Hi Camm, thanks for your reply.
I need to do some more test runs to get logs of the additional failing tests. The last time we tried using maxima-gcl was over a year ago. Can we also create a fasl file with GCL and would you consider putting that into the package? Then I could already test with such a file. This is also complicated by the version mismatch, sage still uses maxima 5.35. In a first test I already got 74 new failing tests because I switched to maxima-ecl 5.38. Many of them were caused by changes in the formatting of maxima output, so the tests need to be updated. (Sage has both a pipe interface to maxima and the fasl interface and I think they are both used.) Anyway, I have to figure out which failures are caused by the changed version and which by GCL. Best, Tobias On 10/16/2016 10:19 PM, Camm Maguire wrote: > Greetings! > > Why don't you send me a description of what you've done with current > maxima and sage so far and what errors you see, so I can take a look. > > Take care, > > Tobias Hansen <than...@debian.org> writes: > >> Hi Camm, >> >> I updated the debdiff to maxima 5.38.1-2. >> >> I really hope we can make some progress on this topic. I read again the >> private mail that you sent me in March 2015 and I must say that I don't >> really understand everything. We tried using maxima-gcl with Sagemath, >> but we get Sagemath test failures that are not trivial to fix for us. I >> also think that we need the fasl file. (Is that what you called >> common-lisp-controller library in your mail?) Unfortunately we have >> nobody with any Lisp experience. Because of this, changing the way >> Sagemath and maxima work together seems hard. >> >> We are currently working a lot on packaging Sagemath for Debian stretch, >> as can be seen from the amount of traffic on our mailing list [1]. I >> think we have a realistic chance to get everything that is required done >> before the freeze. >> >> maxima is now one of two issues that we don't really know how to >> resolve. (The other one is pari, which is resolved if pari 2.9 is >> released in time for stretch.) >> >> Best, >> Tobias >> >> [1] https://lists.alioth.debian.org/mailman/listinfo/debian-science-sagemath >> >> >
