January 2010 - Pharo-dev - lists.pharo.org

Re: [Pharo-project] Getting feedback on an headless VM

by Simon Denier

On 18 janv. 2010, at 18:17, Lukas Renggli wrote: >> Problem is the process starts but obviously got stuck after that. I got no feedback on what's going wrong (I guess some error happens and the system is waiting for user input). > > Presumably it cannot find the sources or changes file, or their > permissions are wrong. This was the right track, the sources file was not in the right place and it could not find it. Thanks Lukas! Now about the Linux, I'm sorry I cant help much, I finally manage to run the 3.10.3 and it appears it is really ok with closure. > > More details you can find in the Seaside book. Not that this is not > particularly related to Seaside, but it contains all essentials to run > a VM headless: > > http://book.seaside.st/book/advanced/deployment/deployment-apache/run-vm > > Lukas > > > > >> >> I tried with a fresh 10500 image and the simplest script I could think of to test: >> (FileStream forceNewFileNamed: 'foo.txt') nextPutAll: 'bar'; close >> >> ....same thing. >> >> Of course it works on my local machine with UI. >> >> Is there a way to debug this, like having the debugger to dump the stack in the terminal. I tried Cmd-. but it just stops the process (like Ctrl-c) with no feedback. >> >> >> -- >> Simon >> >> >> >> >> _______________________________________________ >> Pharo-project mailing list >> Pharo-project(a)lists.gforge.inria.fr >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> > > > > -- > Lukas Renggli > http://www.lukas-renggli.ch > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project -- Simon

Jan. 20, 2010

Re: [Pharo-project] question about O2 vs OB

by Fernando olivero

Ok, thanks for the answer Lukas. Then i would ask the O2 developers, why is the fork needed? Just curious, because if 02 copies a lot from OB, to add more functionality why that extra features cant be part of OB? My point is that i don't understand if OB is no extensible enough for building O2, or just because it was easier to fork. Many thanks, Fernando On Jan 20, 2010, at 2:14 PM, Lukas Renggli wrote: > O2 is a fork of OB. > > The duplication of the complete code-base is necessary, because O2 > makes incompatible changes to the OmniBrowser framework. The way O2 is > implemented makes it impossible to run both together on the same core > model without conflicts. > > Lukas > > -- > Lukas Renggli > http://www.lukas-renggli.ch > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Jan. 20, 2010

Re: [Pharo-project] question about O2 vs OB

by Lukas Renggli

> so is Lukas' statement still true and both are in conflict or are > these conflicts now resolved and it is possible to run both together? That was the answer to the question "why they don't use the same codebase". Don't know if this was the actual question? According to the announcement they should be able to run next to each other now. I haven't tried it though. Lukas -- Lukas Renggli http://www.lukas-renggli.ch

Jan. 20, 2010

[Pharo-project] [update 1.1] #11173

by StÃ©phane Ducasse

11173 ----- - Polymorph Fixes for control button appearance. Stef

Jan. 20, 2010

Re: [Pharo-project] O2: tester wanted!

by Adrian Lienhard

Thanks, Alex I gave it a quick try. Installation into PharoCore 10505 worked and briefly clicking through the browser I haven't found any obvious problems. However, there are 6 undeclared classes referenced in the code. Very likely when this code is run, O2 will fall over. Also 2 tests have errors and one fails. HTH, Adrian On Jan 19, 2010, at 23:11 , Mariano Martinez Peck wrote: > On Tue, Jan 19, 2010 at 11:03 PM, Alexandre Bergel < > alexandre.bergel(a)inria.fr> wrote: > >> Dear All, >> >> For people who do not know, O2 is the browser David has implemented in >> the last few years. It replaces class-categories on the left-hand side >> pane with a list of packages. We are still away from having packages >> as real first class entities, but this is a good step toward. Note >> that some people find O2 a bit slower than other available code >> browsers. >> >> David and I spent some time on fixing O2. O2 has now no dependencies >> to OB anymore. >> We would like that people try it and submit bug reports on the google >> bug tracker. In a fresh 10505, you can load O2 by evaluating: >> >> Gofer new squeaksource: 'MetacelloRepository'; package: >> 'ConfigurationOfO2'; load. (Smalltalk at: #ConfigurationOfO2) perform: >> #loadDefault >> >> > Yes, but be aware that that piece of code load only the core, as you can see > here: > > group: 'default' with: #('Core' ); > group: 'Core' with: #( 'OmniBrowser2' 'O2-Standard' 'O2-Morphic' > 'O2-Enhancements' ); > group: 'Dev' with: #( 'Core' 'O2-Refactory' 'OCForO2' ); > > > As I remember, most of the problems of O2 were sometimes related to the > integration to RB. So, if you really want to tests it, please, test it using > the RB: > > ((Smalltalk at: #ConfigurationOfO2) project version: '1.0') load: 'Dev' > > > As I already said to Alexandre and David in private, I will put O2 again in > the next image. So, it would be cool if it is tested "completely". > > > >> O2 does not need OB whatsoever. O2 loads in a Core (I tried in a >> 11165). Before and after loading O2, running OB tests give "548 run, >> 547 passes, 1 failures" [*]. >> >> > Yes, if someone can help with that failure would be cool, I have already > reported it. > > > >> Cheers, >> Alexandre >> >> [*] the test have been loaded using (ConfigurationOfOmniBrowser >> project version: '1.1') load: 'Dev Tests' >> -- >> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >> Alexandre Bergel http://www.bergel.eu >> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >> >> >> >> >> >> >> _______________________________________________ >> Pharo-project mailing list >> Pharo-project(a)lists.gforge.inria.fr >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Jan. 20, 2010

Re: [Pharo-project] We remove CodeLoader ?? [WAS] Presenter and Project will die aren't they ?

by Mariano Martinez Peck

And that not depends on HTTPLoader please!!!! On Wed, Jan 20, 2010 at 3:22 PM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr > wrote: > Yes this would be good to have a clean ScriptExecuter class. > > Stef > > On Jan 20, 2010, at 9:48 AM, Mariano Martinez Peck wrote: > > > > > > > On Wed, Jan 20, 2010 at 9:13 AM, Michael Roberts <mike(a)mjr104.co.uk> > wrote: > > if you follow the send into CodeLoader you will see fileIn sent at the > > end of installSourceFile:. All the security manager stuff and request > > stuff can be dropped so just try and copy the minimum that you need > > into the class. > > > > Yes, I thought the same. > > > > Also, the startup at login can be dropped as well, > > since the class can be activated or deactivated by removing it from > > the startup list. > > > > ahh cool :) > > > > I'll take a look later on. > > > > > > Excellent. I created > http://code.google.com/p/pharo/issues/detail?id=1854 > > I wanted to put you in cc but I didn't find you in the list :( > > > > Thanks > > > > Mariano > > > > > > cheers, > > Mike > > > > 2010/1/19 Mariano Martinez Peck <marianopeck(a)gmail.com>: > > > > > > > > > On Tue, Jan 19, 2010 at 6:12 PM, Michael Roberts <mike(a)mjr104.co.uk> > wrote: > > >> > > >> we can remove it but we need to keep the ability to load scripts from > > >> the command line. > > >> > > > > > > Mike: Thanks for you answer. I really don't know too much of > CodeLoader, > > > that's why I ask. I checked that about running from command line. You > are > > > right. > > > > > > > > >> > > >> so, your code is fine but you can't zap CodeLoader since it's still > > >> referenced for the load. It would be easier just to make a new class > > >> that only loads scripts from the command line, and zap the rest. > > >> > > > > > > Yes, maybe that's a better idea. However, I think that maybe we even > don't > > > need a new class. Maybe we can do something like this: > > > > > > startUpAfterLogin > > > | scriptName loader isUrl | > > > self readDocumentAtStartup ifTrue: [ > > > scriptName := (SmalltalkImage current getSystemAttribute: 2) > > > ifNil:['']. > > > scriptName := scriptName convertFromSystemString.. > > > ] > > > ifFalse: [ scriptName := '' ]. > > > > > > scriptName isEmptyOrNil > > > ifTrue:[^ self]. > > > > > > self loadAndInstallSources: scriptName. > > > > > > > > > And then, we define ProjectLauncher >> loadAndInstallSources: > aScriptName > > > > > > Here we should do a fileIn, but I am not sure how to do that. Can you > help > > > me ? > > > > > > What do you think ? Is it worth this or keep CodeLoader ? I reall > don't > > > LIKE AT ALL to use something that depends on HTTPLoader to do > that...wtf > > > ??? Look CodeLoader >> createRequestFor: > > > > > > Please, take into account I am very newbie with this stuff, so maybe > there > > > are a lot of problems I am not seeing. I just asked :) > > > > > > Cheers > > > > > > Mariano > > > > > > > > >> cheers > > >> Mike > > >> > > >> 2010/1/19 Mariano Martinez Peck <marianopeck(a)gmail.com>: > > >> > So....we will remove CodeLoader then ? This is the moment to speak > :) > > >> > > > >> > I ask about this because I am trying to analize who uses > ImageSegments > > >> > and > > >> > all that stuff in the code and see how easily or good is to remove > > >> > ImageSegment to a separate package. And, CodeLoader uses > ImageSegment. > > >> > > > >> > As far as I looked, the only user or CodeLoader, as Michael said is > > >> > ProjectLauncher >> startUpAfterLogin > > >> > Which....I think it will also die in a moment if Project will die > > >> > too.... > > >> > > > >> > Anyway, but do you think about changing startUpAfterLogin to this: > > >> > > > >> > startUpAfterLogin > > >> > | scriptName loader isUrl | > > >> > self readDocumentAtStartup ifTrue: [ > > >> > scriptName := (SmalltalkImage current getSystemAttribute: 2) > > >> > ifNil:['']. > > >> > scriptName := scriptName convertFromSystemString. > > >> > scriptName isEmpty ifFalse:[ > > >> > "figure out if script name is a URL by itself" > > >> > isUrl := (scriptName asLowercase beginsWith:'http://') > or:[ > > >> > (scriptName asLowercase beginsWith:'file://') > or:[ > > >> > (scriptName asLowercase beginsWith:'ftp://')]]. > > >> > isUrl ifFalse:[scriptName := 'file:',scriptName]]. > > >> > . ] > > >> > ifFalse: [ scriptName := '' ]. > > >> > > > >> > scriptName isEmptyOrNil > > >> > ifTrue:[^ self]. > > >> > loader := CodeLoader new. > > >> > loader loadSourceFiles: (Array with: scriptName). > > >> > loader installSourceFiles. > > >> > > > >> > > > >> > > > >> > On Tue, Jan 12, 2010 at 9:56 AM, StÃ©phane Ducasse > > >> > <stephane.ducasse(a)inria.fr> wrote: > > >> >> > > >> >> ok we got convinced :) > > >> >> > > >> >> On Jan 12, 2010, at 8:20 AM, John M McIntosh wrote: > > >> >> > > >> >> > Ah well yes Michael suffered thru managing, testing, and working > with > > >> >> > all those browsers & plugins. > > >> >> > I suffered thru trying to figure out how it all worked... At > least > > >> >> > on > > >> >> > the mac. > > >> >> > > > >> >> > Here's the current issue, because: > > >> >> > > > >> >> > Microsoft shafted netscape by abandoning the netscape API, then > > >> >> > shafted > > >> >> > apple by not providing an alternative > > >> >> > we had the strange mixture of a solution for os-x, windows, X11. > > >> >> > Then > > >> >> > of course Microsoft IE on macintosh > > >> >> > would change behaviour from version to version to screw with > FireFox > > >> >> > & > > >> >> > Mozilla etc. Let alone worrying about Opera which > > >> >> > kinda worked. Fortunately Microsoft abandon the Mac browser > market, > > >> >> > and Apple introduced Safari which again was slightly different. > > >> >> > > > >> >> > Now on 10.6 we had this problem an abandon API for plugins, that > > >> >> > needed > > >> >> > to move forward, also into the 64bit world. > > >> >> > So Apple changed things again to either let you run Safari in > 32bit > > >> >> > using old plugins (busted), or using a modified 64bit plugins > api, > > >> >> > but > > >> >> > not both at the same time. > > >> >> > > > >> >> > The words busted, non-existant, abandon, (implied ugly), appear > > >> >> > quite a > > >> >> > few times in the sentences above. Also on every > > >> >> > browser version change, or operating system major version change > > >> >> > something was just slightly funky enough to bring down > > >> >> > the house of cards. > > >> >> > > > >> >> > So, time is better spent in getting 64bit squeak working, since > there > > >> >> > is > > >> >> > plenty of hours still needed to fixup all the plugins, > > >> >> > and check that the VM is actually sane... > > >> >> > > > >> >> > > > >> >> > On 2010-01-11, at 1:59 PM, Michael Rueger wrote: > > >> >> > > > >> >> >> On 1/11/2010 10:58 PM, Michael Rueger wrote: > > >> >> >> > > >> >> >>> forget squeak/pharo in the browser. > > >> >> >> > > >> >> >> for those not in the know: > > >> >> >> I'm one of the poor guys who actually spent a *lot* of time > making > > >> >> >> it > > >> >> >> possible ;-) > > >> >> >> > > >> >> >> Michael > > >> >> >> > > >> >> >> _______________________________________________ > > >> >> >> Pharo-project mailing list > > >> >> >> Pharo-project(a)lists.gforge.inria.fr > > >> >> >> > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > >> >> > > > >> >> > -- > > >> >> > > > >> >> > > > >> >> > > =========================================================================== > > >> >> > John M. McIntosh <johnmci(a)smalltalkconsulting.com> Twitter: > > >> >> > squeaker68882 > > >> >> > Corporate Smalltalk Consulting Ltd. > > >> >> > http://www.smalltalkconsulting.com > > >> >> > > > >> >> > > > >> >> > > =========================================================================== > > >> >> > > > >> >> > > > >> >> > > > >> >> > > > >> >> > > > >> >> > _______________________________________________ > > >> >> > Pharo-project mailing list > > >> >> > Pharo-project(a)lists.gforge.inria.fr > > >> >> > > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > >> >> > > >> >> > > >> >> _______________________________________________ > > >> >> Pharo-project mailing list > > >> >> Pharo-project(a)lists.gforge.inria.fr > > >> >> > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > >> > > > >> > > > >> > _______________________________________________ > > >> > Pharo-project mailing list > > >> > Pharo-project(a)lists.gforge.inria.fr > > >> > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > >> > > > >> > > >> _______________________________________________ > > >> Pharo-project mailing list > > >> Pharo-project(a)lists.gforge.inria.fr > > >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > > > > > > > > _______________________________________________ > > > Pharo-project mailing list > > > Pharo-project(a)lists.gforge.inria.fr > > > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > > > > > > _______________________________________________ > > Pharo-project mailing list > > Pharo-project(a)lists.gforge.inria.fr > > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > > > _______________________________________________ > > Pharo-project mailing list > > Pharo-project(a)lists.gforge.inria.fr > > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >

Jan. 20, 2010

Re: [Pharo-project] Something I don't understand with Method Finder

by Mariano Martinez Peck

But Marcus, in this case it is not even used as symbol but not as selectot. It is not at all in the image. Neither selector, neither symbol, neither a comment, NOTHING. It it just in the .changes if you load it with a text editor and I can see: "Preferences class removeSelector: #enableGently:!" On Wed, Jan 20, 2010 at 3:23 PM, Marcus Denker <marcus.denker(a)inria.fr>wrote: > > On Jan 20, 2010, at 2:45 PM, Mariano Martinez Peck wrote: > > > Hi folks! If you look for example "gently" in Method Finder, you will > see that it lists, for example, "enableGently:" and "disableGently:" > > > > However, those methods are not implemented, there are no senders, and I > even did a "method source with it " and I found nothing. > > > > In my final shut, I read the .changes and I found: > > > > "Preferences class removeSelector: #enableGently:!" > > > > Now I wonder...is this a feature of Method Finder to look also in > .changes or it is a bug ? Anyway, it is not clear when you find a message > but you don't have implementros or senders, it means that it was in > .changes. > > > > What do you think? > > It's broken... I think it searches all symbols, not all > symbols-used-as-selectors. > > Marcus > > -- > Marcus Denker -- http://www.marcusdenker.de > INRIA Lille -- Nord Europe. Team RMoD. > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >

Jan. 20, 2010

Re: [Pharo-project] how to get nice native fonts in 1.0

by StÃ©phane Ducasse

tx Stef On Jan 20, 2010, at 10:51 AM, Simon Denier wrote: > > On 20 janv. 2010, at 09:27, Damien Cassou wrote: > >> On Wed, Jan 20, 2010 at 8:48 AM, StÃ©phane Ducasse >> <stephane.ducasse(a)inria.fr> wrote: >>> I have to give some demos of seaside today and I would like to have nice fonts in 1.0 >>> and I cannot find the preferences to say that the fonts can be looked up on my disc. >>> and I'm looking for some nice combination (but the code snippets on the wiki do not work). >> >> FreeTypeFontProvider current updateFromSystem. >> >> font := LogicalFont familyName: 'DejaVu Sans' pointSize: 10. >> codeFont := LogicalFont familyName: 'DejaVu Sans Mono' pointSize: 9. >> titleFont := LogicalFont familyName: 'DejaVu Sans' pointSize: 11. >> >> Preferences >> setListFontTo: font; >> setMenuFontTo: font; >> setCodeFontTo: codeFont; >> setButtonFontTo: font; >> setSystemFontTo: font; >> setWindowTitleFontTo: titleFont. >> > > > > Yep, this is more or less the script I use in my images (Pharo 1.0) > > <Screen shot 2010-01-20 at 10.48.37.png> > -- > Simon > > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Jan. 20, 2010

Re: [Pharo-project] Something I don't understand with Method Finder

by Marcus Denker

On Jan 20, 2010, at 2:45 PM, Mariano Martinez Peck wrote: > Hi folks! If you look for example "gently" in Method Finder, you will see that it lists, for example, "enableGently:" and "disableGently:" > > However, those methods are not implemented, there are no senders, and I even did a "method source with it " and I found nothing. > > In my final shut, I read the .changes and I found: > > "Preferences class removeSelector: #enableGently:!" > > Now I wonder...is this a feature of Method Finder to look also in .changes or it is a bug ? Anyway, it is not clear when you find a message but you don't have implementros or senders, it means that it was in .changes. > > What do you think? It's broken... I think it searches all symbols, not all symbols-used-as-selectors. Marcus -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

Jan. 20, 2010

Re: [Pharo-project] Parrot VM

by StÃ©phane Ducasse

ok so this is totally different then. Stef On Jan 20, 2010, at 9:20 AM, Stefan Marr wrote: > You mean how their approaches compare? (I would guess, there is no direct involvement of people in both projects.) > > PyPy is a generator approach, your language specification i.e. the interpreter is used > to generate a VM. > You use a high-level language i.e. RPython to implement a classic interpreter, which is then transformed by a powerful tool-chain to your C implementation (and a JIT). Stuff like GC is just generated for you. > The result is one specific VM, which is generated for your language. > > > With Parrot it is much like with a classic VM like JVM or CLR/.NET. > You are using one existing VM. > Language developers use the Parrot Assembly language (PASM) or the Parrot > Intermediate Representation (PIR) which can then be compiled to the bytecode set. > Furthermore, they can extend the bytecode set with their own instructions (primitives). > So, you are basically compiling your target language while using all the runtime features of the Parrot VM. So, here the 'power' is in the VM, not in a transformation tool chain. > > Best > Stefan > > > > On 20 Jan 2010, at 09:02, StÃ©phane Ducasse wrote: > >> stefan >> >> do you know how this is linked with pypy? >> >> Stef >> >> On Jan 20, 2010, at 8:50 AM, Stefan Marr wrote: >> >>> >>> On 20 Jan 2010, at 06:12, Miguel Enrique CobÃ¡ Martinez wrote: >>> >>>> Parrot has just been released: >>>> http://www.parrot.org/news/2010/Parrot-2.0.0 >>>> >>>> I would like to know, from the VM experts from the community, if this VM >>>> can be used to run Smalltalk. >>> Well, as someone mentioned, there seems to be Smalltalk implementation... >>> >>> But, probably it is hard to get it fast. >>> Parrot uses a very unique approach. >>> Actually, they try to provide a language runtime for all kinds of languages, and the right tools to build them: http://www.linux-mag.com/cache/7373/1.html >>> >>> Eliot mentioned primitives, well, there is no concept of a primitive in Parrot. >>> Instead, they have a unification of bytecodes and primitives. The idea is, to give you the tools to develop the bytecode you need for your language, which can be loaded dynamically during execution. >>> It will be interesting to see, how they can get their JIT fast with this strategy. >>> >>> Best >>> Stefan >>> >>> >>>> The site says that >>>> >>>> "Parrot is a virtual machine aimed at running all dynamic languages.". >>>> >>>> but I don't know anything about VM so this sounds to me a bit like black >>>> magic. >>>> >>>> Is there something in the squeak vm that is specific to the way >>>> smalltalk works or a generic virtual machine (maybe with a upper layer >>>> understanding Smalltalk specifics) can be used. >>>> >>>> Thanks for the answers >>>> >>>> -- >>>> Miguel CobÃ¡ >>>> http://miguel.leugim.com.mx >>>> >>>> >>>> _______________________________________________ >>>> Pharo-project mailing list >>>> Pharo-project(a)lists.gforge.inria.fr >>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >>> >>> -- >>> Stefan Marr >>> Software Languages Lab >>> Vrije Universiteit Brussel >>> Pleinlaan 2 / B-1050 Brussels / Belgium >>> http://soft.vub.ac.be/~smarr >>> Phone: +32 2 629 3956 >>> Fax: +32 2 629 3525 >>> >>> >>> _______________________________________________ >>> Pharo-project mailing list >>> Pharo-project(a)lists.gforge.inria.fr >>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> >> >> _______________________________________________ >> Pharo-project mailing list >> Pharo-project(a)lists.gforge.inria.fr >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > -- > Stefan Marr > Software Languages Lab > Vrije Universiteit Brussel > Pleinlaan 2 / B-1050 Brussels / Belgium > http://soft.vub.ac.be/~smarr > Phone: +32 2 629 3956 > Fax: +32 2 629 3525 > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Jan. 20, 2010