[RANT] Come on people! ;)
Michael Latta
lattam at mac.com
Tue Dec 21 17:31:43 UTC 2004
Stef,
Without a strong leadership Squeak is going to drift as various
interests get more or less effort applied to them. I helped start a
software company in 1986 built on VW. It soon because too much effort
to keep the product up on the latest image from ParcPlace. So we
baselined on an image and did work that benefited our customers. We
had a stable base because it was not changing. We could selectively
move pieces from the newer images down to our image. At intervals we
would migrate our code base to the new image to get their bug fixes and
features. This seems to be the way of applications in Smalltalk. You
can not go long telling your customers "we spent the last 6 months
moving to a new image so here is the new release with little in the way
of new features". The very productivity in Smalltalk that helps us
all, can cause issues for a product trying to get work done. That is
why I hope the community heads in the direction of using a minimal core
and layering on that packages. This makes the task of tracking the
pieces you do need much easier, and the effort of moving forward with
new versions of the core and select packages easier.
Michael
On Dec 21, 2004, at 2:48 AM, Stéphane Rollandin wrote:
> Doug Way wrote:
>
>> Anyway, um, I think Goran's main point was along the lines that you
>> can't simply try to convince other Squeakers (or the Guides) to start
>> working primarily on gearing Squeak toward the biggest group
>> possible.
>
>
> but this was not my point. my concern is: what are the stable parts of
> Squeak which I can securely build on ?
>
> examples of recent questions arising about parts of Squeak I'm using
> (and relying) on:
> BookMorphs - someone wanted to deprecate them for project (?)
> Morphic - will it be replace by Tweak ?
> Connectors - v1.9 don't load easily anymore and v2 is a complete
> rewrite which does not seem stable yet ?
> Regexp - will the plugin version be maintained ?
>
> etc. (there are others !).
>
> I'm developing a rather ambitious open source application for musical
> composition and I would like it to live a long life. it's very nice to
> see all this dynamic of experimentation taking place, but at the same
> time some strong and solid ground should be maintained to allow for
> the construction and long term maintenance of reliable applications on
> top of Squeak.
>
> Squeak is a wonderful experimental field, but it would be much better
> as a multimedia development framework if the experimentations and new
> features did respect some generally agreed "shape" which I believe is
> still to be precisely defined and accepted.
>
>
> hoping that this makes sense,
>
> Stef
>
More information about the Squeak-dev
mailing list
|