Minh Nguyen wrote: > Hi David, > > On Thu, Aug 13, 2009 at 5:03 PM, Dr. David > Kirkby<david.kir...@onetel.net> wrote: >> Juanjo, >> >> the main ECL developer has made new source release in the last 24 hours. >> I've created a new ECL spkg. You can find it in the directory: >> >> http://sage.math.washington.edu/home/kirkby/Solaris-fixes/ecl-9.8.3/ >> >> I've updated >> >> http://sagetrac.org/sage_trac/ticket/6564 >> >> which has the title "[with spkg; needs review] Update ECL to the latest >> 9.8.3, needed for Solaris." >> >> I've rebuilt ECL and Maxima on Solaris and all seems to go ok. > > Given that t2.math is down at the moment, reviewing your spkg would > have to wait for some time.
The Sun T5240 (hostname 't2') is back up now, so the *building and installation* of ECL can be tested. It should build and install without issues. However, ECL 9.8.3 can not be added to Sage without other changes, due to a bug in the version of Maxima in Sage, which is Maxima 5.16.3. An improved parser in recent ECL releases detected the Maxima code was invalid and would not accept it - unlike all the other LISP interpreters out there, which are less fussy. Juanjo (the main ECL developer) made the Maxima developers aware of this and sent them a patch. So the latest Maxima (5.19.0) does not have this bug. So whilst it is possible to test the building and installation of ECL easily, one can't use this package in isolation. Both Maxima and ECL have to be updated together in order to get a work Sage build. http://sage.math.washington.edu/home/kirkby/Solaris-fixes/maxima-5.19.0/maxima-5.19.0.spkg http://sage.math.washington.edu/home/kirkby/Solaris-fixes/ecl-9.8.3/ecl-9.8.3.spkg Some changes in Maxima cause some doctest failures. One is simply that Maxima can now evaluate an integral it once could not. File "/export/home/drkirkby/sage/sage-4.1.1.rc0/devel/sage/sage/calculus/functional.py", line 216: sage: integral( exp(-x^2)*ln(x), x, 0, oo) Expected: integrate(e^(-x^2)*log(x), x, 0, +Infinity) Got: -1/4*(euler_gamma + 2*log(2))*sqrt(pi) others are due to improved formatting. But there are other issues too, so changes will need to be made to other packages and doctests. This updating of ECL and Maxima for Solaris support is not trivial. Note, I do not have any patch which can make the old version of Maxima in Sage work with the new ECL. I looked at trying to create one, but it needs more ECL knowledge than I have. In any case, I think an update to Maxima is well overdue. But it needs work on some packages outside of both ECL and Maxima. > I've had little to not input at all in > porting any of the spkg's to Solaris. But I really appreciate that you > have taken the time to do the porting work. Reading through your > comments and emails, I've learnt many valuable tips. One in particular > is to avoid Bashism in shell scripts so as to make one's shell scripts > more portable to other platforms. This has made me more aware of > non-portable features I use my shell scripts, prompted me to read up > on ways to write portable scripts, and helped me to learn about > virtues of portability. A big thank you to you. Thank you. I've often thought it would be a great idea if Sun made available machines for public access so people could test their code on Solaris. HP/Compaq used to do this. I made use of their systems to test some code on operating systems I did not have access to. Clearly Sun have gone some way to this with the donation of the Sun T5240 to the Sage project. But there are many programs out there (nothing to do with Sage) which don't work well on Solaris, due to bad programming. I do have at home machines which have AIX, tru64, HP-UX and IRIX installed. I personally have no interest in starting a port to any of them, but others might do. IRIX and tru64 are effectively dead, but AIX and HP-UX are not. Apparently some of the new p-series machines from IBM have processors running at 5 GHz. (Perhaps IBM would donate one to the Sage project!) > If we ever meet in > meatspace, either at a Sage Days conference or somewhere else, I would > buy you your favourite beverage. > Cheers. I'll take you up on that one. Dave --~--~---------~--~----~------------~-------~--~----~ To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
