On 05/14/2010 04:23 AM, AGK wrote:
Thanks for the update William. It seems like we're
getting very close
to release now, which is great to hear.
We're very excited, too.
On 14 May 2010 06:25, William
Pietri<william(a)scissor.com> wrote:
we will be getting
together with Rob H. and the rest of the ops ninjas to discuss release
dates.
Are you referring to the developers? Suggestion: just set a firm date
and have the dev's do it by then. My understanding is that we have
some on payroll, so ultimately there's little need for a discussion if
the only benefit is more time-wasting.
In this case, we're just looking to get on the ops calendar. Picking an
arbitrary date is reasonable when you have flexible scope, but for as
long as I've been involved this project has always been looking to
release with the absolute minimum necessary for starting the trial.
With that kind of product plan, picking an arbitrary date means you have
a high risk of either releasing without the minimum feature set or
taking a hit in code quality, which has to be cleaned up later. Either
of those wastes time, so we've waited until we were pretty close to done
to start formally talking with ops about a release date. We have been
casually talking with them, though, so they are aware of what's coming,
and I expect things will come together nicely.
One consequence of launching with a minimum-necessary feature set is
that after launch, a bunch of people will say, "Why didn't you do
[obvious feature X]?" Some will say that nicely, and some will say that
as if we were fools for missing such an obvious improvement. At that
time, I'd love it if everybody who has been pushing (quite rightly) for
the soonest release possible will help us explain that additional
features would have meant later release dates.
William
