On Fri, 26 Oct 2001 10:49:00 -0500 Michael Kaply
<[EMAIL PROTECTED]> wrote:
>
> As far as the hangs, crashes, etc. go, we very seldom see reproduction
> scenarios for these so we can't reproduce them.
>
> Bugs like "Warpzilla hangs" and "warpzilla crashes" are not very useful.
>
> There are tons of these that are machine specific, and with OS/2 having
> no useful tools for doing stack traces when you crash and things like
> that, we don't have much to look at.
When my system creates a popuplog.os2 entry, it also creates a
a pdump.000 (or suchlike) file. The PMDF facility can be used
to extract *some* information (maybe even a stack dump) from
this file. (I have 'DUMPPROCESS=E' in my config.sys.)
It would help users if the Mozilla developers would take time out
to create a short "cookbook" list, telling users what information
should be gathered (and telling *how* to do that) from a "crash",
so that a report of that "crash" is made as valuable as possible.
mikus
p.s. I really dislike formal reporting mechanisms like Bugzilla.
In my opinion, they rerely result in a "fix" to a problem.
An example: In earlier days of OS/2, I found an instance
of a clearly SOFTWARE bug. I submitted a problem report.
IBM closed that report, with the notation: "This problem
was reported on an EISA system. We do not have any EISA
system on which to reproduce it." Another example: From
the first beta of Netscape 4.xx on, my browser window was
YELLOW. I submitted a problem report, even identifying the
incorrect color_scheme_label from which the value was being
taken. Result -- I'm still living with a yellow window.
