On Dec 25, 12:39 am, "William Stein" <wst...@gmail.com> wrote:
> On Wed, Dec 24, 2008 at 10:44 PM, Justin C. Walker <jus...@mac.com> wrote:
>
>
>
>
>
> > On Dec 24, 2008, at 20:36 , William Stein wrote:
> >> On Wed, Dec 24, 2008 at 1:48 PM, mabshoff <mabsh...@googlemail.com>
> >> wrote:
> > [snip]
> >>>>>> I want the Sage "optional" packages to *all* install fine on every
> >>>>>> system that we officially support Sage on.
>
> >>>>> No chance this is going to work.
>
> >>>> 1. Just because you don't know how to do something doesn't mean it
> >>>> can't work.
>
> >>> You are kidding me?
>
> >> Obviously we're misunderstanding each other here. I'm not claiming
> >> that every spkg should be *supported* to run on every OS. However,
> >> when one tries to install on a non-supported platform, the failure to
> >> install should be very graceful.
>
> > I don't think you can achieve this. To be able to fail gracefully on
> > a randomly chosen platform means that you will (and will continue to)
> > test on every conceivable platform, to assure that the failure will be
> > graceful. Even "supported" platforms may have configurations that
> > were not imagined at the time a release is put together, so saying
> > that *any* platform will be supported to some level of "grace" is a
> > significant commitment.
>
> > Do you mean something other than the above?
>
> I definitely do not mean the above. Here's what is stated in the official
> Sage README.txt:
>
> "OFFICIALLY SUPPORTED PLATFORMS:
> Building of Sage from source is regularly tested on
> (minimal installs of) the following platforms:
>
> PROCESSOR OPERATING SYSTEM
> x86 32-bit Linux -- Debian, Ubuntu, RHEL5, Fedora
> Core, CentOS, Suse, Mandriva
> x86_64 64-bit Linux -- Debian, Redhat, Suse
> ia64 itanium2 64-bit Linux -- Redhat, Suse
> x86 Apple Mac OS X 10.5.x
> ppc Apple Mac OS X 10.5.x
>
> Use Sage on Microsoft Windows via VMware.
> We do not always test on OS X 10.4, but Sage should work there fine."
>
> We will be doing the following:
Why? I did not see any discussion about this. I am not opposed to more
testing or properly defining the term "optional" or making known
broken configurations fail gracefully, i.e. phc on Itanium/Linux [i.e.
I am in spirit pretty much on board of what you want to do], but
starting a white list of known good configurations seems wrong. There
are many more platforms out there than we can conceivably test on, so
allowing the occasional build failure seems worth the trouble. As long
as issues are reported we can at least try fix them. What do you
propose to do on Linux distributions where we do not have access to?
Issue a warning? Refuse to build? Make the refusal overwritable via
some env variable? I just think this warrants some discussion before
going off and ripping the optional repos apart despite you BDFL
status ;)
> (1) make the above list much more precise, by nailing down the exact
> OS version and hardware in each case, and deleting the parenthesis
> around "minimal installs of".
Things will still go wrong since one can easily construct
configurations that blow up. Making it more precise by saying we test
regularly on $FOO, but please report any issues seems much more
sensible.
> (2) Test installation of all optional packages on those platforms
> before each release.
+1
> -- William
Another thing: For whatever reason you decided to violate policy and
put a bunch of unreviewed spkgs (I counted three, but I didn't look
too hard) into the optional repo, two of which (polymake and
pyopenssl) are broken [pyopenssl ironically on sage.math :)]. I just
overwrote polymake with a working and reviewed copy, while I have an
improved version of pyopenssl on the ticket. I do not think review
policy should be violated and the argument that "things are better
than before" does not fly since nobody cares for the KASH.spkg for
about two years we could have easily waited a couple more days.
Cheers,
Michael
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to
sage-devel-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
