2nd monthly community report!
ducasse at iam.unibe.ch
Mon Apr 7 16:13:44 UTC 2003
Excellent this is really good to have such as report! I can forward it
On Monday, April 7, 2003, at 02:31 PM, goran.hultgren at bluefish.se wrote:
> Monthly Squeak community report, 2003-03-11 - 2003-04-07
> NOTE: This is a report based mostly on my personal view on what is
> on, please remember that. I may for example simply be wrong about stuff
> or miss things - it has been known to happen. ;-) I/we haven't decided
> on a format for these reports, this time I am testing a narrative style
> (much inspired by the Debian Weekly News but without the urls - didn't
> have time for those). Content is 99% extracted from the mailinglist.
> NOTE 2: The basic idea behind these reports is to make it easier for us
> all to know what we are doing and where we are heading.
> Welcome to the second monthly report on the Squeak community!
> Göran (me that is) wrote up and posted a draft on the "Squeak
> It was mostly met with constructive responses and even though many
> thought this should be available on www.squeak.org (we can put a link
> there) it will probably end up as a page on the Squeak Swiki with a
> script autoposting it once a month to the list. This way it is easily
> edited just like the bigger FAQ.
> A little thread discussed different Graph implementations (think Graphs
> as in Collections, not graphics) and there was an idea to try to
> implement a common base since several people obviously had code for
> - we will see what happens.
> Jerry Balzano started a thread about how eToys (and perhaps to some
> extent Squeak) is meant to be learned and taught by teachers. Jerry
> managed to put words to the frustration that many feel in this regard
> and the replies from among others Alan Kay did make things fall a bit
> more into place.
> Alexandre Bergel posted about his work on a new module system called
> "Classbox Model". Not too much discussion arose from it, but people are
> probably waiting for it to become usable before taking a closer look.
> does sound interesting. See http://scgwiki.iam.unibe.ch:8080/SCG/559
> Many threads discussed the upcoming layered model for Squeak 3.6 and
> beyond. As all should know by now Squeak 3.6 will be the first release
> that actually is split into packages. Of course, we will start small
> this does mean that official Squeak will not simply consist of a
> monolithic image. Instead Squeak is moving towards a small kernel image
> on top of which several Squeak packages can be installed. All these
> packages will be sorted into three layers - minimal, basic, full.
> The names for these three layers have been under long discussions and
> the three I just used above are just... three of them. :-) I have no
> idea which names we should try to settle on. But we do seem to have
> settled on having these three layers, roughly what they should contain,
> the dependency rules for them and a few other things.
> We also discussed the pipeline model used in the Linux distribution
> Debian and how those concepts could work for Squeak. Nothing decided
> our processes will need to mature more until we can say if it would be
> something to strive for.
> Doug Way announced the release plan for version 3.5 and explained why
> are doing a new release so shortly after version 3.4. This was because
> few BUGs needed correction and we also wanted to try the new harvesting
> model and test the release process. The planned release date was set to
> 4th april.
> The 3.5 plan panned out pretty good into beta and gamma but got hung in
> the final days due to the so called "WeakArray bug" - which eventually
> actually was discovered to be really old. Due to other factors the
> release was given one more week anyway so 3.5 *will* be released on
> friday the 11th april. And that's a promise. :-) It will also be
> accompanied with a rough plan on what version 3.6 will bring which in
> turn is planned to be released 1st of August.
> The oldtimers on the list got a bit sentimentally carried away and
> started trading warstories about ancient mythical Smalltalk hardware.
> :-) It all started with some questions on the basic Squeak history
> answered promptly by Alan Kay and eventually turned into a discovery
> tour on different available processors and hardware today suitable for
> building Smalltalk machines.
> The Guides announced their support for having SCG (the guys in Berne)
> the Steward of the kernel. It was not a big deal but some of us Guides
> thought it would be a good thing to show our support in a more official
> MCP (Morphic Cleanup Project) reported some progress and they are
> digging into Morphic. Assumably a lot of that work will go into 3.6.
> KCP is also starting to produce results and this started a discussion
> about how to handle their contributions through harvesting. Daniel and
> Stéphane have exchanged quite a bit of postings and hopefully they are
> coming to some form of an agreement.
> Harvesting is hard and at the same time as we are trying to establish a
> process for this that can scale and move smoothly with as little stress
> on the harvesters as possible, the KCP team also needs a process that
> doesn't slow them down too much. I think the last word was that we will
> go forward and see how it works.
> Diego Gomez Deck started doing some really good looking visual
> improvements to windows and menues in Morphic. Very nice work!
> a lot of traffic on the list which shows people are indeed interested
> how things look.
> Anthony Hannan announced a new version of Compiler using (among other
> things) SmaCC (Smalltalk Compiler Compiler) internally and then "all
> hell broke loose" :-) due to the fact that we ended up discussing
> licenses. And that subject is a dead sure list-killer! The discussion
> did turn up a whole bunch of interesting facts though and I think we
> will *at the very least* compile all that into the FAQ. I think more is
> to be expected on the subject, the discussion has moved over to the
> SqF-list where it probably belongs. If anything more decisive is
> concluded it will of course be posted on squeak-dev for feedback.
> A thread about some user interface behaviour of the package loader (the
> combination of the filters obviously tend to confuse in some situations
> - and some people aren't even aware of the filters) turned into a
> about user interface design. I assume we will revisit these details
> I get SM1.1 working, since that will lead to changes in the package
> loader anyway.
> And to round this monthly report off - a series of short anouncements
> no particular order (updates to existing packages not included):
> - New SUnit 3.1 released now available on SourceForge
> - A new Swedish :-) free Smalltalk book on Stephane's book site
> - Bryce Kampjes gives us a taste of a new compiler called Exupery
> written in pure Squeak that is meant to generate native machine code
> - A serial terminal implementation by Ned building on top of Ians work
> - Code for locking down an image for end usage by Ned
> ...and finally the new Squeak 3.4 CD by Marcus Denker.
> (Note that I may have missed some cool stuff, in that case - simply
> reply and add!)
> Well, over and out for another month of intense Squeaking!
> regards, Göran & the other Guides
Prof. Dr. Stéphane DUCASSE
"if you knew today was your last day on earth, what would you do
different? ... especially if,
by doing something different, today might not be your last day on
"The best way to predict the future is to invent it..." Alan Kay.
Open Source Smalltalks: http://www.squeak.org,
Free books for Universities at
Free Online Book at
More information about the Squeak-dev