Josh,
* Josh Berkus (josh@agliodbs.com) wrote:
> In some cases the other solution is "we need to search for a better
> solution". But if you say "the proposed solution is bad" without even
> proposing criteria for a better solution, then you are *de facto* saying
> that the problem isn't important, whether or not you would like ot
> pretend that you're saying something else. If a problem is important,
> then it is worth solving. If it's not worth solving, then it's not
> important.
Or you're simply saying that other things hold priority over this
particular problem. Sure, that makes it *less* important than other
things (if priority is your measure of importance, and it may or may not
be) but it is not the same to say that something is unimportant.
> This is just as true of bugs in our process as it is of bugs in our
> code; if we release without fixing a bug, then we are making a concrete
> statement that the bug is not important.
Or that the priority of the release is *more* important. Things are not
all either red-or-blue here (to use politically correct colors).
> For the past two years, we've
> proceeded without fixing the bugs in our process, which is a material
> statement that most contributors don't feel that the process is buggy.
It's not hard to imagine that developers might feel that bugs, code,
hacking, etc, are of a higher priority than very vaugue and extremely
challenging process 'bugs'.
> If the community doesn't think there's a
> problem, then clearly I'm in error for proposing fixes.
For my 2c, I agree that there's a problem, but I've got a ton of other
tasks, not to mention $dayjob that makes it unlikely that I'll find time
to come up with alternate proposals. I will also say that I feel that
this process is still an *improvement* over the previous process, which
is really the only other one we've actually tested. Perhaps that means
we should just try different things out and see if we can't build the
best process out there through supervised learning.
Thanks,
Stephen
