Re: Comment

[Steve Langasek <vorlon@debian.org>]

On Tue, May 29, 2007 at 01:13:25PM +0800, Uwe Dippel wrote:
> Steve Langasek wrote:

> >Anyway, this is a bug in a specific package.  The QA team isn't 
> >responsible
> >for finding all bugs in individual packages; ask the maintainer of the lpr
> >package why it wasn't noticed before release...

> Steve, you get me / it wrong. QA isn't responsible for finding all bugs. :)
> QA is responsible, though, for QA.

There is no charter for the QA team which implies that we have assumed
responsibility for testing Debian in any particular configurations or with
any particular hardware.  By and large, the QA team works on *known* bugs
that have gone unaddressed by package maintainers, as well as picking up
packages that have no maintainer and working on systematic *automated*
detection of bugs in the archive.  Sorry, but "do parallel printers work"
just doesn't fall under that umbrella.

> And at least in my case, all three boxes broke beyond repair by means
> Aunt Tilly has at her disposition. And that ought better not happen
> ... :)

"Aunt Tilly" isn't likely to be setting up lpr in the first place.

> >>Nvidia legacy is broken, actually preventing me from 
> >>using that system at all (#423592).

> >This is a bug report filed on a package that's not in etch.  Not sure what
> >you would expect the QA team to do about it.

> It very much *is* in Etch. I am getting it from the repository of Etch; 
> and it was working in Sarge.

No, this bug report is filed on the package nvidia-glx-legacy-71xx, and
there is no package by that name in etch.  Perhaps this bug also affects
some package that is in etch, but that's not clear from the state of that
bug report.  (And I don't use nvidia binary drivers anyway, so I'm
disinclined to go digging myself to figure out which package that is.)

> >>My individual settings for xfce4 are 
> >>completely lost, actually making it impossible to use xfce4 with the old 
> >>settings; since they prevent the xfce4 menu from appearing.

> >Not something that I ever like to see, but it's not at all unheard of for
> >upstream changes to invalidate user configurations from previous versions
> >without providing an upgrade path.  Debian doesn't have the resources to
> >compensate for every upstream who leaves their users in the lurch like 
> >this,
> >sorry.

> Again, you get it / me wrong. Being a FOSS developer myself (though at a 
> very humble level), I am too well aware about such predicaments.
> There could have been a compensation, though, kind of: Either leaving 
> functional packages of previous versions hanging about; or; and this is 
> the least, to solicit the user input before overwriting the binaries; 
> warning that xfce4 will be non-functional and no settings saved. Then 
> you'd have .xfce4 moved to .xfce4_saved or similar, and informed the 
> user likewise.

For all I know (as someone who doesn't use xfce4), it might be that no one
else had problems with their particular configurations migrating over from
sarge to etch, though?

-- 
Steve Langasek                   Give me a lever long enough and a Free OS
Debian Developer                   to set it on, and I can move the world.
vorlon@debian.org                                   http://www.debian.org/