Fedora Core 4 test3 freeze warning

[Arjan van de Ven]

> 
> > Ignoring the bug reports doesn't motivate testers very much...
> No one sane is going to suggest that there is enough skilled manpower
> right now to follow up on each individual bugreport AND get things
> actually fixed. But I don't think its fair to say that people are
> actively ignoring bug reports. Let's talk numbers.
> I get a total of 1602 bug against fc3t1-3
> i get a total of 320 assigned/new/reopened against fc3t1-3
> 
> Those numbers don't tell me that developers are trying to ignore
> anyone, those numbers tell me the system is overwhelmed.


as someone who has been on the receiving end of bugs until a few months
ago I can say that we don't ignore bugs. *HOWEVER* bugs do have various
priorities. And that's not always the priority the reporter thinks it
is. "OH MY GOD THEY BROKE MY WEIRDASS ISA SOUND CARD" may well be very
important for you (you don't get to enjoy your mp3^Wogg collection) but
it's a defect that would hit maybe 0.00005% of the users, and as such
for the other end of the bugreport isn't too important compared to the
other bugs. I can imagine you saying "then close it wontfix then I know
where I am". Well. Funny you say that ;-). If I do that as developer to
100 bugs, the result will 40 hate mails describing the assumed
profession of my mother, 30 reopens with "but this IS important and you
are a jerk to think it's not" and a lot of hassle. After some time, I
can tell you, you just don't do that anymore and leave the bugs dangling
instead.

Another angle is the amount of information in the bug. Bugs with very
little useful aren't exactly motivating to work on, simply because
there's no way to even begin to start (basic info like version numbers
and hardware info etc missing). Asking for that info goes unanswered
like half the time at least.

Another thing (at least for the kernel) is that there is a lot of broken
hardware out there. Some by design, some really defective. But hey the
kernel oopses so ...  Or some binary module is used that happens to
corrupt random data.  I'd estimate that 20% to 30% of the incoming bugs
is in this class; absolutely unreproducable and by nature unfixable. Yet
if you close it people don't like to hear their hardware sucks, and you
get "If I sacrifice 3 chickens, hold my finger exactly THERE then
mandriva works, but fedora doesn't" and a portion of the hatemail I
described before already. Leaving these bugs open is useful however,
because if you see the same bug pattern (same oops) happen more than a
few times then maybe the combined bugs are sufficient to have a lead. 


-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : http://lists.fedoraproject.org/pipermail/devel/attachments/20050421/6fe6d2b1/attachment-0002.bin