optimization
Concerning Gr�bner, I was told that there are better things than Mathematica
out there. I wonder what's really important to you: speed or other things as
well?
Concerning numerics, I'm not to sure what you are looking for. There was one
suggestion by Raymond Toy, who said
> As one of the tests in f2cl, there is a copy of minpack, a numerical
> optimization pacakge for non-linear and/or least-squares problems.
> This could be added, I suppose
I'd be glad to help.
Martin
From: Christopher C. Stacy
Subject: Re: Mathematica interface
Date:
Message-ID: <u1xz8yt9j.fsf@dtpq.com>
>>>>> On 09 May 2003 10:45:46 +0200, Martin Rubey ("Martin") writes:
Martin> * Macsyma is dead, it is Maxima now.
What does that mean?
······@dtpq.com (Christopher C. Stacy) writes:
> >>>>> On 09 May 2003 10:45:46 +0200, Martin Rubey ("Martin") writes:
>
> Martin> * Macsyma is dead, it is Maxima now.
>
> What does that mean?
The company that sold Macsyma does not exist anymore (I don't know since when)
On http://maxima.sourceforge.net/ you can read:
Maxima is a descendant of DOE Macsyma, which had its origins in the late 1960s
at MIT. It is the only system based on that effort still publicly available and
with an active user community, thanks to its open source nature. Macsyma was
the first of a new breed of computer algebra systems, leading the way for
programs such as Maple and Mathematica. This particular variant of Macsyma was
maintained by William Schelter from 1982 until he passed away in 2001. In 1998
he obtained permission to release the source code under GPL. It was his efforts
and skill which have made the survival of Maxima possible, and we are very
grateful to him for volunteering his time and skill to keep the original
Macsyma code alive and well. Since his passing a group of users and developers
has formed to keep Maxima alive and kicking. Maxima itself is reasonably
feature complete at this stage, with abilities such as symbolic integration, 3D
plotting, and an ODE solver, but there is a lot of work yet to be done in terms
of bug fixing, cleanup, and documentation. This is not to say there will be no
new features, but there is much work to be done before that stage will be
reached, and for now new features are not likely to be our focus.
Martin
(Unfortunately we do not have the latest Macsyma sources, which would be good
because a lot of bugs don't appear in the commercial version...)
Martin Rubey <········@yahoo.de> wrote in message news:<···············@invite02.labri.fr>...
> ······@dtpq.com (Christopher C. Stacy) writes:
>
> > >>>>> On 09 May 2003 10:45:46 +0200, Martin Rubey ("Martin") writes:
> >
> > Martin> * Macsyma is dead, it is Maxima now.
> >
> > What does that mean?
>
> The company that sold Macsyma does not exist anymore (I don't know since when)
>
> On http://maxima.sourceforge.net/ you can read:
>
> Maxima is a descendant of DOE Macsyma, which had its origins in the late 1960s
> at MIT. It is the only system based on that effort still publicly available and
> with an active user community, thanks to its open source nature. Macsyma was
> the first of a new breed of computer algebra systems, leading the way for
> programs such as Maple and Mathematica. This particular variant of Macsyma was
> maintained by William Schelter from 1982 until he passed away in 2001. In 1998
> he obtained permission to release the source code under GPL. It was his efforts
> and skill which have made the survival of Maxima possible, and we are very
> grateful to him for volunteering his time and skill to keep the original
> Macsyma code alive and well.
Does this mean the Maxima code base contains a significant portion of
Macsyma code ? What % ?
> Since his passing a group of users and developers
> has formed to keep Maxima alive and kicking. Maxima itself is reasonably
> feature complete at this stage, with abilities such as symbolic integration, 3D
> plotting, and an ODE solver, but there is a lot of work yet to be done in terms
> of bug fixing, cleanup, and documentation. This is not to say there will be no
> new features, but there is much work to be done before that stage will be
> reached, and for now new features are not likely to be our focus.
>
> Martin
>
> (Unfortunately we do not have the latest Macsyma sources, which would be good
> because a lot of bugs don't appear in the commercial version...)
Who has the latest code? Was it maintained and expanded following Mr.
Shelter's release of Macsyma code under GPL in 1998?
Olivier
> Does this mean the Maxima code base contains a significant portion of
> Macsyma code ? What % ?
Maxima is basically a version of Macsyma. As far as I know, the
commercial Macsyma and DOE Macsyma came from basically the same code
base in 1982. The commercial code was updated over the years, and a
version of DOE Macsyma became Maxima and was developed at UT Austin.
The commercial code base fixed lots of bugs and provided a substantial
number of updates, but both versions share the same roots and for the
most part have the same syntax and many of the same internals. Of
course, now that Maxima is open source many more changes have taken
place, but to the user the command line Macsyma and command line
Maxima will look very similar, since they are basically different
branches of the same original application.
> > (Unfortunately we do not have the latest Macsyma sources, which would be
> > good because a lot of bugs don't appear in the commercial version...)
>
> Who has the latest code? Was it maintained and expanded following Mr.
> Shelter's release of Macsyma code under GPL in 1998?
The commercial company didn't collapse until 1999 if memory serves -
there is a post somewhere in these archives where Richard Petti
details what happened to Macsyma Inc. See Dr. Fateman's post above
for who owns it currently. As far as the Maxima project knows, the
commercial code is now lost to the world. There has been no
indication at all that the owners of the commercial Macsyma assets are
interested in releasing them to the open source community. Whether or
not it will ever re-emerge as a commercial product with a development
team and support team remains to be seen, but regaining market share
from Mathematica and Maple is not a task to be taken lightly.
Regardless, the open source nature of Maxima has allowed a volunteer
group to begin to update it, fix the many bugs that exist in the older
code base it is based off of, and document the system. The open
source Maxima code base is under VERY active development, so in a
sense it could be considered the latest. The commercial program
probably still has a lot fewer bugs at this stage, but I have not
heard of any active development on it since Macsyma Inc. sold it.
CY
From: Christopher C. Stacy
Subject: Re: Mathematica interface
Date:
Message-ID: <u3cjitbyf.fsf@dtpq.com>
>>>>> On 13 May 2003 19:10:11 -0700, CY ("CY") writes:
CY> The commercial code base fixed lots of bugs
Yes, _many_ _thousands_ of bugs.
CY> and provided a substantial number of updates,
"Updates" makes it sound like a few new odds and ends.
A better way to characterize it would be that that commercial
version was equivalent to three or five times the power and
features of the original program; it had major improvements
in functionality, and whole new areas of functionality.
The commercial version represents about 100-150 man years
of additional development done by some of the top people
in the field.
> >>>>> On 13 May 2003 19:10:11 -0700, CY ("CY") writes:
>
> CY> The commercial code base fixed lots of bugs
>
> Yes, _many_ _thousands_ of bugs.
It is unfortunate that the buglist for Macsyma was never public - it
would have been a great resource for the current Maxima project.
> CY> and provided a substantial number of updates,
>
> "Updates" makes it sound like a few new odds and ends.
> A better way to characterize it would be that that commercial
> version was equivalent to three or five times the power and
> features of the original program; it had major improvements
> in functionality, and whole new areas of functionality.
I've never really used the commercial one - Macsyma Inc. died when I
was a beginning undergrad, and in any case I don't have the $$$ to buy
a copy. I emailed David Schmidt of Symbolics and he said that while
they do offer the 2.4 Windows version of Macsyma for $500, that does
not come with printed documentation or support. They do not offer the
Unix or Linux versions. Adding the notice that Macsyma is available
to the website is on his todo list.
I am glad to hear this - if it isn't ever going to be open source, I
want the commercial Macsyma to survive as a commercial product. For
users with the resources and need for the additional power of the
commercial version, it will be a more powerful solution than Maxima
for a very long time.
> The commercial version represents about 100-150 man years
> of additional development done by some of the top people
> in the field.
Well, I don't know where the Maxima project will wind up in the long
run, or if we will ever equal the full power of the commercial
version. I agree right now it's a long ways behind, but it has one
key advantage that the commercial version will most likely never have:
it is free and open source. It can be developed as long as there is
interest in doing so, and does not need to worry about revenue
sources. Unless active development is resumed on Macsyma, Maxima also
is the only cross platform version of Macsyma around.
I would be curious - what would you say are the major improvements
made by Macsyma as opposed to Maxima? (The obvious one, although I
suspect not the most interesting one, is the interface/plotting -
that's already on our long range radar.) If you have had experience
with both, that insight might allow for the creation of a good todo
list for Maxima. We know there is a great deal of work to be done,
but a list of improvements from someone who knows both systems would
be a help.
CY
"CY" <···········@hotmail.com> wrote in message
·································@posting.google.com...
> > >>>>> On 13 May 2003 19:10:11 -0700, CY ("CY") writes:
> If you have had experience
> with both, that insight might allow for the creation of a good todo
> list for Maxima. We know there is a great deal of work to be done,
> but a list of improvements from someone who knows both systems would
> be a help.
>
> CY
The commercial one is much more stable, does not crash as much, and has
templetes where you can enter in values /wo typing commands. (via a menu)
I think it also ran a lot faster, and did not need 2 B restarted as much.
On Sun, 18 May 2003 04:09:02 +0000, Franz Kafka wrote:
>
> "CY" <···········@hotmail.com> wrote in message
> ·································@posting.google.com...
>> > >>>>> On 13 May 2003 19:10:11 -0700, CY ("CY") writes:
>> If you have had experience
>> with both, that insight might allow for the creation of a good todo
>> list for Maxima. We know there is a great deal of work to be done,
>> but a list of improvements from someone who knows both systems would
>> be a help.
>>
>> CY
>
> The commercial one is much more stable, does not crash as much,
We're working on that :-). Please report any crashes you encounter to the
Maxima bug list - we definitely want to know about them.
> and has templetes where you can enter in values
> /wo typing commands. (via a menu)
Hmm. That's an interesting idea. New GUI work is a ways down the road,
but that idea is worth keeping in mind.
> I think it also ran a lot faster, and did not need 2 B restarted as much.
Interesting. What commands have you found to be slow on Maxima? By
restarting, do you mean it had crashed, or had something else happened?
CY
"CY" <···········@hotmail.com> wrote in message
···································@hotmail.com...
> On Sun, 18 May 2003 04:09:02 +0000, Franz Kafka wrote:
>
> >
> > "CY" <···········@hotmail.com> wrote in message
> > ·································@posting.google.com...
> >> > >>>>> On 13 May 2003 19:10:11 -0700, CY ("CY") writes:
> >> If you have had experience
> >> with both, that insight might allow for the creation of a good todo
> >> list for Maxima. We know there is a great deal of work to be done,
> >> but a list of improvements from someone who knows both systems would
> >> be a help.
> >>
> >> CY
> >
> > The commercial one is much more stable, does not crash as much,
>
> We're working on that :-). Please report any crashes you encounter to the
> Maxima bug list - we definitely want to know about them.
>
When I go into Lisp mode using to_lisp();
and attempt to execute (macsyma-top-level) to get
back out. I have to restart Minima.
>
> Interesting. What commands have you found to be slow on Maxima? By
> restarting, do you mean it had crashed, or had something else happened?
When I am ploting a function it takes much longer than it did when I was
using a 386/Win machine with Macsyma.
Some times I am not sure that it is working.
Sometimes when I enter a command it doen't do anything. (& if
I enter it again it might/might not work)
>
> CY
I have my old Macsyma guides so I can stress test it.
& when I type
integrate(x^x,x);
it returns 'integrate(x^x,x) as the answer. don't know if this is correct.
Are there any parts of Macsyma missing from Maxima--or do I have to test
every function in my manuals.
plus--when it prints math stuff in ANSI while Macsyma printed in
math symbolics -- your prob. working on this thou.
On Sun, 18 May 2003 05:11:21 +0000, Franz Kafka wrote:
>> We're working on that :-). Please report any crashes you encounter to the
>> Maxima bug list - we definitely want to know about them.
>>
> When I go into Lisp mode using to_lisp();
> and attempt to execute (macsyma-top-level) to get
> back out. I have to restart Minima.
Hmm. I have Maxima builds on CMUCL 18e, gcl 2.5.2, clisp 2.29, and sbcl
0.7.14.
On cmucl and sbcl, (macsyma-top-level) works. Note that you can also use
(run) to return. GCL does die, and that is a serious bug. Clisp doesn't accept
(macsyma-top-level), but will accept (run).
I'm assuming your using a gcl build?
I've submitted this to our bug database. Thanks for finding it.
>> Interesting. What commands have you found to be slow on Maxima? By
>> restarting, do you mean it had crashed, or had something else happened?
>
> When I am ploting a function it takes much longer than it did when I was
> using a 386/Win machine with Macsyma.
>
> Some times I am not sure that it is working.
Ah, yes. Plotting. Our plotting is a not so minor disaster. Some work
has been done recently on that, and much more is planned.
> Sometimes when I enter a command it doen't do anything. (& if
> I enter it again it might/might not work)
Hmm. Can you provide some example commands that exhibit this behavior?
> I have my old Macsyma guides so I can stress test it.
>
> & when I type
>
> integrate(x^x,x);
>
> it returns 'integrate(x^x,x) as the answer. don't know if this is correct.
Well, I get an ascii art rendering of the same thing. It means Maxima
doesn't know how do to that integral. Does commercial Macsyma think it
can?
> Are there any parts of Macsyma missing from Maxima--or do I have to test
> every function in my manuals.
Probably some parts are missing, but I don't know what they would be.
Also it can sometimes be trick figuring out what was considered
a working part of Macsyma in the manuals. If this is Macsyma Reference
Manual version ten I have managed to get a copy of it myself - if you
would rather not mess with it I will probably eventually go through it.
> plus--when it prints math stuff in ANSI while Macsyma printed in
> math symbolics -- your prob. working on this thou.
There are a couple partial solutions, although the better ones exist only
on Linux right now. TeXmacs (http://www.texmacs.org) has a Maxima mode
which provides very nice sessions (although the interface can be fragile)
and there are some very nice Emacs modes. If you are comfortable with TeX
you can generate Maxima output in TeX form, which can be rendered very
nicely. (That's what TeXmacs uses, IIRC.)
We eventually plan to do a whole new GUI, and at that point the plan is to
address 2D interactive input and 2D output. The project has decided the
priority is fixing math bugs though, so we have quite a ways to go before
the focus of our efforts will be GUI stuff.
CY
In Lisp I get this error when I type (macsyma-top-level)
Error: The function MACSYMA-TOP-LEVAL is undefined.
Fast links are on: do (si::use-fast-links nil) for debugging
Error signalled by EVALHOOK.
Backtrace: system:universal-error-handler > EVALHOOK
Broken at SYSTEM:TERMINAL-INTERRUPT.
when I type (run) I get this message.
MAXIMA>>(run)
Maxima restarted.
(C1)
Error: Could not connect
Fast links are on: do (si::use-fast-links nil) for debugging
Error signalled by LISP:LET.
Backtrace: system:universal-error-handler > evalhook > run > lisp:let >
catch > catch > macsyma-top-level > funcall > lisp-eval > progn > progn >
user::setup > lisp:let* > open-socket > lisp:let > SYSTEM:SOCKET
(C1)
I can still use macsyma but I get a Error sending to Maxima::
can not find channel named "sock176" you may need to Restart
Selecting File/Restart restarts Macsyma.
when I execute to_lisp();
I get a socket error and have to restart.
Hitting <ctrl><g> gets me to the Lisp Listener.
Sometimes when I enter Lisp with to_lisp();
I can not type in commands and get
responces back.
Some features that may be missing are unit conversions, tensors, some ode
solvers.
It is also hard to compare for two other reasons:
the ode function in Macsyma is equiv. to ode2 in Maxima -- found this out by
testing / observing.
plot2d(sin(x),[x,-%pi,%pi]);
is
plot(sin(x),x,-%pi,%pi); in Macsyma
mat: [a, b, c; d, e, f];
works in Macsyma and genrates this error:
at: [a, b, c;
^
(C1) Incorrect syntax: Too many ]'s
d, e, f]
^
(C1) Incorrect syntax: Premature termination of input at ;.
;
^
in Maxima.
On Sun, 18 May 2003 19:19:30 +0000, Franz Kafka wrote:
> In Lisp I get this error when I type (macsyma-top-level)
>
> Error: The function MACSYMA-TOP-LEVAL is undefined.
> Fast links are on: do (si::use-fast-links nil) for debugging
> Error signalled by EVALHOOK.
> Backtrace: system:universal-error-handler > EVALHOOK
>
> Broken at SYSTEM:TERMINAL-INTERRUPT.
That's the error I got in clisp. I've reported it.
> Sometimes when I enter Lisp with to_lisp();
> I can not type in commands and get
> responces back.
Hmm. Are you working in Xmaxima? (macsyma-top-level) doesn't seem to
work for me in Xmaxima, but it fails quietly. I've reported this as well.
> Some features that may be missing are unit conversions, tensors, some ode
> solvers.
There is some work ongoing on unit conversions and new ode stuff. There
is a tensor package but it has not yet been tested.
> It is also hard to compare for two other reasons:
> the ode function in Macsyma is equiv. to ode2 in Maxima -- found this out by
> testing / observing.
>
> plot2d(sin(x),[x,-%pi,%pi]);
> is
> plot(sin(x),x,-%pi,%pi); in Macsyma
Plotting is probably going to be erratic - it's a known weak
area in Maxima.
> mat: [a, b, c; d, e, f];
> works in Macsyma and genrates this error:
Hmm. OK, that's Maxima's parser deciding the ; character in the middle of
the command is an end of line. It looks like Macsyma introduced a new
notation for matrix entry. I've entered this as lower priority bug.
If you continue with more extensive testing, you should enter your bugs
here:
http://sourceforge.net/tracker/?group_id=4933&atid=104933
That's our central archive for all Maxima bugs. Rule of thumb: Detail is
good, more detail is better.
CY
From: Christopher C. Stacy
Subject: Re: Mathematica interface
Date:
Message-ID: <uaddky5c1.fsf@dtpq.com>
>>>>> On 17 May 2003 20:46:41 -0700, CY ("CY") writes:
>> >>>>> On 13 May 2003 19:10:11 -0700, CY ("CY") writes:
>>
CY> The commercial code base fixed lots of bugs
>>
>> Yes, _many_ _thousands_ of bugs.
CY> It is unfortunate that the buglist for Macsyma was never public - it
CY> would have been a great resource for the current Maxima project.
It was a commercial investment of millions of dollars that
generated the bug list and the fixes. Why would a company
that had spent that kind of money give that away to the public?
It's unfortunate that commercial Macsyma is not being maintained and
developed at this moment. However, it is quite possible that it may
go back into active development. It's still a product that people
are paying money for, right now, today.
Perhaps people could come up with the money to purchase Macsyma,
and then place the sources under an MIT/BSD type license.
That would certainly push their effort at improving the program
forward by many years.
······@dtpq.com (Christopher C. Stacy) writes:
> Perhaps people could come up with the money to purchase Macsyma, and then
> place the sources under an MIT/BSD type license. That would certainly
> push their effort at improving the program forward by many years.
I mentioned the idea of collecting money from the public for buying Macsyma
several times in conversation with people who know the situation better
than me (Richard Fateman, Dick Petti). Their answers were not stimulating.
They apparently hope that the situation clarifies in a different way
(whichever that is).
Nicolas.
Nicolas Neuss <·············@iwr.uni-heidelberg.de> writes:
> ······@dtpq.com (Christopher C. Stacy) writes:
>
> > Perhaps people could come up with the money to purchase Macsyma, and then
> > place the sources under an MIT/BSD type license. That would certainly
> > push their effort at improving the program forward by many years.
>
> I mentioned the idea of collecting money from the public for buying Macsyma
> several times in conversation with people who know the situation better
> than me (Richard Fateman, Dick Petti). Their answers were not stimulating.
> They apparently hope that the situation clarifies in a different way
> (whichever that is).
>
> Nicolas.
I should be more precise: RJF was only sceptical if a sufficient amount
could be raised and referred me to Petti who is apparently in contact with
Topping. IMO, the interesting point would be to know how much this code
would cost. If someone could find that out, and if the price is not
exorbitant, one might think about a collective effort.
Nicolas.
From: Christopher C. Stacy
Subject: Re: Mathematica interface
Date:
Message-ID: <u7k8uti4l.fsf@dtpq.com>
The Maxima code is based on a version of the Macsyma sources
circa 1982. Subsequent to that, Macsyma was under development
for another decade, and those improvements are available in the
commercial version of the program.
From: Christopher C. Stacy
Subject: Re: Mathematica interface
Date:
Message-ID: <uy91a3knt.fsf@dtpq.com>
>>>>> On 13 May 2003 14:48:03 +0200, Martin Rubey ("Martin") writes:
Martin> * Macsyma is dead, it is Maxima now.
Martin> The company that sold Macsyma does not exist anymore
Martin> (I don't know since when)
The company that sells Macsyma is:
Symbolics Technology, Inc.
P.O. Box 10862, Burke, VA 22015
Phone: 703-455-0430
This company has been in business for many years as a sucessor
to the bankrupt "Symbolics, Inc.". They acquired Macsyma when
"Macsyma, Inc." went bankrupt about a half-dozen years ago.
Their President, Dave Schmidt, talked about the state of Lisp Machine
technology at the recent Lisp Users and Vendors (LUV) conference.
From: Christopher C. Stacy
Subject: Re: Mathematica interface
Date:
Message-ID: <ullxa1szq.fsf@dtpq.com>
>>>>> On Tue, 13 May 2003 15:56:22 GMT, Christopher C Stacy ("Christopher") writes:
>>>>> On 13 May 2003 14:48:03 +0200, Martin Rubey ("Martin") writes:
Martin> * Macsyma is dead, it is Maxima now.
Martin> The company that sold Macsyma does not exist anymore
Martin> (I don't know since when)
Christopher> The company that sells Macsyma is:
Christopher> Symbolics Technology, Inc.
Christopher> P.O. Box 10862, Burke, VA 22015
Christopher> Phone: 703-455-0430
Christopher> This company has been in business for many years as a sucessor
Christopher> to the bankrupt "Symbolics, Inc.". They acquired Macsyma when
Christopher> "Macsyma, Inc." went bankrupt about a half-dozen years ago.
Christopher> Their President, Dave Schmidt, talked about the state of Lisp Machine
Christopher> technology at the recent Lisp Users and Vendors (LUV) conference.
Actually, the company was just recently renamed back to the original
"Symbolics, Inc.". The contact information is correct. They are
actively selling Macsyma (sold some copies in the past two weeks).
It costs $500, and comes with PDEASE and other bundled products you'd like.
Christopher C. Stacy wrote:
>>>>>>On Tue, 13 May 2003 15:56:22 GMT, Christopher C Stacy ("Christopher") writes:
>>>>>
>
>>>>>>On 13 May 2003 14:48:03 +0200, Martin Rubey ("Martin") writes:
>>>>>
> Martin> * Macsyma is dead, it is Maxima now.
>
> Martin> The company that sold Macsyma does not exist anymore
> Martin> (I don't know since when)
>
> Christopher> The company that sells Macsyma is:
>
> Christopher> Symbolics Technology, Inc.
> Christopher> P.O. Box 10862, Burke, VA 22015
> Christopher> Phone: 703-455-0430
>
> Christopher> This company has been in business for many years as a sucessor
> Christopher> to the bankrupt "Symbolics, Inc.". They acquired Macsyma when
> Christopher> "Macsyma, Inc." went bankrupt about a half-dozen years ago.
> Christopher> Their President, Dave Schmidt, talked about the state of Lisp Machine
> Christopher> technology at the recent Lisp Users and Vendors (LUV) conference.
>
> Actually, the company was just recently renamed back to the original
> "Symbolics, Inc.". The contact information is correct. They are
> actively selling Macsyma (sold some copies in the past two weeks).
> It costs $500, and comes with PDEASE and other bundled products you'd like.
I don't know what "actively selling" means,. The last I heard
Kalman Reti would send you a CD for Windows, maybe with printed
documentation.
But how you would find this out is unclear. Support may not be
available.
The Macsyma company was purchased by Andrew Topping. I think he also
purchased the assets of Symbolics. I believe
Macsyma Inc never filed for bankruptcy, and so the company probably
exists, though it may not have any employees.
see
http://www.ma.utexas.edu/pipermail/maxima/2002/001247.html
From: Christopher C. Stacy
Subject: Re: Mathematica interface
Date:
Message-ID: <ubry6ti76.fsf@dtpq.com>
>>>>> On Tue, 13 May 2003 14:52:39 -0700, Richard Fateman ("Richard") writes:
Richard> I don't know what "actively selling" means...
To me, "actively selling" means "people are calling and asking
for the product, and the company is delivering it", and that
this actually happens with some frequency (for a little tiny
company, say, a few times a month). That is a very different
reality than what Martin Rubey was claiming in his newsgroup
post, which is that the product was not available and the
company did not even exist.
······@dtpq.com (Christopher C. Stacy) wrote in message news:<·············@dtpq.com>...
> >>>>> On Tue, 13 May 2003 14:52:39 -0700, Richard Fateman ("Richard") writes:
> Richard> I don't know what "actively selling" means...
>
> To me, "actively selling" means "people are calling and asking
> for the product, and the company is delivering it", and that
> this actually happens with some frequency (for a little tiny
> company, say, a few times a month). That is a very different
> reality than what Martin Rubey was claiming in his newsgroup
> post, which is that the product was not available and the
> company did not even exist.
Have a loo at: http://www.scientek.com/macsyma/main.htm
From: Christopher C. Stacy
Subject: Re: Mathematica interface
Date:
Message-ID: <uk7cts9id.fsf@dtpq.com>
My knowledge is first-hand and is current as of May 14,2003,
so I think it's more accurate than any web site you're pointing me to.
From: Richard Fateman
Subject: Re: Mathematica interface / for sale? / how different? /power
Date:
Message-ID: <3EC3E7CE.1010908@cs.berkeley.edu>
(1) The web site seems to be from a (former?) Korean distributor of Macsyma,
which (at least without my downloading more character sets) is
hard to read. Has no prices or ordering procedure.
(2) Regarding cstacy's claim that the commercial Macsyma is
3 to 5 "the power" of the non-commercial one, there are no clear
measures of power, as far as I know.
(With great power comes great responsibility :)
The most obvious and, to my mind, one of the worst ways of
comparing complex systems is speed. I've run some timings
comparing Macsyma and Maxima. Let's dispense with that first.
Is it 3 to 5 times faster? In general, no. But there may be
places where it is faster. Places where it is slower, since it
was restricted to running on those lisp systems that were chosen
by Macsyma Inc. before about 1999, and there are some faster ones now.
Does it have 3 to 5 times more code? Hard to say from here.
Do large pieces of (admittedly neat) user interface code count
as "power"? Does cstacy count the (several) user interfaces for
Maxima on the other side of the ledger? Or the use of plotting
routines, etc imported into Maxima?
Does rewriting Matlab syntax into a Macsyma parser count as "power"?
Adding numerical routines?
Some of the system routines have clearly been substantially
rewritten, but even the circa 1982 ones are, in some cases, more
"correct" than Mathematica and Maple. Some other changes seem to
be mostly cosmetic. And there are some changes or additions that
are likely to be so esoteric as to be meaningless to most people.
Nevertheless, among all this there are lots of features and fixes that
would be nice to have, whether by re-implementation or by
somehow opening up the Macsyma Inc source.
Of course Mathematica and Maple are not standing still, and the number
of cases where they still blunder while Macsyma of 1982
or 2003 has an advantage, are fewer and fewer.
Maxima may be "behind" Macsyma in power at the moment, but it is not static.
It is not clear how many tech developers there are at Mathematica or
Maple, (or other systems such as Axiom, Mupad, Reduce[also in lisp!],
etc etc.) but Macsyma Inc's version has zero, and Maxima has quite
a few on sourceforge.
Andreas Holz wrote:
> ······@dtpq.com (Christopher C. Stacy) wrote in message news:<·············@dtpq.com>...
>
>>>>>>>On Tue, 13 May 2003 14:52:39 -0700, Richard Fateman ("Richard") writes:
>>>>>>
>> Richard> I don't know what "actively selling" means...
>>
>>To me, "actively selling" means "people are calling and asking
>>for the product, and the company is delivering it", and that
>>this actually happens with some frequency (for a little tiny
>>company, say, a few times a month). That is a very different
>>reality than what Martin Rubey was claiming in his newsgroup
>>post, which is that the product was not available and the
>>company did not even exist.
>
>
> Have a loo at: http://www.scientek.com/macsyma/main.htm
From: Christopher C. Stacy
Subject: Re: Mathematica interface / for sale? / how different? /power
Date:
Message-ID: <u3cjfdhqy.fsf@dtpq.com>
What is the adgenda of people seeking to dissuade people
from using the commercial version of Macsyma?
My adgenda is that I hate to see a fine piece of
software not used, but I have no connection with
the company that sells it.
······@dtpq.com (Christopher C. Stacy) writes:
> What is the adgenda of people seeking to dissuade people from using
> the commercial version of Macsyma?
>
> My adgenda is that I hate to see a fine piece of software not used,
> but I have no connection with the company that sells it.
["agenda"]
I agree with Chris.
It's one thing to advocate the use of a particular product.
It's quite another to try to sabotage the reputation of another.
It appears that Macsyma is not easily available at the moment, but as I
remarked a while back on this newsgroup, I believe the reason to be a
medical situation involving its owner. For privacy reasons, I haven't
been more specific. However, I consider it poor form to be having a
negative conversation about the product generally just because it is
temporarily unavailable, and especially so because there is reason to
believe the owner cannot reasonably respond and has more important things
to worry about.
If people want to use this so-called "Maxima", they should do so, I guess.
But I doubt that Macsyma is really dead, and it certainly does not need
people who are ill-informed speaking for it and trying to make it go away.
There were a lot of people actively employed on Macsyma for a long
time after the split, including (for varying intervals of time): me
(Kent Pitman), John Aspinall, Bill Dubuque, Rick Brenner, Howard
Cannon, Jeff Golden, Dick Petti, and probably others I'm forgetting.
It would be a shame to lose all of that investment.
"Kent M Pitman" <······@world.std.com> wrote in message
····················@shell01.TheWorld.com...
>
> There were a lot of people actively employed on Macsyma for a long
> time after the split, including (for varying intervals of time): me
> (Kent Pitman), John Aspinall, Bill Dubuque, Rick Brenner, Howard
> Cannon, Jeff Golden, Dick Petti, and probably others I'm forgetting.
> It would be a shame to lose all of that investment.
>
I have used Macsyma in about 1989--it had a much better
user interface that Minima--I am using Minima right now
because my version of Macsyma will not work with
Windows 9x or 2000 or XP
but if a new version of Macsyma was released I'd prob.
buy it--it is much much cheaper that Mathematica has
have a lot more featurs--including a built in Lisp compiler :)
I hope that the new version would come with a way to interface
/w other Lisps. Peter Novig in PAIP wrote about how Macsyma/Minima could be
written in Lisp.
I esp. liked the templates where you could fill in parts of the
program--this is not in Minima. (It cost me under $300 when I bought it--I
feel that that's a fair price.)
From: Martin Rubey
Subject: Re: Mathematica interface / for sale? / how different? /power
Date:
Message-ID: <whfwugr9uzy.fsf@invite02.labri.fr>
Dear Chris, dear Kent,
In fact, I didn't want to start a fight here. I would be interested what's
going on with the commercial Macsyma, too. I didn't know that it was still
possible to buy it, therefore I'm thankful for the hint Chris gave. I just
didn't know, excuse my inappropriate wording, please! It seems that it is
*very* difficult to give any simple fact the adjective "true" or "false"...
Considering how this thread started, I really wanted to suggest a lisp based
product instead of Mathematica or Maple, nothing more. Given that Maxima is
under active development, and is quite OK to use, I suggested it. You have to
be careful, because there are quite some -mathematical - bugs included, but
this goes also with MMA, which I used quite extensively also. In any case, from
the MMA team I got the response "sorry, but we can't fix this" concerning a
*very* annoying bug (In[1]= Sum[Binomial[0,k],{k,0,n}] Out[1]= 0, its still
there as of 4.1), with maxima I just fixed the bug. (well, there is one where
we were only able to fix the symptoms, so far)
Concerning Macsyma versus Maxima, I think that a coexistence could be very
fruitful. Unfortunately (in my opinion), we do not have a sign of life (is this
english? - sorry, not native) of the vendors. When it comes to it, I hope we
can cooperate in some way.
Sorry again for the possibly inappropriate word "dead". Maybe "sleeping" is
better? Is there technical support available?
How is http://www.scientek.com/macsyma/main.htm related to Symbolics
Technology, Inc? Are there several vendors of Macsyma out there? (I can't find
any info on the scientek homepage, probably since its all korean which I don't
know...)
Martin
Martin Rubey <········@yahoo.de> wrote in message
>
> Concerning Macsyma versus Maxima, I think that a coexistence could be very
> fruitful. Unfortunately (in my opinion), we do not have a sign of life (is this
> english? - sorry, not native) of the vendors. When it comes to it, I hope we
> can cooperate in some way.
I'd say we have a life sign, but the patient is comatose. Which is
still good news - Symbolics might someday bring it back to its glory
days.
> Sorry again for the possibly inappropriate word "dead". Maybe "sleeping" is
> better? Is there technical support available?
No technical support. According to Dave Schmidt, the source code and
presumably other goodies are in cold storage, and in any case they
would have to hire programmers to work on it again.
> How is http://www.scientek.com/macsyma/main.htm related to Symbolics
> Technology, Inc? Are there several vendors of Macsyma out there? (I can't find
> any info on the scientek homepage, probably since its all korean which I don't
> know...)
I dunno. The font set is very bizarre. It looks to be some
informational stuff on Macsyma, but I can't tell if they were a
reseller at one point. I don't think it's another version of Macsyma,
or related to Symbolics. Although maybe Symbolics could use some of
that info for their Macsyma stuff, when they get the webpage up.
There HAVE been many commercial branches in the past, but I believe
they all fizzled.
CY
On Thu, 15 May 2003 17:39:31 -0400, Kent M Pitman wrote:
> ······@dtpq.com (Christopher C. Stacy) writes:
>
>> What is the adgenda of people seeking to dissuade people from using
>> the commercial version of Macsyma?
>>
>> My adgenda is that I hate to see a fine piece of software not used,
>> but I have no connection with the company that sells it.
>
> ["agenda"]
>
> I agree with Chris.
>
> It's one thing to advocate the use of a particular product.
>
> It's quite another to try to sabotage the reputation of another.
That's most definitely not the intent, at least on my part. The only
reason I found out about Maxima in the first place was one of my
professors showed me a commercial Macsyma. It's a very good program. I
don't think anyone disputes that. The new information is that it can
still be purchased. For several years nothing had been heard about it, so
it was assumed it was gone. If it is now back, that is unexpected and
excellent news.
> It appears that Macsyma is not easily available at the moment, but as I
> remarked a while back on this newsgroup, I believe the reason to be a
> medical situation involving its owner. For privacy reasons, I haven't
> been more specific. However, I consider it poor form to be having a
> negative conversation about the product generally just because it is
> temporarily unavailable, and especially so because there is reason to
> believe the owner cannot reasonably respond and has more important things
> to worry about.
Again, the point was not to bash Macsyma. Most people would readily admit
Macsyma is better mathematically and interface wise than Maxima right now.
But if it isn't available then it doesn't help much.
> If people want to use this so-called "Maxima", they should do so, I guess.
It is a real program, and for many uses it is a very good program.
Because it is open source, it is also likely to get better with time.
> But I doubt that Macsyma is really dead, and it certainly does not need
> people who are ill-informed speaking for it and trying to make it go away.
We don't want it to go away. We want it back. The indications were that
it had gone away. I hope it's back. But it has been nearly invisible for
several years.
> There were a lot of people actively employed on Macsyma for a long
> time after the split, including (for varying intervals of time): me
> (Kent Pitman), John Aspinall, Bill Dubuque, Rick Brenner, Howard
> Cannon, Jeff Golden, Dick Petti, and probably others I'm forgetting.
> It would be a shame to lose all of that investment.
I most definitely agree. But virtually all the indications were that it
had been lost. I still don't quite see how it is going to compete
commercially against Mathematica and Maple, but I hope it does.
CY
······@dtpq.com (Christopher C. Stacy) wrote in message news:<·············@dtpq.com>...
> What is the adgenda of people seeking to dissuade people
> from using the commercial version of Macsyma?
>
> My adgenda is that I hate to see a fine piece of
> software not used, but I have no connection with
> the company that sells it.
Uh - I wasn't aware of any such agenda. Sorry if you got that
impression. If the commercial Macsyma does come charging back into the
game, I for one will welcome its return. The problem has been for the
past four to five years that getting a copy seemed to range anywhere
from very difficult to impossible. The recent confirmation here that
it is still available is a rather significant surprise to many people,
myself included. There has been no sign of its being out there other
than a very occasional usenet post. If that situation is now
beginning to change (and I'm seeing more signs that such is the case)
more power to them.
I think Macsyma and Maxima might actually coexist very nicely
together. Maxima is appropriate for student lab use, where lots of
copies are needed to do relatively simple tasks, and the license costs
of a commercial system would prohibit its being used there. Then,
when those students become researchers and hit the limits of Maxima,
they can turn to Macsyma for a commercially supported and developed
system that still lets them use most of their Maxima acquired syntax
and skills. Indeed, a commercially supported Macsyma would probably
help Maxima too, since companies wanting to leverage students'
knowledge of Maxima would most likely want a commercial company to
support their software and they would be more comfortable with
Macsyma. If the commercial companies are happy, then a feedback loop
of college and business use might get started.
For this to work though, commercial Macsyma will most likely need to
be maintained. New Windows versions keep coming out, and teaching
labs don't always run Windows either. (My own undergrad lab used
Linux.) Even if the math abilities of Macsyma can be regarded as
stable, there are usually mundane issues when making it actually work.
So fingers crossed for the Symbolics folk - I hope they can make it
live again.
CY
"Richard Fateman" <·······@cs.berkeley.edu> wrote in message
·····················@cs.berkeley.edu...
>
>
> Christopher C. Stacy wrote:
> >>>>>>On Tue, 13 May 2003 15:56:22 GMT, Christopher C Stacy
("Christopher") writes:
> >>>>>
> >
> >>>>>>On 13 May 2003 14:48:03 +0200, Martin Rubey ("Martin") writes:
> >>>>>
> > Martin> * Macsyma is dead, it is Maxima now.
> >
> > Martin> The company that sold Macsyma does not exist anymore
> > Martin> (I don't know since when)
> >
> > Christopher> The company that sells Macsyma is:
> >
> > Christopher> Symbolics Technology, Inc.
> > Christopher> P.O. Box 10862, Burke, VA 22015
> > Christopher> Phone: 703-455-0430
> >
I looked in Principals of Artificial Intelligence Programming by
Peter Novig printed when Macsyma was still being activly sold.
It gave instructions for downloading Maxima from a ftp site.
sourceforge.net was not around back then--that does not
mean that Maxima was not avail. back then just harder to
locate--it had no web site, just an ftp site.
BTW, Macsyma handles much better--quicker, less likely to crash or hangup
when being used.
But, AFAIK Maxima is being activly supported--I don't know about Macsyma--&
yes you can get Macsyma for a Lispm for $500, don't know the windows ver.
price.
I don't know what the myth about Maxima taking over Macsyma users is all
about--they were both avail. in the mid 80's.
How new is the Windows ver. of Macsyma. www.macsyma.com is down--is it going
to ever be back up, maxima.sourceforge.net is not
down. (More people use Maxima now becasue it is more avail.)
goto www.symbolics.com I think they sell Macsyma for about $500.
Don't know if the Lispm ver. has extra/lacks some features of the Windows
ver.
From: Christopher C. Stacy
Subject: Re: Mathematica interface
Date:
Message-ID: <uiss9xd0h.fsf@dtpq.com>
>>>>> On Sat, 17 May 2003 16:57:07 GMT, Franz Kafka ("Franz") writes:
[a bunch of misinformation]
Why are you so compelled to spread misinformation?
What is your fucking problem?
"Christopher C. Stacy" <······@dtpq.com> wrote in message
··················@dtpq.com...
> >>>>> On Sat, 17 May 2003 16:57:07 GMT, Franz Kafka ("Franz") writes:
> [a bunch of misinformation]
>
> Why are you so compelled to spread misinformation?
> What is your fucking problem?
Why the attitude? Are you looking for Erik to tell you what
you should do with your pathetic life?
Look at on page 900 of Peter Novig's PAIP isbn=1-55860-191-0
It says Maxima could be downloaded from
rascal.ics.utexas.edu
In 1992 about 5 years before Maxima claims to be released
on maxima.sourceforge.net
This leads me to one of two conclusions that Dr. Novig was
lieing and/or invented a time-machine that allowed people
in 1992 to download software that was released in 1997 :)
or that you don't know what the fuck your talking about--
I happen to believe Dr. Novig
From: Christopher C. Stacy
Subject: Re: Mathematica interface
Date:
Message-ID: <uel2wy5k1.fsf@dtpq.com>
Most of the information that you gave was wrong.
In particular, whether Macsyma is available,
what platforms it runs on, etc. Not to mention
how many people use which versions.
Your just have crap running out of your mouth.
Aren't you embarassed knowing that it's all indelibly
recorded for posterity? What motivates you to do keep doing that?
>>>>> "Reinhard" == Reinhard Oldenburg <··········@gmx.de>:
Reinhard> Has anybody used the MathLink interface of Mathematica from
Reinhard> common lisp? Any hints of examples, software or discussion
Reinhard> papers?
I'm following up to the first posting in this thread which has gone way
off-course. I haven't seen any (serious) replies to this posting -- does
this mean that nobody has tried the above?
It turns out the original poster is not the only one interested in such
an interface, and no, Maxima is not an equivalent solution.
--J.