Greg Stein <email@example.com> writes:
> Does it need to be do darned large? Your last big commit (2024) blew away
> practically a whole day of Pilato's time to review it. The notification
> stuff is *perfectly* decomposable. Change an API at a time.
> Or maybe change the notification functions to support their new
> functionality in one commit, then start using that in further commits.
> I'm concerned. Your commit rate has been near zero the past few days, which
> is kind of implying a power plant RSN. That scares me :-)
It *is* a power plant :-(, even with my leaving the diff/merge
notification for another day. Freakin' huge.
The trouble is the ugly state the interfaces are in during the
interim. You end up with some things taking *both* trace editors and
notify func/baton pairs. Or in the best case, you end up with two
different styles of interface in the same file -- fine, if temporary,
but the reality is that when it comes time to finish the change, I (or
someone else) will have to mentally recover the tmp expertise I've got
in my head because I'm doing it right now.
But you know... having said all that, I think you're right. I'm going
to go back and decompose it. The full change is absolutely enormous,
not only in terms of affected code, but in time spent making sure that
the new output is *exactly* the same as the old output as far as the
regression test suite is concerned. That's led to all sorts of hair,
such as discovering bugs in some of the output, bugs that the test
suite either didn't catch or was coded to handle.
I mean, enough is enough. It would, in a certain sense, be very
healthy for the code to have this all done now, but it's not at all
healthy for our schedule. I'm just going to do the checkout, update,
and switch interfaces, which is *all* we ever needed for modules, and
leave the rest as a patch attached to issue #662. Let's try not to
let that patch go stale, that's all.
To unsubscribe, e-mail: firstname.lastname@example.org
For additional commands, e-mail: email@example.com
Received on Fri Jun 14 00:06:47 2002