Re: Behavior Trends in Nautilus and other Desktop Apps

[Ben Liblit <liblit cs wisc edu>]

Karsten Br�elmann wrote:
> Oh, I didn't expect you guys to actually have subscribed to the
> [gnome-bugsquad] list, judging from the first post. I did, however,
> hope it. :)

When Jason and I looked at the <gnome-bugsquad> list archives, it seemed 
that discussion there mostly concerns Bugzilla administration and policy 
questions.  That's why we didn't send a copy of Jason's original message 
to that list.  If you think there are a fair number of people on that 
list who would be interested in what we're doing, though, we can 
certainly join the list and try to get some discussion going.  Do you 
think we should do that?

> Anyway, I did NOT say what you quote out of context above.

Ah, you are right.  I never intended to put words in your mouth, and I 
sincerely apologize for misrepresenting your comments.

You are also right when you point out that bugginess is not the same as 
crashiness.  One bug can cause many failures, and one bug in a library 
(even a GTK+ theme) that Nautilus uses can cause Nautilus to fail 
through no fault of its own.  Our findings affirm that Nautilus 2.14 on 
Fedora Core 5 crashed often; they do not necessarily suggest that it had 
more bugs.

I'll let Jason reply to some of the more technical points in your 
message, but I do want to emphasize that (1) we are not just looking at 
stacks, and (2) we do understand and exploit the distinction between 
bugs and failing runs.  CBI is very much about finding the root causes 
of the most common failures, with an explicit bias toward finding the 
few bugs that cause the most problems for the most people.

Thanks for your feedback, and again I am sorry for misquoting you earlier.

-- Dr. Ben, the CBI guy