[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: supporting closed source operating systems?

Jeff Spaleta wrote:
On Thu, Jul 10, 2008 at 4:26 PM, Karsten 'quaid' Wade <kwade redhat com> wrote:
I am only pointing out the precedents and differences in this case, I've
no idea yet if we are in danger of slipping down a slope.

I haven't seen anyone put forward any rationale by which to limit
which libraries we distribute compiled against mingw.  The libraries
so far suggested are specifically targeted so libvirt can be more
easily developed. But there's no bright line painted as to why we
should not allow all libraries to be rebuild and packaged if people
desire them. If we let the relatively narrow set of libraries be
rebuilt for the virtualization development needs as part of our
repository, then we are opening the door for all libraries to be
rebuild and packaged. I haven't seen a credible argument to limit
mingw rebuilds to just what libvirt needs.  And quite frankly we
shouldn't be setting limits on intended use.  If we allow any mingw
built libraries into a repository we control.. then we should let all
libraries be rebuilt with mingw.


Here's why:

I want Fedora to be a premier platform for developing software. I want Fedora to attract developers of software to it. Having mingw and a non-excluded range of OSS libraries furthers this goal by allowing developers who might have had to develop on a proprietary OS or with proprietary libraries to now build on Fedora with OSS libraries instead. If these developers come to enjoy using Fedora, they will see more reason to enhance their experience on Fedora leading to more *contributors* to OSS software.

We should not get into the business
of trying to determine if one sort of development need for a library
is more worthwhile than any other.  So if we let in what libvirt needs
rebuilt in mingw, I'm pretty sure we are going to be pressed into
allowing more libraries down the line, as people find more reasons to
want to cross compile.  That's potentially a lot of additional space
to be consumed in the main repository we ask mirrors to carry.  I'm
not sure any of the mingw rebuilt items should be in the main
repository.  I'm not saying they shouldn't be allowed to exist..but I
am saying that its probably best to distribute them as their own
collection outside of the main repository

So here's a question:

We've already established that mingw doesn't fit the model of secondary arches but does it fit the model of EPEL and OLPC? Can we have a repo/sub project that has its own branches of packages in CVS and builds that target windows using mingw on Fedora? Is that something we have enough interested parties to do?

Note: This may or may not scale very well when we think of expanding it to the realm of all cross-compilers. But we probably aren't talking about potentially rebuilding every library in Fedora for the Lego MindStorm :-)


Attachment: signature.asc
Description: OpenPGP digital signature

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]