Migration to new SourceForge url scheme now inevitable, solution

[Dmitry Marakasov]

* Pav Lucistnik (pav at FreeBSD.org) wrote:

> > > I've been following this discussion closely since several of my ports fetch 
> > > from Sourceforge.  Is it safe to assume that some global solution will be 
> > > applied to the ports tree?  Or are we maintainers going to need to submit PRs 
> > > for affected ports once a solution is agreed upon?
> > 
> > This should be done globally, or else we'll end up with 90% unfetchable
> > SF ports for 8.0 release. I'm preparing the patch currently.
> 
> Once you have a patch, send it over for eyeball-review and approval.
> Thanks for attacking this!

Automated run is done, patch is here: [1]. It's not the final
version, as it turned out to be more complex as I thought. There
are things like SF:foo SF:bar (we can do this, right?), and
${MASTER_SITE_SOURECFORGE:C/$/:foo}, which should be fixed manually.
Also there are slave ports, for masters of which you don't want to
substitude ${PORTNAME} in some cases. There's a log [2] which lists
potential issues. Tomorrow I'll look them through and manually fix
remaining ports, run a quick fetchability test and present a final
patch.

[1] http://people.freebsd.org/~amdmi3/sfp.patch
[2] http://people.freebsd.org/~amdmi3/sfp.log

PS. Btw, SOURCEFORGE_EXTENDED and SOURCEFORGE_JP still use an old
scheme. Because of that SFE can no longer include SF, and honestly, I
would like it to go away, as there are more than enough official SF
mirrors. Also need to check if our SF mirror list can be extended with
more mirrors, last time I had that `select mirrors' screen on
sourceforge it looked like there are much more mirrors than we have. 

-- 
Dmitry Marakasov   .   55B5 0596 FF1E 8D84 5F56  9510 D35A 80DD F9D2 F77D
amdmi3 at amdmi3.ru  ..:  jabber: amdmi3 at jabber.ru    http://www.amdmi3.ru