On Tuesday, March 27, 2018 at 8:39:50 AM UTC-4, Erik Bray wrote: > > On Sun, Mar 25, 2018 at 4:49 PM, Simon King <simon...@uni-jena.de > <javascript:>> wrote: > > Hi! > > > > On 2018-03-25, Volker Braun <vbrau...@gmail.com <javascript:>> wrote: > >> On Sunday, March 25, 2018 at 2:51:46 PM UTC+2, Dima Pasechnik wrote: > >>> > >>> one can install autotools on archlinux systemwide > >>> > >> > >> Also, autotools aren't even required to build Sage. > >> > >> Whats the point of delaying a release for weeks/months to fix an > optional > >> package? Presumably you agree that broken standard packages should have > the > >> highest priority, so broken optional packages necessarily have a lower > >> priority. > > > > In my previous post I didn't understand that the actual point of this > > thread apparently is not "How much support are we supposed to provide > for > > optional packages?", but "Are optional packages important enough to > > constitute a blocker?". > > > > As a maintainer of a former optional package, I repeatedly found it > > annoying that the package got broken by changes in SageMath. It feels > > like EACH TIME I fix upstream to make it work with the latest Sage > version > > and open a ticket for upgrading the package, BEFORE THE REVIEW IS > FINISHED > > there will be yet another change in Sage that breaks the just-fixed > package > > yet again. For that reason, I lost the impetus at some point, and now > the > > package doesn't properly work since several years. > > > > Therefore I do believe that Sage development should treat optional > packages > > with more respect. > > I agree--I think there should be at least one buildbot run > per-platform--maybe not for every issue but at least run once a week, > that tests building all optional packages and running tests that use > them (i.e. tagged with # optional - <packagename>). > > A build with broken optional packages could then be considered a broken > build. > > We also tried to improve this by having "existing" optional packages tested by default in users' setups when sage -testall or whatever was run, but maybe that hasn't yielded the bevy of bug reports it should have. Yes, if an optional package fails on all reasonable platforms we/build creator has access to, that is a blocker. Simon has been remarkably patient regarding his package, by the way.
-- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at https://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
