Hi all!

First - Hilaire - this is an interesting post indeed. I will now perhaps
sound like I am "correcting" you, but that is not my intention. I will
though reply with my view of the issues you mention - but your view is
of course very valid - because it is the view of a newcomer.

Hilaire Fernandes <hilaire at ext.cri74.org> wrote:
> Hello,
> 
> Regarding the recent loong thread related to the future of Squeak and 
> suggested road map, I would like to give some points of view as a newbie 
> to Squeak and Smallktalk in general.

Thanks for doing this. It does teach us a lot.

> An outsider may have complete different point of view worth reading.

Indeed.
 
> Mark Guzdial was wondering what may be the expectation of newcomers to 
> Squeak. As a new comer, I can jump there.
> 
> Probably the expectation of new comers just depends on how is presented 
> Squeak to them. To me it has alway been presented as a multimédia tools 
> with advanced programming concepts. So I just expect that Squeak has 
> good and stable features in that field. If I discover that such features 
> just disapear or too easily become obselete I will quit and spend my 
> time to learn some  other environment. I have to admit, that I was 
> really worried when I read that a super-mega-inovative stuff named T**k 
> will soon replace Morph. I was just spending time learning it...

Morphic will probably be around for a loooong time - like Cobol. :) It
is cool and neat in lots of ways, but it is currently also a bit messy
(to say the least). No need to worry - instead we should rejoice that
there is new stuff in the works. :)

And since the primary developers behind Tweak are old time Morphic
people it should probably retain the best values of Morphic. Btw,
regardling multiple mailinglists - Tweak has one of its own. :)

> As a newbie appearing to be a math teacher, I was interested by the 
> quote of the www.squeak.org: 'If you are a student, parent, or teacher, 
> please jump over to our newly redesigned SqueakLand.org website, and 
> download some great educational projects.' However as a French speaking 
>   person I was surprise to see that I18n was poorly or/and not clearly 
> in place in Squeak. I later discovered that people were working on that, 
> but not on maintstream, but aside and in an isolated way. Eventually 
> with the help of others I volunteer to help in localisation in French 
> but with the strange feeling that it was just hugly hacks (comparing to 
> the way i18n and l10n is professionnaly conducted in other project of 
> the free software community). So to sum up, as a newbie I fell that 
> 'Squeak in education'  does not make big-enought sense as long as I18n 
> is not a top concern in Squeak, at least in non-english speaking countries.

This doesn't sound like my impression of i18n in Squeak. Sure, it is not
yet in the official release (being 3.7) but I think it looks good in the
upcoming releases. Squeak is being heavily used in non-english talking
countries (especially Spain and Japan I think) so I think this will
improve steadily. And French also has good representation in the
community.

[SNIP]
> There are some other point of vew or feedback I would like to share, but 
> at a meta-development level or the organisational level of the Squeak 
> development.
> 
> - When a new comer hit the Squeak reference page of Squeak it gets to 
> http://www.squeak.org. Do we need to comment about its quality?
> After a few click it will get to http://minnow.cc.gatech.edu/squeak it 
> looks a bit better, much better, not because the information is well 
> organised (it is a nightmare) but only because there is a lot of 
> valuable information to find.
> 
> So to say the true, the first contact with Squeak through the web is not 
> very good. Until now, the only professional looking web site I found was 
> in German http://www.squeak.de, the one from Small-land is also quite 
> nice, but the German one is much better structured. (hints: 'Provide 
> guidance in the User Interface, in particular newbie need that')

Yes, a known fact. Want to help with our web presence? :)

> - It looks like Squeak is an orphaned project. Where are the leaders? 
> Where are the founding fathers? Beside demonstrating how the UI was 
> invented by themseleves can they help the Squeak community to be organised?

It is not orphaned. The original team is around in satellite projects -
at least a few of them.
Some are involved in the Croquet project (written in Squeak).

The current "leadership" is looser than the first few years. We now have
6 "Guides" that try to guide the decisions being made in the community,
me being one of them. This approach was selected in order to keep the
community intact. In retrospect it has worked "somewhat".

Now there is an ongoing debate that might turn into a new kind of
leadership - we will see.

But the project is *very much* alive, no doubt about that. It is just
more loosely organized at the moment.

> - Only one devel. list covers the different *core* aspect of Squeak. 
> Squeak is like a whole system to itself covering VM, compiler, library, 
> UI, multimedia. I cannot image that the 
> linux-kernel+GCC+libc+Xwindow+Gnome could have been developped with only 
> one developper mailing list.
> At some point it really shows that Squeak developpement is poorly 
> organised in the community.

I disagree with this. One of the beauties of Squeak (and also one of the
primary design goals of Smalltalk/Squeak) is that it should be
comprehensible by a single person. The fact that the list talks about a
wide variety of Squeak components is a strength IMHO. Also, there are
other lists - one for the VM, one for Croquet, one for Tweak, one for
Squeakland, one for Spoon/Squat etc etc.

> - why there isn't sub development groups (with clearly identified 
> leaders) covering some of the core points of Squeak:  VM, class-core, 
> i18n, GUI, multimedia, E-Toy, code cleaning, documentation, etc.

There are. Clearly defined may be a different story, but some of these
groups exist.

> Each sub-groups will work on its own space area (mailing list, etc) 
> following a roadmap defined in common, throught a Squeak foundation. An 
>   additional group, the "release team", would help in coordinating 
> transveral communication and in sighting the development group to follow 
> the release schedule.
> All these points are not new and not from me. For example the Gnome 
> community is following something like that. Look at 
> http://developer.gnome.org/dotplan/ and the link at the right

Sure, this would be one way to organize. Just don't think it is easy. I
am also not sure if it is the best route, even though managing Squeak in
"modular parts" is one of the primary objectives in a lot of the work we
do.

> Oh, a few last words, again related to technical aspect of Squeak and 
> newbie.
> A simple mandatory rule to follow when adding new classes to the core of 
> Squeak will be that only well documented class can get in. I can assure 
> you that newbie will thanks you.

You wouldn't believe how many times I have screamed on this list about
that to the point of pissing people off. :) And you would be surprised
how many actually think it is better to get new undocumented code that
*works* into the image than putting it on hold because it isn't
documented. I am not of that view, but that is a different topic.

It just serves as an example of the fact that organizing/leading this
community is hard, in many different ways.

> Hilaire

regards, Göran