> Date: Mon, 18 Nov 2002 06:41:28 -0500
> From: Ray denenberg <[log in to unmask]>
>
> Yes, you're right. As you say above you're "way out of the loop".
:-) OK.
> > > Sorry, that won't work. has to be
> > > http://www.loc.gov/zing/cql/xcql/
> >
> > Why? The former seems more correct to me. If it's just because you
> > have trouble persuading your web server to honour the aliases, I don't
> > think we should let that be a factor. We should give these babies the
> > _correct_ names, and worry about linkage later.
>
> Honestly, isn't it far too late to be discussing this when we're
> seriously considering announcing this later today, regardless of the
> technical merits, speaking of which, I think the choice between the
> two is relatively insignificant and somewhat arbitrary.
Just more evidence that the release is premature. I agree with you
that this particularly choice is not an important one, but the very
fact that so much uncertainty is floating around speaks volumes.
> > Let's postpone. It's the only intellectually honest thing to do.
>
> I haven't seen that anyone's responded to this suggestion, is there
> anyone else who agrees? I though that the consensus late yesterday
> was to push on. I'll be at the office in a few hours, will put up
> the new schemas, and I'll be there all day. We can wait until late
> aftenoon to decide.
A cost-benefit analysis is called for here. Ray's made the point that
it would be "embarrasssing" to postpone for, say, a week. Yes, I
suppose it would. However, as I've argued before, no-one's even going
to _remember_, let alone care, a few months down the line. But if we
release something that (worst case) is broken, or (more likely) works
with fewer toolkits than we'd hoped, or (very likely) turns out to
have a lot of little contradications and holes in it, due to the
hastiness of release, then people surely WILL remember, and the
credibility of SRW -- maybe the whole ZING initiative -- will be
blown out of the water.
Yes, if we release today, we might get away with it. It's even
possible that we're _likely_ to get away with it. But the
consequences if we don't are much, much, much greater than those if we
do the sensible thing and postpone.
I do think that the imposition of this deadline has been a very good
thing: it's focussed minds wonderfully, and brought about a frenzy of
necessary activity which might otherwise never have happened. With
that said, the result of that activity has simply been too much
instability to move immediately forward. What it's shown us is that
the specs weren't as mature a week ago as we thought. This very
morning, XCQL has changed yet again -- not due to caprice, but because
of a real problem with a real toolkit. If we go public today and
another one of these pops up tomorrow, we're screwed.
Having just caught up on those fifty-odd messages, I've seen calls for
a postponement from Sebastian, Matthew and Ralph as well as my own.
At the very least, that should server as a big warning flag.
OK, I'm done. Your call.
_/|_ _______________________________________________________________
/o ) \/ Mike Taylor <[log in to unmask]> www.miketaylor.org.uk
)_v__/\ "Sad as a lonely little wrinkled balloon ..." -- Paul Simon,
"Crazy Love, volume II"
|