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

Re: Is Sid for broken stuff? Is it too much to ask for testing the packages?



On Thu, Dec 12, 2002 at 03:13:40AM +0100, Bernd Eckenfels scribbled:
> On Thu, Dec 12, 2002 at 02:53:15AM +0100, Marek Habersack wrote:
> > trap "mail -s "$MAILSUBJECT" root@localhost < ${MAILFILE}; rm -f $MAILFILE"
> > 
> > Do you honestly think that the maintainer installed script with that code
> > and that the code worked for him? 
> 
> if it is only conditionally executed this might be the case, yes.
Agreed. But if you add some new code, even conditionally executed, then it
is fair to assume that you test the code before letting anybody else use it?
This wasn't the case here.

> I do install all my packages before uploading (on a testing system most of
> the time). Most of the time I also run:
> 
> (last version is on system)
> - update
> - remove
> - update
> - purge
> - install
I do the same.

> But even in that sequence a broken purge script could leave stuff behind,
> whih will make a broken install script to work, whereas all other users will
> get a installation failure (for example directory missing). Sometimes I may
True.

> also miss some update testing, since the current package I upgrade is not
> the last released version but my last working version. So if I did multiple
> internal versions of a package and the last change introduced an error which
> only occurs while upgrading older packages, this would be untrapped.
Also true, but note that the cases I pointed out were all much, much, much
simpler.

[snip]
> Of course packages with less dependecies tend to be very forgiving, but you
> never know.
On another note - I expect the postgres maintainer to be an especially
cautious and responsible person, because the package he maintains is
depended upon by dozens of other packages and any failure to test new
functionality of the packaging scripts or the upstream code might cause a
chain reaction resulting in a lot of work for other developers, which is
definitely not right. Therefore I propose that if a maintainer lacks time to
properly maintain a package then he should either find somebody to help
her/him in maintaining the package (for example find people who would be
willing to test the packages before s/he uploads them) or to abandon the
package and pass it to somebody with more time in their hands.

marek

Attachment: pgprxY8jNKPRL.pgp
Description: PGP signature


Reply to: