[ On Wednesday, June 18, 2003 at 12:59:11 (+0200), Matthias Scheler wrote: ]
> Subject: Re: What is a CRITICAL bug in send-pr
>
> That's the point. A PR should be marked as "critical" if it is critical
> for the whole project.

Nope, that's very wrong, at least if you're talking about what
submitters should do.

You are confusing the severity and importance of a bug report with the
priorities of the project as a whole.  Such confusion is incredibly
dangerous in a project as large and widely used as a whole multi-user
general-purpose operating system.

The severity and importance of a bug report initially reflect the
priorities of the submitter.  If you believe otherwise then you are
mistaken.

It would be downright stupid to assume that submitters can properly
assign the severity and importance of a bug w.r.t. the project as a
whole.  It's not even a terribly good idea to ask them to try.  It is
far smarter to ask them to assign the severity and importance based on
their own specific needs.

Now ideally there would be a second pair of severity/importance values
that the project could assign to the bug so that reports could be
generated based on both the submitter's priorities and the projects
priorities, either computed together somehow or separately.

However GNATS doesn't work that way, at least not out-of-the-box, so
it's the responsibility of the "responsible" parties to decide both a
general policy w.r.t. managing the priority of reports, as well as of
course to decide how severe and important a bug is w.r.t. the priorities
of the project as a whole vs. the resources available to work on all
problems and goals.

-- 
								Greg A. Woods

+1 416 218-0098;            <g.a.woods@ieee.org>;           <woods@robohack.ca>
Planix, Inc. <woods@planix.com>; VE3TCP; Secrets of the Weird <woods@weird.com>