>
> > Well, that depends if sourceforge has more intelligent bug
> query methods
> > than simple keyword searches. If you can only keyword
> search, the current PR
> > database might be just as good. If sourceforge will allow
> me to search in a
> > more intelligent way, it may be worth the effort.
>
> Can you define "more intelligent"?
>
No, :-) but I can take another shot at what I'm envisioning.
It's the same problem with finding stuff on the Internet. You type a
keyword, and the search engine returns everything that's remotely related,
sorted on how much their sponsoring was.
I know that a lot of research is being put into better and more intelligent
search engines for the Internet. For something like a bugs database, it must
be much easier to do, because the range of subjects is so much smaller. A
kernel in the FreeBSD bugs database is those three-odd megs of code sitting
on /, not a plant's seed of some sort.
You could have a new interface to the PR database. You type your
description, and a search engine coughs up a few (possibly closed) PR's that
are related, presenting them to the user with the question: "does one of
these match your problem?". If yes, you have one less PR to root around for,
and you file his/her e-mail address under "me too". If no, you might even go
so far as to say: "why not?", if your engine is sure they must be.
That's what some people are already doing manually now: "I file this PR.
It's similar to xxx/1000, but not quite, because my foo bars the foobar()."
All I'm suggesting is automating this process and making it standard (and
simple!) practice.
I will admit that my description is vague, but it stems from the deeper
feeling that there must be more than this. :-)
Kees Jan
==============================================
You are only young once,
but you can stay immature all your life
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message
