It would be great to have a texmacs interface a la Q and - http://www.texmacs.org/tmweb/manual/webman-write-itf.en.html with cmd line option "-texmacs" (see Remark 2.2 in the above document).

Maybe using pure_norl.cc ? Priority: low

ty:)

This issue was closed by revision 01b932fa9064.

I've added some minimal support now, see http://docs.pure-lang.googlecode.com/hg/install.html#texmacs-mode.

This is really rather minimalistic but seems to work fine so far. If you have any suggestions on how to improve it then please let me know (patches are even better ;-).

BTW, while testing I noticed that at least with the stock Ubuntu 11.04 package of texmacs (1.0.7.7), texmacs is slow like molasses on startup. (Once it's running, it seems snappy enough, though.) Do you know a remedy for that?

Dear Dr. Gr�f that was pretty quick :) believe me, it wasn't my intention to urge you in any way but I'm truly grateful to have it.

I'm going to rewrite all my FriCAS/Yacas code to Pure (ambitious but feasible; it will be worthwile since the "rule" command of Axiom/FriCAS is miles away from Pure's capabilities) so I'll try to copy the FriCAS texmacs mode, which is based on scheme commands instead of latex. Otherwise a good tex() function in Pure would be useful too, maybe I'll try this approach first. I'll share all my code with the community as soon as it is useable but for the moment I'm still in the learning phase ;)

Regarding Ubuntu/Texmacs: I have tried Ubuntu 12.04 with Texmacs 1.0.7.14 and it works perfectly (in fact it's quite rapid). I think that maybe Ubuntu 11 but for sure Texmacs 1.0.7.7 are simply too "old" (current texmacs version is 1.0.7.16).

Again, many thanks and best wishes Kurt

Am 24.09.2012 16:20, schrieb pure-lang@googlecode.com:

Am 24.09.2012 16:27, schrieb pure-lang@googlecode.com:

I compiled the dev code in VirtualBox Ubuntu 11.04, testing with Texmacs 1.0.7.7 but I couldn't recognize any delay in startup.(?) Nevertheless, I'm sure 1.0.7.7 is out of date and one shouldn't use versions before 1.0.7.14.

sorry for my messy English in comment#2: I meant of course "grateful to have the interface now" not "having urged you ..."

Ah, thanks for the info. Maybe I'll have to grab the latest source package from Launchpad and rebuild it myself.

I just downloaded the latest source from the texmacs website and compiled it myself; works great now. Apparently the older version did some extra stuff to look up ttf files on startup, which took a long time with the big font directories I have here.

Concerning the porting of FriCAS/Yacas code; I agree that it's ambitious. ;-) The term rewriting stuff is readily available in Pure, of course, but you'll probably also need at least a few other algorithms such as Buchberger and Risch to do anything useful in computer algebra. Anyway, having some more comprehensive examples from this realm would be great, so I'm looking forward to it!

Albert

Ah, good, I supposed something in this direction.

I meant rewriting my own code, not parts of those systems (that could really be a life's work :). I was using almost only the rewriting capabilities of those applications and only very few algebraic built-ins. The code deals mostly with the calculus of constructions and exterior algebras without specified dimensions, so ordinary CAS are useless and rewriting is the clue.

Kurt

Am 24.09.2012 22:01, schrieb pure-lang@googlecode.com:

Ok, I see. Yes, I guess that this makes it somewhat more feasible. :) So good luck, let me know how it goes.

Am 24.09.2012 22:27, schrieb pure-lang@googlecode.com:

Thanks, it's about 20k code lines.

But you are right, there are many algebraic algorithms suitbale for rewriting and a simple CAS wouldn't be really hard to implement. Although Axiom for example would be out of reach (the type inference alone would be quite a task), I dare to allege that Yacas, for example, could be implemented with ease, since it's library is entirely rule based http://sourceforge.net/p/yacas/code/2623/tree/trunk/scripts/. BTW it's GPL. Kurt