On Nov 23, 1:03 pm, "John Cremona" <[EMAIL PROTECTED]> wrote:
Hi, > Before I jump on the bandwagon, are there any arguments at all against > this? Not that I can see. The main reason we chose clisp over gcl for example that it worked better than any other lisp implementation supported by Maxima (and I am being generous when I write 'working' here). There are other lisp implementations out there that support building Maxima and are even faster than ecl for many things, i.e. sbcl for example, but those are not self hosted, i.e. without a lisp compiler you cannot compile sbcl. Since back in the day when Maxima was added to Sage it did not support ecl the only choice was clisp and considering the number of build and run time problems we had it is a miracle that it took so long to dump clisp. To add to the list of clisp issues: Most gcc 4.x releases lead to Maxima segfaulting when clisp is build with more than "-O0", so performance is worst than it could be. On Solaris/Sparc *no* gcc after gcc 3.3 seems to result in a working clisp binary. I have also played with clisp 2.46 and when I run Maxima's test suite under valgrind I found a couple bad things that clisp was doing. Unfortunately I no longer have that test setup and I must have deleted the valgrind logs, so I never reported the issue upstream. One problem with ecl is that it requires a working gcc when you want to compile lisp code, but the disadvantage here is greatly out weight by the fact that all the bits we need from Maxima "just work" - so I couldn't care less about the lisp compilation problems. I would assume that the number of people installing Sage to use it for its lisp are about zero :) So in the end that leads me to believe that the Sage project is gaining here by replacing clisp by ecl. > I can see the benefits, and the only cost is in someone other > than me making it work...which makes it rather easy to vote +! Yes, I always vote +1 on nice things other people do :) > John Cheers, Michael --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
