Dear all.
I tried to compile the Maxima computer algebra system (version 5.16.1)
with Lispworks Personal 5.1.1. Running configure.lisp, (maxima-compile)
and (maxima-load) all passed without problems, but after (cl-user::run)
I got the followiung error message:
Undefined function HARLEQUIN-COMMON-LISP:GETENV called with arguments
("MAXIMA_LANG_SUBDIR").
Does anybody know how to circument this?
Yours,
Reinhard Oldenburg
On Sun, 17 Aug 2008 19:30:55 +0200, Reinhard Oldenburg <oldenbur@math.uni-frankfurt.de> wrote:
> I tried to compile the Maxima computer algebra system (version
> 5.16.1) with Lispworks Personal 5.1.1. Running configure.lisp,
> (maxima-compile) and (maxima-load) all passed without problems, but
> after (cl-user::run) I got the followiung error message: Undefined
> function HARLEQUIN-COMMON-LISP:GETENV called with arguments
> ("MAXIMA_LANG_SUBDIR"). Does anybody know how to circument this?
You probably want to use LW:ENVIRONMENT-VARIABLE instead.
huh...interesting...I didn't know a version of Macsyma was even still around...20+ years ago I traveled to Los Alamos (with my machine) and ported DOE-Macsyma to the TI Explorer Lisp Machine...twice (first time was Rel 2 (1986), second was the much more CL Rel 3 (1988)). one of the things I did was take a self-test mechanism (which I think originated with Bill Shelter at UT, but didn't work, and his macsyma attempt didn't work either) and make it behave cleanly, and create a set of self-tests which you could batch-run to verify build-tests. this was important because that was some god-awfully sloppy code, that was also just about incomprehensible, and written for previous dialects of Lisp that had subtle differences (heck, on the Explorer itself I got weirdly different behaviors, depending on how I compiled things (defsystem vs file vs function) that took me a while to figure out) haven't thought about that in years...wonder if I still have that old monster on backup tape...it'd be a shame for that code/tool to just vanish...there ought to be a way to "productize" an open version of it, I'd think. what is the origin of this Macsyma version? -- clint hyde On Aug 17, 2008, at 1:30 PM, Reinhard Oldenburg wrote: > Dear all. > > I tried to compile the Maxima computer algebra system (version > 5.16.1) with Lispworks Personal 5.1.1. Running configure.lisp, > (maxima-compile) and (maxima-load) all passed without problems, but > after (cl-user::run) I got the followiung error message:
Clinton Hyde <chyde@cox.net> writes: > huh...interesting...I didn't know a version of Macsyma was even still > around...20+ years ago I traveled to Los Alamos (with my machine) > and ported DOE-Macsyma to the TI Explorer Lisp Machine...twice (first > time was Rel 2 (1986), second was the much more CL Rel 3 (1988)). > > one of the things I did was take a self-test mechanism (which I think > originated with Bill Shelter at UT, but didn't work, and his macsyma > attempt didn't work either) and make it behave cleanly, and create a > set of self-tests which you could batch-run to verify build-tests. > this was important because that was some god-awfully sloppy code, > that was also just about incomprehensible, and written for previous > dialects of Lisp that had subtle differences (heck, on the Explorer > itself I got weirdly different behaviors, depending on how I compiled > things (defsystem vs file vs function) that took me a while to figure > out) > > haven't thought about that in years...wonder if I still have that old > monster on backup tape...it'd be a shame for that code/tool to just > vanish...there ought to be a way to "productize" an open version of > it, I'd think. > > what is the origin of this Macsyma version? Bill Shelter got the DOE to open up the source of the 1982 version under the GPL. It's still enhanced and maintained by a group of people after his death. Too bad all the Symbolics enhancements and man years of development on it will be locked up forever. -- The method of "postulating" what we want has many advantages; they are the same as the advantages of theft over honest toil. - Russell
Rainer Joswig <joswig@lisp.de> writes: > Am 19.08.2008 um 19:37 schrieb Lieven Marchand: >> Bill Shelter got the DOE to open up the source of the 1982 version >> under the GPL. It's still enhanced and maintained by a group of people >> after his death. Too bad all the Symbolics enhancements and man years >> of development on it will be locked up forever. > > Well, that's only a question of money. Currently no buyer with > sufficient cash has been identified. AFAIK that's the main problem. You're probably better informed about Symbolics than I but I'd heard the Symbolics assets were tied up in probate with the death of the previous owner. Of course, that's also nothing cash and lawyers can't solve but it can make it a lot more expensive. -- The method of "postulating" what we want has many advantages; they are the same as the advantages of theft over honest toil. - Russell
On Tue, 19 Aug 2008 21:40:42 +0200, Lieven Marchand <mal@wyrd.be> wrote: > You're probably better informed about Symbolics than I but I'd heard > the Symbolics assets were tied up in probate with the death of the > previous owner. The problem is that the previous owner (who is now dead) owed the IRS a lot of money and there's no heir, so the executor keeps the assets and tries to find someone who is willing to pay enough money for them to make the IRS happy. It seems this hasn't happened so far and probably it never will. Edi.
> The problem is that the previous owner (who is now dead) owed the IRS > a lot of money and there's no heir, so the executor keeps the assets > and tries to find someone who is willing to pay enough money for them > to make the IRS happy. It seems this hasn't happened so far and > probably it never will. > > Edi is right. This is being handled, and there's nothing anyone else can do about it. Please do NOT try to get involved, or else the trustees of the estate will think that they can start a bidding war. We just have to wait until they give up. -- Dan -- ________________________________________ Daniel Weinreb dlw@alum.mit.edu http://danweinreb.org/blog/ http://ilc2009.scheming.org/
Tim, Tim Bradshaw wrote: > > Having used several Symbolics systems (I'm in the, very > long-drawn-out, process of giving my last SB machine away) as well as > LispWorks, both fairly extensively I'll offer a controversial opinion: > for most purposes LW is already as productive a development > environment as the LispM was. I think the idea was just to pick up the changes that were made to Macsyma by Symbolics, rather than use Genera as a development environment. > > > It might be that Macsyma would still be interesting (but, well, how > does it compete with Mathematica as a general purpose algebra system, Indeed. My impression was always that you had to be rather expert to use Macysma. However, there are still users out there. > and how large is the market for algebra systems anyway?), but I think > the rest of the assets of Symbolics would have negligible, or more > likely negative, value to LW. Yeah, I'm also dubious that much of Genera could be used easily by anyone there days. There were lots of good ideas in the software that one could learn about by using it, but actually transplanting the code may not be that useful. -- Dan -- ________________________________________ Daniel Weinreb dlw@alum.mit.edu http://danweinreb.org/blog/ http://ilc2009.scheming.org/