On Fri, 17 Dec 2004, Greg A. Woods wrote:

> > You're more than welcome to find us twice to four times the head count of
> > *experienced* pkgsrc users to handle pkg-submission PRs, including field
> > testing and cleanups.  That's what pkgsrc-wip does:  it brings the packages
> > into a repository where others can do field testing and help clean them up,
> > rather than letting them stagnate in a uuencoded tarball.
>
> How many people really use pkgsrc-wip for testing?  Do you/we have any
> way of even guessing?

I don't think we can get a real number of "users" as SourceForge doesn't
have aggregate CVS checkout stats.  (Not that such stats would be revealing
anyway, given that many users of pkgsrc and pkgsrc-wip have dynamic IPs.)

There are, however, 136 registered developers with commit privileges to
pkgsrc-wip.  You could be one too.  8-)

> Personally I won't touch pkgsrc-wip with a 10-foot pole.

Then you can sit and wait for things to show up in main pkgsrc and we'll
choose to keep using what works.

> I suppose those who really love the bleeding edge will appreciate it, but
> it seems like a lot of effort just for that.

It's certainly better than having packages stagnate in PRs because of lack
of human time to review them.  And since this thread was about package
submissions stagnating in PRs, I think the point is blindingly obvious.

> > In the process, we've actually gained some main pkgsrc folks who "learned by
> > doing" in the pkgsrc-wip tree.  The pkgsrc-wip effort is actually quite
> > successful so far.
>
> I'm not sure I see how that's any different than what might be achieved
> by simply being more liberal with commit privileges in the official
> tree.

Less chaos in main pkgsrc.  Those folks who are new to adding things to
pkgsrc will be less likely to trip up and cause pkgsrc-wide problems if they
are sandboxed into a more liberal source tree.  (As a sysadmin, you should
certainly understand that stance.)

If you look at some of the stuff in pkgsrc-wip, or even some of the stuff
sitting in wait in PRs, you'd understand why this is necessary.  pkgsrc is
not an easy beast to learn, and it has many pitfalls.  Major mistakes are OK
in pkgsrc-wip, since people are in the process of learning the ins and outs
of pkg creation.  Major mistakes are definitely not OK on a broad basis in
main pkgsrc.

So pkgsrc-wip gives newer users a playground to create and test packages,
allowing broader peer review and collaboration than would typically be
available through stagnant PRs.

Did you know that X11_TYPE=pkgsrc (using X11 from the x11/XFree86-*
packages, rather than the base system) was implemented first in pkgsrc-wip?
So were quite a few other concepts now in use in main pkgsrc.

-- 
-- Todd Vierling <tv@duh.org> <tv@pobox.com>