Re: bug rating

[Graham Percival]

On Thu, Dec 10, 2009 at 08:22:32AM +0100, Werner LEMBERG wrote:
> is it correct that all fixes, regardless of its annoyance, get a `low
> priority' in case it won't become part of the next `milestone'
> release?

That's not quite correct.  There's no functional difference
between Postponed, Low, and Medium.  In the case of your recent
two issues, I more-or-less assigned them randomly between Medium
and Low.  If you'd entered them yourself as both Medium, or both
Low, I wouldn't have said anything.

We essentially only have two levels: stop-stable-release, and
not-stop-stable-release.

> I consider this categorization a bit coarse,

Yes and no.  I'd rather move to having 4 levels:
- High: should be fixed ASAP (crashes, regression failures)
- Medium: will stop a stable release.  (I'd consider maybe 10% of
  the current "medium" items to be medium on this scale)
- Low: the normal priority.  Sorry, but we just don't have many
  bug fixers!  I favor honesty over trying to make users happy
  about assigning their pet issue a "higher priority" flag that
  nobody pays attention to.
- Postponed: there's every indication that nobody will touch this
  for the next year or more.

but so far I've been waiting for feedback about my Regression-bugs
email.

> and I would like to see at least one more level to mark bugs as
> `annoying' or something like that.

I disagree here.  Every bug is "annoying" to the person who
reported it.  This would only lead to arguments over what's more
annoying.  I'm sure that somebody considers our lack of a handheld
media CSS for the new website to be horrible!

Do bug fixers look at the priority levels?  Shortly after I posted
the links to the grid view, IIRC two people started working on
regressions, so maybe this is starting to happen.  But I doubt
there's any difference in how Frogs consider items between medium
and low priority.  Their main interest is "how hard will it be to
fix?", not "does somebody find this annoying".

Don't misunderstand me: I think it would be great to rank bugs
based on severity, annoyance, or "how ugly does it look".  But
that won't happen until we double or triple the bug fixers -- both
doubling their numbers, but most importantly doubling their
knowledge of lilypond architecture.


Let me turn this around: you are one of our top 10 bug hunters.
If you had no previous connection to any of the issues, how would
you decide which bug(s) to work on?  Would you seriously just
start working on whichever item *I* said was most important / most
annoying ?  or would you try to find an item that appealed to
*you* personally?

Cheers,
- Graham