On Tue, 1 May 2001, Eric W. Sink wrote:
>
> There's a very simple, obvious characteristic which distinguishes
> immature dev teams from strong ones:
>
> Strong product teams know how to finish a release
> that does not satisfy all their hopes and expectations.
>
> The software industry is filled with lame teams that don't know
> how to finish anything. Many companies, including mine, make a
> great deal of money charging obscene hourly rates to clean up
> after bozos who ran their projects into the ground. Most of the
> time, the project team sees the ship date on the horizon and
> starts to get scared:
>
> "What if people don't like version 1.0?"
>
> "We can't ship without YYY."
>
> "Without feature XXX, everyone will say that 1.0 is crap."
>
> Sorry folks, those arguments don't get you anywhere. This line
> of reasoning will simply prevent you from ever shipping anything
> at all. In contrast, the truth is very simple:
>
> Every 1.0 release has several goals. One of those goals
> is to gather a bunch of users who will be unhappy with
> your product.
>
Every user of Linux distros should observe the truth of these words.
RedHat ships what it believes is the current best collection of programs
every 6 months. There is always a new feature that gets missed in every
cut.
Every release is accompanied by moans that the product does not have X or
Y and that some user won't be use until it does. Never-the-less there are
lots of places where RedHat is used. It's probabally on 10's of millions
of computers by now. It's clearly a successful product.
I'm in favour of shipping a solid 1.0 without tables or equations or
background images.
I do regard 1.0 as special. Every release we've made so far has bugs in it
that I know could be fixed in a reasonable amount of time. I want to ship
a release where that is not true relatively soon.
Look at what Mandrake says about Abi 0.7.14 " * * AbiWord - is a lean and
fast full-featured word processor perfect for the most common office
tasks."
So lets do the right thing by our users, give them a close to bug free
experience with just a few more features.
Doing tables will probabally mean breaking abi for a while. I know Mike
Nordell wants to refactor fv_View and PD_Document. Jesper is itching to do
major surgury on the backend. If we ship a nice 1.0 we all get to have
more fun hacking :-)
Cheers
Martin
