> On the other hand,
> if lots of other beta-testers with lots of different hardware
> configurations come across it, it may be bad enough to warrant
> delaying the release to fix it, in order not to inconvenience a
> large number of users.

Do we realy have this notion of showstopper bugs? If, how are they collected
right now and why isn't the list publicaly available?

We could do a vote to select open bugs for this list ;-) (no, not realy
kidding) - but that would need a slightly reduced count of open bugs in the
first place. I think most developers have a good subjective estimate for the
impact of bugs once they are sufficiently analyzed/reproducible.

Anyway, we should have such a list and we should have objective criterias
(i.e. "any bug introduced after being in alpha phase must be fixed for the
release or the change will be backed out").

In real life software engineering we do have such a list. My boss makes it
up and besides the objective criterias he takes votes from paying clients.
So every time he drops the phone, the list will be shuffled.


Martin