Houston, we have a problem.

Andreas Pour pour at mieterra.com
Fri Sep 6 02:31:05 BST 2002

Marc Mutz wrote:
> On Friday 06 September 2002 01:08, Christian Parpart wrote:
> <snip>
> > > > > Proposal: Use a more reliable keyserver.
> > > >
> > > > Do you really think that telling this to a user is an acceptable
> > > > answer?
> > >
> > > As a short time solution: Yes, absolutely.
> >
> > You forgot that this may be also caused by having a slow bandwidth
> > connection to the Internet. Ingo and Marc, I understand your problem,
> > but don't be so indefinitely stubborn in your descission.
> <snip>
> I don't see stubbornness in the mail you just replied to (see below).


The answer "use a more reliable keyserver" to the problem "KMail freezes
forever" seems, based on my limited understanding, quite absurd, for at
least the following reasons (some of which may not actually cause a
problem, but as I mentioned not everyone understands the internals as
well as you):

  * the keyserver may not be the problem - instead, any one of the
following problems could occur:
    o Internet connection goes down
    o Internet proxy goes down
    o Internet has slow response
    o DNS server has slow response or fails to respond
  * the keyserver may have a problem one day unexpectedly
  * the problem may be in the route to the keyserver, with lost packets
or so
  * how do you intend to communicate this "solution" to users?  Since
KMail is frozen, I guess the user can't access Help, even if it (as
opposed to wasting time at bugs.kde.org) were a likely response to a
frozen UI.
> Don't you think it would be way more prudent to make a patch fixing the
> problem or - failing that - at least to not write such senseless me-too
> mails?

I think this "make a patch" response gets out of hand.  It is a fine
response to a feature request, but not to a bug.  Or do you think a
maintainer of a KDE app should always be able to say, "I won't fix a bug
unless you send a patch"?  Where does that lead us?

One very practical reason not to use this approach is that it would take
me, e.g., 100 times as long to fix a bug as it would take you, at the
very least, and I'm sure it would be the same if you tried to fix a bug
in some code I wrote.  In any complicated code, an "outsider" fixing a
bug is quite likely to introduce new bugs, since getting up to speed on
how everything relates together is very time consuming and difficult. 
That is why apps have maintainers and people do not just randomly work
on any project.
> > > I know all this. And I would very much like to fix all those
> > > problems. But unfortunately I only have so much time to work on the
> > > code. Maybe I should spend less time closing bogus bug reports and
> > > duplicate wishes and supporting KMail users. NOT!

It does seem a rather strong case has been made for why Don's patch is
not appropriate at this juncture.  Any particular reason Don has not
presented his point of view?



More information about the kde-core-devel mailing list