On Thu, 26 Sep 2002 11:39:36 +0200 Thierry Vignaud
<[EMAIL PROTECTED]> wrote:

> what's more sometimes we can be in mad mood because of multiples
> identical reports or "support requests disguised as reports", invalid
> bug reports, ...
> 
> these 3 items may explain why sometimes people can think we don't care
> about them or their reports: this is just because we're not perfect.
> this was not planned.
> 
> anyway we're sorry if you (the communauty) can get annoyed by feeling
> we don't care about you (ie by not get enough feedback from your
> feedback, ideas, ...).

Personally, I expected no explicit feedback at all (having come across this
problem many times before - if developers spent much of their time saying
"done" less would be "done") and the best feedback was to see something I'd
reported fixed. There's considerable good faith involved which was duly
repaid; for example, I felt the problems I had with RC3, USB and solid state
devices were obvious enough and had been reported sufficiently clearly, I
knew that things were happening to the kernel and was sure those problems
would be fixed by kernel changes for 9.0 final: they were!

I understand there's a problem with this implicit contract, though, when
people feel they're being ignored. In quite a number of cases the resolution
was from someone not from Mandrakesoft to take the problem on and solve it
on the list, but I'm struggling to come up with a general solution; my
initial suggestion was for an obvious "tag" (eg REPOST) in an email header
but, if that were done, everyone would use it to try to flag their own
favourite problem and real problems would be drowned out in the shouting.

The problem with bad-quality bug reports is also very difficult to tackle.
Granted, you need people who are expert in the internal workings of Linux,
but you also need people like me who tackle softer issues (eg glitches with
the Gnome UI). Given this diversity there are always going to be
inappropriate reports made; the problem with making bug reporting difficult
is that bugs will be missed. (I would prefer too many reports, with
redundancy and irrelevancy, rather than too few, with incomplete coverage).

Also, as someone who has developed with Unix (Solaris, AIX) extensively in
the past and who is uncovering a brave new world in the handling of
hardware, I believe it would be very useful if the type of output considered
helpful for various types of bug was listed. For example, I wouldn't have
known of lspcidrake -v had I not subscribed to this list.

Alastair

PS The idea that people sign up to looking after certain packages is
excellent.

Reply via email to