On Wed, Feb 18, 2015 at 10:40:24AM +0100, Julien Puydt wrote: > >>- they also have backported patches to add cb_pari_err_handle from > >>upstream's repository (see end of this mail). > >> > >>Can we get those in debian too? > > > >This patch is breaking the ABI and API, thus this is not possible to > >apply it. > > Since it's backported from upstream, it will be in some version at > some point : do you have an ETA?
I expect PARI 2.9 to be released in about one year. But at this point SAGE will require other patches to be applied anyway. > >Beside SAGE could do the same thing without patching PARI, by > >using the standard PARI error trapping mechanism (iferr_env and > >global_err_data). This would be much better. > > That would be the sensible thing to do, but sage developers have a > terrible tendency to patch upstream instead of using upstream. On > some occasions, I have proposed them to replace an upstream patch > with a one-line patch to their code, because the api to configure > what they wanted was right there... Yes, they should create a small C library that would be linked with libpari and would provide the changes they like instead of patching libpari directly. > It looks like they've been toying with pari error handling for months: > http://trac.sagemath.org/ticket/14894 I have been trying to explain to the SAGE developers how to use pari error handling since 2005... > Is there a better way to do what they want ? I expect they should do something like if (setjmp(&iferr_env)) { GEN E = pari_err_last(); char *s = GSTR(GENtoGENstr(E)); fprintf(stderr,s); ... } before the point they are recovering using cb_pari_err_recover() at this time. Cheers, -- Bill. <ballo...@debian.org> Imagine a large red swirl here. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
