Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 29 May 2004 03:04:12 +0200
From:      Oliver Eikemeier <eikemeier@fillmore-labs.com>
To:        Ade Lovett <ade@FreeBSD.org>
Cc:        FreeBSD ports <ports@FreeBSD.org>
Subject:   Re: Ports with a wrong PKGORIGIN
Message-ID:  <40B7E18C.1050907@fillmore-labs.com>
In-Reply-To: <A11C5230-B104-11D8-B3B3-000A956B6386@FreeBSD.org>
References:  <200405282358.i4SNwfeF061182@8ball.rtp.freebsd.org> <A11C5230-B104-11D8-B3B3-000A956B6386@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
Ade Lovett wrote:

> On May 28, 2004, at 16:58, Oliver Eikemeier wrote:
> 
>> ** The following ports have a wrong PKGORIGIN **
> 
> [snip]
> 
> Any particular reason these are not either going to the port maintainer, 
> or at least cc'd to them?  Perhaps also the committer who last touched 
> the Makefile.  Might make things a little more directed.

Send a patch, the script is at ports/Tools/scripts/chkversion.pl ;)

PKGORIGIN and PKGVERSION errors are mostly obvious typos, so every
committer should be obliged to fix them in order to keep the tree
consistent, it shouldn't be a question of `ownership'.

I can understand that the last committer is held responsible for
build errors that are sometimes hard to trace, but concerning
PKGORIGINs there is only one obvious choice. It's like fixing an
INDEX build when the last committer typed devel/getext before going
to bed: there is no point in waiting till the next morning, just
fix it.

Another problem is that the wrong PKGORIGINs might occur in slave
ports or after a repocopy, where the last committer has nothing to
do with the resulting problems.

> [snip]
> 
>> - *games/sarien* <hsn@netmag.cz>: emulators/sarien
> 
> Also, the fact that your script is generating this line suggest that it 
> would be a SMOP to offer up a new CATEGORIES line in the report, or even 
> as a diff against the Makefile to make things even simpler.

Should be trivial: just make `games' the first category. Good idea, I'll
have a look into this in the next revision.

I expect the error rate dropping rapidly when people are a little more
aware of the importance of PKGORIGINs.

> Depending on how reliable it got, it could (possibly, perhaps, no 
> stonings yet, please) even auto-commit.

Hmmmm... dangerous. And I have to leave my ssh key in the custody of
a script, which might be a security breach.

> Just a few random thoughts.

Thanks for the feedback.

-Oliver



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?40B7E18C.1050907>