Two important issues... - on Issue 2

goran.hultgren at bluefish.se goran.hultgren at bluefish.se
Fri Feb 14 14:24:24 UTC 2003


<diegogomezdeck at consultar.com> wrote:
> Hi Göran and list,
> 
> > Issue 2
> > ========
[BIG SNIP of the issue]
> I feel we'll be not able to make the next step.

I hope you are wrong. :-)

> Don't misunderstand me. I *really* want to see (and to be part on) the next
> step! but "We have too much ballast in our balloon".
> 
> Every time somebody suggest a change we hear a lot of voices saying "Don't
> please! Compatibility!".
> 
> Traits is a good example. I'm not sure that Traits is the way to go, but
> I'm sure we have to try! But this list is more motivated in un-important
> discussions (like native widgets, access to relational databases,
> deployment as executables, etc).

Well, un-important to some - important to others. Personally I feel both
the "planes" are important (to me). I would like to have access to
native widgets AND the cool new Morphic or whatever paradigms. ;-)

I want to have Traits AND somehow be able to ignore them if I want to.
In the end Traits may turn out so popular though that I can't ignore it
- but at that time I probably don't want to either...

> The are a lot of other things to try (examples: remove of direct access of
> variables, more metaprogramming facilities, etc, etc, etc).

True. But Traits seems to be one of the current crop of "experimental
extensions to Smalltalk" that most seamlessly could be accepted. We have
AspectS (boy, would I like to play with all these things sometime),
PerspectiveS etc too and... well, I don't know.

I just want to think that we can come up with a "good solution" on how
to be able to accommodate these "extensions" and still not "tear down
the house".

> Solution? I haven't one. I'm only describing my feeling with Squeak since
> SqC had leaved Disney.

I think Squeak is moving in the right direction. We don't have the
strong hand of SqC anymore, but the community has stepped up to the
plate and I sure think it looks like good sofar.

And this new "distributed" community driven Squeak (more like a *real*
open source project I think) may turn out to scale even better than the
tight team at SqC in the end. Who knows? And perhaps we can do a better
job at documentation, unit tests etc than them because we aren't rushing
against the clock like they did... ;-)

regards, Göran



More information about the Squeak-dev mailing list