On Sunday, March 25, 2018 at 9:57:20 PM UTC+1, Dr David Kirkby wrote: > > > On 25 March 2018 at 18:06, William Stein <wst...@gmail.com <javascript:>> > wrote: > >> >> opinions. >> >> My initial intention with ooptional packages was definitely that they >> do *not* have as much support as standard. >> > > The developers guide. > > http://doc.sagemath.org/html/en/developer/packaging.html > > says: > > " > > - *optional* packages are subject to the same requirements, they > should also work on all supported platforms. If there are optional > doctests > > <http://doc.sagemath.org/html/en/developer/doctesting.html#section-optional-doctest-flag> > > in the Sage library, those tests must pass. Note that optional packages > are > not tested as much as standard packages, so in practice they might break > more often than standard packages. > - for *experimental* packages, the bar is much lower: even if there > are some problems, the package can still be accepted. > > " > > I would interpret that as meaning if a test reveals it does not work, then > it should not be optional, as optional packages should work on all > supported platforms. In this instance, it seems there are some problems, so > it should either be fixed or changed to experimental. > > > TLDR; "supported platform" and "blocker ticket" are merely engineering terms. There are not and cannot be as precise as mathematical theorems :-) The list of supported platforms (https://wiki.sagemath.org/SupportedPlatforms) is severely outdated. As well, no attempt to promise that a particular version of gcc (or other compiler) can be used to build Sage on a particular supported platform is made.
In such a situation, the statement "Sage is fully supported on platform X" has at best only fuzzy meaning. And it's OK, as it seems that the commitment to fully support Sage on every flavour of Linux out there is not realistic, no matter how popular this flavour is. From this (and other) discussions here, it seems that "X is supported" came to mean "there is at least one patchbot running platform X". This is too rigid and not really the case, in my view. I would actually like to see the definition of "supported platform" to be made realistic, and mean "popular distributions of Linux on x86/x86_64, ..." without any particular emphasis to versions, toolsets used, etc. (e.g. the already recognised fact that compilers can be broken ought to be extended to other tools, e.g. make --- cf. the recent ticket related to guile-supporting make on Archlinux) ------------------------ Given this, there should be no tickets made blockers merely on the basis that Sage broke on your favourite patchbot or laptop... Dima -- 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.
