Re: Bug reporting proceedure, was Re: Bug#24066: libc6: rsh segfaults as , a result of new libc 2.0.7r2
>>>>> "JG" == John Goerzen <jgoerzen@complete.org> writes:
JG> The fact is -- if we require research beforehand, there will be
JG> FAR fewer reports. I for one will not submit any (or very few at
JG> least) bug reports if this happens. This will end up hurting
JG> Debian seriously.
I agree. I see (I may be narrow-minded in this respect :) only few
alternatives:
a) A better, if that's a proper word, bug reporting system should be
created. A user should be able to see in seconds if the bug she has
noticed has been reported already. This should be *easy* in addition
to being quick (rude, though *far* from complete, but quick way to do
this would be to create a command foo, which would take a
package as an argument. foo xbase would open lynx/netscape and go to
appropriate url, where all xbase related submitted bugs lie
(i.e it would go to Debian bug tracking system (yes, I like nested
parentheses))). One problem is that it isn't always so easy to spot
the package which caused the problem (eg. slrn hangs, is it slrn, is
it slang, or is it ncurses, or perhaps libc?)
Users won't bother checking bug submissions if it isn't very
easy. Remember that most computer users are very lazy in this respect,
and so am I. However, if a developer would contact me, I'd be more
than willing in helping him in fixing the bug (sometimes I might even
mail a bug-fixing diff to her, who knows :)
b) A user doesn't have to check whether someone else has spotted the
problem too. This method has some advantages, as well as drawbacks
too:
+ Developer(s) can derive something from the number of
bug reports of particular bug. Like whether it happens with
all users, does it behave always the same etc.
+ Developer(s) get probably more detailed view of the bug,
because different people notice different things and explain
it in varying detail / way etc.
- The self-evident drawbacks: bug archives become huge and as
such harder to browse, users waste time telling the same
story one after another, developer(s) get annoyed and start
to -<sob>- ...and users go through that "Hey, I did it! I
discovered what's wrong with.. ..whaddayameanyouknewit?
Awww. Duh". Now, that's tough for a poor user.
c) Let's create only bug-free software.
* Sure. Forgive me, I'm quite tired already. Time to sleep(7200).
--
# Ed GSM: 040 5960810 URL: http://lodge.ton.tut.fi/%7Eed/
*Purr*
-- Mira, the Cat
--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: