On Thu, 15 Nov 2001 09:27:29 -0800 Richard Maine <[log in to unmask]>
wrote:
> Malcolm Cohen writes:
> > Anyway, we're all in violent agreement that F77-F90 was too big a gap (13
> > years), but we're only looking at a 7-year gap here (1997-2004) for F2002.
> > Regrettable though the slipping of the schedule is, it's not (yet!) a
> > complete disaster. IMO.
>
> But it will be if we throw everything out and start over. :-(
>
> There does come a time when you slow things down by pulling features
> out. We are long past that time. The standard is "complete" from a
> feature standpoint. We are in the integration phase of polishing and
> making sure it all fits together consistently. Pulling substantial
> features out now would cause a *MAJOR* slip. Little things are a
> different matter (enums, for one example, which really don't have much
> in the way of global interactions as they are done now), but to pull
> out, say OOP, my personal estimate would be a 2 year slip to get it
> out of the document, restart the integration, and get back to where we
> are now. That, of course, presumes that there was immediate unanimous
> agreement at the start of the next meeting that this was the thing to
> do - more realistically, it would take several years to agree on the
> redirection, making the total perhaps 5 years of slip. I'd call that
> a complete disaster.
>
> It is all too possible to get caught in a Dilbert-esque trap of
> spending years arguing about why things are going so slowly and
> how to speed them up. And how we could proceeed with much more
> speed and less contention if everyone else would just agree to do
> it my way. Neither J3 members or posters to this list (there being
> some overlap there), are immune to this tendency. (And I must
> include myself).
>
> If one thinks that having feature X (for whatever X) in the standard
> is actively a bad idea, that's one matter. Depending on the value of
> X, I might even agree. But don't think that pulling features out at
> this late date is a way to speed anything up. When you get this
> close to "shipping", then any change means delay, deletions not being
> necessarily any beter than additions.
>
> And yes, I'm afraid that in terms of standards processes, "we" (J3)
> are very close to "shipping." About two more meetings if I recall
> correctly (and I could be off a little, but not much - didn't bother
> to check the exact dates), before the first review draft is supposed
> to ship. The review and approval process just takes that long.
>
> (Malcolm, of course, knows all this. I'm not addressing this at him.)
I could not agree more with Richard. Now is not the time to start pulling
features in a counter productive attempt to speed things up.
IMO one of the major reasons why F8X did not appear until 1991 and then very
greatly changed was the disaster of the grand compromise at Scranton when the
committee paniced over a NO vote on its first formal ballot and chucked out
vast amounts of material, some of it critical. We then took three years
editing the revised document to obtain a correct removal and a further three
years re building a half way workable language. In my opinion if we had spent
two years correcting the flaws in the 85 draft and intergrating the then
included functionality properly we would have had F88 and it would have been
a better language. Ah Well! Hindsight and wouldn't it have been great if we
hadn't!!!!
>
> --
> Richard Maine | Good judgment comes from experience;
> [log in to unmask] | experience comes from bad judgment.
> | -- Mark Twain
--
Lawrie Schonfelder
Honorary Senior Fellow, University of Liverpool
Home: 1 Marine Park, West Kirby, Wirral, UK CH48 5HN
Phone: +44(151)625 6986
|