On Jul 6, 2009, at 9:33 PM, Hernan Wilkinson wrote:

> Hi Stef,
>  have you thought about have not only first class instance variables  
> but first class variables? (no matter the scope).
yes this was included. Now we want to really experiment and provide a  
real and working solution.
I just mentioned that because Smalltalk deserves a better  
instantiation. :)

>  I think that would be interesting... observers on variables would  
> be pretty easy to implement, type inference (or type  
> recollection)... not sure about the trade-off cost/convenience,

We did a simple implementation with marcus and the implementation we  
came up with did not have **any** runtime penalties.
Now the classBuilder is so brittle that to make sure that we would  
have a robust implementation we would have to rewrite it and this is
a real nightmare... look at the comment.

for now we just created an extra instance variable that serves to  
allocate our instances :)

> but looks interesting... accessing local variables could be made  
> sending messages to thisContext... anyway, just an idea.

so far having plain well working first class instance/class variable  
would be a really good achiemenet
with a good mop. Would be great. Imagine people could implement tweak  
like field without requiring copying the compiler and adding XML  
syntax :)


>
> On Mon, Jul 6, 2009 at 6:31 AM, Stéphane Ducasse <[email protected] 
> > wrote:
> Hi nevin
>
> our goal is not predominancy :) It was never. I would be happy that
> squeak offers a really good multimedia
> platform. Our goal is to deliver good open source smalltalk.
> Our goal is to get agile, rethink the system, use good software
> engineering practices (systematic smallLnt rules),
> tests....
> We want to propose/accept/evaluate new solutions:
>        first class package,
>        first class instance variables, better refactoring support,
>        may be modules, security, new compiler....
>
> Smalltalk deserves a clean and powerful implementation and
> we will work on that hard.
>
> Stef
>
> PS: you can forward this answer to squeak-dev if you feel the need.
>
> On Jul 6, 2009, at 3:17 AM, Nevin Pratt wrote:
>
> > There's no right or wrong answer to the following three questions.
> > And,
> > every response is meaningful, because it helps reveal the general
> > sense
> > of direction that the community is feeling.
> >
> > ***********************
> > Question #1: Of the various Squeak forks (such as Croquet, Pharo,
> > etc.),
> > what is your feeling of their relative popularities at this point in
> > time, and how does that compare with the current popularity of the
> > base
> > Squeak distribution?
> >
> > Question #2: This is related to the above-- what is your feeling of
> > the
> > relative popularity *trend* of each of the forks, and how does that
> > compare with your feeling of the *trend* of the future popularity of
> > the
> > base Squeak distribution?
> >
> > Question #3: What is your justification for your answers to the  
> above
> > two questions?
> > ************************
> >
> > I'll begin with my own answers...
> >
> > ...I'm beginning to feel like the Pharo fork currently now has maybe
> > 50%
> > or more of the popularity of the base Squeak distro, and I'm  
> beginning
> > to feel like Pharo is headed to become the dominant distribution in
> > the
> > future, surpassing the original Squeak distro.  And I base that
> > feeling
> > on several observations:
> >
> > 1. The email traffic comparison on the "Pharo-project" email list
> > compared to the email traffic on "The general-purpose Squeak
> > developers
> > list", and the observed trend in such traffic.
> >
> > 2. The number of developers posting on the Pharo-project email list.
> >
> > 3. The specific developers posting on the Pharo-project email list
> > (known Squeakers who have contributed heavily to Squeak in the  
> past).
> >
> > 4. The fact that Seaside development has moved to Pharo.
> >
> > At one time I wondered if Croquet would become the dominant Squeak
> > fork,
> > for the simple reason that it had a business organization  
> "officially"
> > supporting it, similar to what Squeak had in the Disney days.  But
> > now I
> > think I would bet my money on Pharo.
> >
> > What do you think?  And why?
> >
> > Also, remember, there's certainly nothing wrong with several
> > "dominant"
> > forks existing.  But of those, one of them will still be the most
> > popular overall.  And I'm curious about thoughts of which one that  
> is,
> > and which one that will become, and why you think so.
> >
> > Nevin
> >
> >
> >
> >
> >
> >
> > _______________________________________________
> > Pharo-project mailing list
> > [email protected]
> > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>
>
> _______________________________________________
> Pharo-project mailing list
> [email protected]
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>
> _______________________________________________
> Pharo-project mailing list
> [email protected]
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project


_______________________________________________
Pharo-project mailing list
[email protected]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Reply via email to