October 2009 - Pharo-dev - lists.pharo.org

Re: [Pharo-project] [squeak-dev] Re: Selection in new file list / file system api

by Damien Cassou

On Thu, Oct 8, 2009 at 2:03 AM, Torsten Bergmann <astares(a)gmx.de> wrote: > Damien also proposed a very small API that allows you to browse > the file system easily including tests, just check > > http://code.google.com/p/pharo/issues/detail?id=1146 Forget about that. This is a very *very* small API that provide the very strict minimum to browse a file system. -- Damien Cassou http://damiencassou.seasidehosting.st "Lambdas are relegated to relative obscurity until Java makes them popular by not having them." James Iry

Oct. 8, 2009

[Pharo-project] RFBServer problem

by Dennis Schetinin

Using pharo1.0-10466-BETAweb09.10.1 + RFB-DamienCassou.28 (latest): the image stops to respond after closing connection remotely. Stopping the server (via VNC) leaves image ok. -- Dennis Schetinin

Oct. 8, 2009

Re: [Pharo-project] calling for a fixer :)

by StÃ©phane Ducasse

be my guest! I like this kind of charming song :) Stef On Oct 8, 2009, at 10:01 AM, Johan Brichau wrote: > > On 07 Oct 2009, at 21:55, StÃ©phane Ducasse wrote: > >> did you open a ticket? > > > Yes, I did: 1288 > but I will probably send in a cleaner fix today (instead of the two > previous ones... sorry about that!) > > ---------------------------- > Johan Brichau > johan.brichau(a)uclouvain.be > > > > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Oct. 8, 2009

Re: [Pharo-project] Selection in new file list / file system api

by Torsten Bergmann

Colin wrote on squeak-dev: >I've been working on a new filesystem API built on top of the >FilePlugin primitives. >... >if anybody wants to take a look it's here: >http://source.wiresong.ca/mc/Filesystem-cwp.22.mcz Damien also proposed a very small API that allows you to browse the file system easily including tests, just check http://code.google.com/p/pharo/issues/detail?id=1146 Bye Torsten -- GRATIS fÃ¼r alle GMX-Mitglieder: Die maxdome Movie-FLAT! Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome01

Oct. 8, 2009

Re: [Pharo-project] calling for a fixer :)

by Johan Brichau

On 07 Oct 2009, at 21:55, StÃ©phane Ducasse wrote: > did you open a ticket? Yes, I did: 1288 but I will probably send in a cleaner fix today (instead of the two previous ones... sorry about that!) ---------------------------- Johan Brichau johan.brichau(a)uclouvain.be

Oct. 8, 2009

Re: [Pharo-project] calling for a fixer :)

by Johan Brichau

On 07 Oct 2009, at 21:38, StÃ©phane Ducasse wrote: > I want benchs :) > > Stef Here they are: Time millisecondsToRun: [1 to: 10 do:[:i | SystemNavigation new allCallsOn: #add: ]]. "current: 2234 2240 2261" "refactored: 2806 2799 2816" Time millisecondsToRun: [1 to: 10 do:[:i | SystemNavigation new allCallsOn: #foobar ]]. "current: 2166 2216 2159" "refactored: 2826 2790 2933" So, an overhead of 20-30%. However, making that refactoring will probably also trigger a lot of changes elsewhere. The current #literalsDo: method iterates over header, method properties and class binding too and it is sent by external methods. In addition, there is the #hasLiteral:, #hasLiteralThorough: and #sendsSelector: methods on CompiledMethod that all perform roughly identical tasks but it is far from clear to me who is expected to do what, especially since the comments are sometimes not in sync with what the method actually does. It seems like this part of the code is another candidate for cleanup. ---------------------------- Johan Brichau johan.brichau(a)uclouvain.be

Oct. 8, 2009

Re: [Pharo-project] Hierarchy browsers & implementers of

by StÃ©phane Ducasse

Ahh ok that! The problem is that imagine that you would see superclasses of ProtoObject class then how would you relate let's say Behavior when you click back on the instance button. I think that the instance class button is somehow misleading. and may be it would be better to have Object class Collection class .... instead of Object Collection and class button pressed Stef On Oct 7, 2009, at 10:32 PM, Petr VÃ¡cha wrote: > I know Smalltalk hierarchy, I just wanted to point out, that Hierarchy > Browser does not provide (maybe I am dumb, but I do not see any) > obvious > way to show the hierarchy to see that Array class inherits from > Behavior. > > PV. > > ----- PÅ®VODNÃ ZPRÃVA ----- > Od: "StÃ©phane Ducasse" <stephane.ducasse(a)inria.fr> > Komu: "Petr VÃ¡cha" <recycler(a)volny.cz> > PÅedmÄt: Re: Hierarchy browsers & implementers of > Datum: 7.10.2009 - 21:22:42 > >> I suggest you to read the chapter on the Smalltalk >> model in the Squeak >> or pharo by example book >> the key is: >> >> when you look for a message you look in the CLASS of >> the receiver >> then its superclasses. >> >> So when you look for new sent to Array you should look >> in the Array >> class superclasses and you will see it contains >> Behavior before ProtoObject >> >> Stef >> >> PS: you can sent to the mailing-list just register >> to it. You do not >> need to have a google account. >> >> On Oct 7, 2009, at 1:00 PM, Petr VÃ¡cha wrote: >> >>> Hi again, >>> when try to find out that Array>>new: bug I found >>> one thing that I >>>> think >>> may be very confusing - somehow mangled information >>> from hierarchy >>>> browsers >>> (see figure). >>> >>> Array>>new: calls primitive and uses self>>basicNew: >>> call as fallback. >>>> Array does not implement basicNew: so I tried to >>> find implementors of >>>> it. Behavior has this implementation. However, when >>> you show hierarchy >>>> of Array, it ends in ProtoObject with no Behavior >>> involved. I think >>>> this to be is somewhat confusing for beginners especially >>> when they >>>> are trying to get into method lookup chain. >>> >>> Petr Vacha >>> >>> >>> <pharo-browsers.jpg> >> > > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Oct. 8, 2009

Re: [Pharo-project] Loading Gemstone Variable (Indexable) classes into Pharo via Monticello

by StÃ©phane Ducasse

On Oct 7, 2009, at 10:18 PM, rkrock(a)locksleycreeksoftware.com wrote: > I have the following in Gemstone: > > FristClass which is a indexableSublass of Object. can you check the fix for the classbuilder in the bug archive and let us know if it fixes the problem? http://code.google.com/p/pharo/issues/detail?id=1270 I guess not. Now I'm wondering how the other smalltalk react to a variableSubclass being subclass. Did you try? Finally what you should know is that the classBuilder in Squeak/pharo is really an ugly beast that one of these will have to be rewritten from scratch Stef > SecondClass which is a subclass of FirstClass. > > When I save this to Monticello and then load it into Pharo, the load > fails > on the 'recompile' because the SecondClass is not of type #variable. > > The load works when I modify the method > MCClassDefinition>>createClass to > use the superclass' type when the superclass type is #variable. > > --- current way ---- > ... > type: type; > ... > > --- hacked way ---- > .... > type: ((superClass typeOfClass isSymbol not and: [superClass > typeOfClass > isVariable]) ifTrue: [superClass typeOfClass] ifFalse: > [type]); > .... > > With this I was sucessfully able to important the code into Pharo. > > I am able to go from Pharo to GemStone with no problems. > > Is there a general solution to this problem? It obviously will not > occur > very often. > > Regards, > > Reg > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Oct. 8, 2009

Re: [Pharo-project] next steps... to we create a release candidate and open a 1.1 alpha?

by Michael Roberts

On Wed, Oct 7, 2009 at 8:19 AM, Adrian Lienhard <adi(a)netstyle.ch> wrote: > Hi Mike, > > On Oct 6, 2009, at 22:36 , Michael Roberts wrote: > >> i would mark [...] > > I agree. Do you want to update the tracker? > i've updated the debugger highlight. i'll mark the decompiler tests later in the week. they don't yet have a # i think. > > I noticed that after the recent closure fixes there are a couple of > new failures and errors. For instance > AuthorTest>>#testDeprecatedSendsRemoved (I guess the reported senders > of a selector changed with some closure changes). > > see http://code.google.com/p/pharo/wiki/BaselineTestResults > yes, i saw and the recent thread. one of the things that is difficult with merging eliot's changes is knowing what is intentional or not between the pharo base and the change. it would be good if we could comment or annotate key 'pharo fixes' (to the compiler & machinery) so that it is obvious in a diff browser. if we genuinely think we have a fix then we can feed it back to eliot. we are likely to want to take compiler changes for sometime, presumably all the way to the JIT. > > ok. I think we should seriously check the remaining errors/failures > before marking them. yes i would like a serious review but of course it requires folks to do it. now you're back from your holiday ;-)... the other thing i thought was a good candidate for 1.0 was the latest MC patch stef posted. those semantics are vital, and look an increment on previous changes we had taken to load the compiler. i just wanted to test it before recommending. cheers, Mike

Oct. 7, 2009

Re: [Pharo-project] Hierarchy browsers & implementers of

by Alexandre Bergel

Hi! you're very right, the browser does not show this. Open a browser, look for the class ProtoObject. You can see that the superclass of ProtoObject is nil, but if you select the 'class' tab, you just see the declaration of instance variable names for the meta-class. This does not show you the superclass of the meta-class. In practice, this is not something that you need to know. Well, maybe for #new and #new: You need to open a workspace and type 'ProtoObject class superclass' to discover that it is Class. Cheers, Alexandre On 7 Oct 2009, at 16:32, Petr VÃ¡cha wrote: > I know Smalltalk hierarchy, I just wanted to point out, that Hierarchy > Browser does not provide (maybe I am dumb, but I do not see any) > obvious > way to show the hierarchy to see that Array class inherits from > Behavior. > > PV. > > ----- PÅ®VODNÃ ZPRÃVA ----- > Od: "StÃ©phane Ducasse" <stephane.ducasse(a)inria.fr> > Komu: "Petr VÃ¡cha" <recycler(a)volny.cz> > PÅedmÄt: Re: Hierarchy browsers & implementers of > Datum: 7.10.2009 - 21:22:42 > >> I suggest you to read the chapter on the Smalltalk >> model in the Squeak >> or pharo by example book >> the key is: >> >> when you look for a message you look in the CLASS of >> the receiver >> then its superclasses. >> >> So when you look for new sent to Array you should look >> in the Array >> class superclasses and you will see it contains >> Behavior before ProtoObject >> >> Stef >> >> PS: you can sent to the mailing-list just register >> to it. You do not >> need to have a google account. >> >> On Oct 7, 2009, at 1:00 PM, Petr VÃ¡cha wrote: >> >>> Hi again, >>> when try to find out that Array>>new: bug I found >>> one thing that I >>>> think >>> may be very confusing - somehow mangled information >>> from hierarchy >>>> browsers >>> (see figure). >>> >>> Array>>new: calls primitive and uses self>>basicNew: >>> call as fallback. >>>> Array does not implement basicNew: so I tried to >>> find implementors of >>>> it. Behavior has this implementation. However, when >>> you show hierarchy >>>> of Array, it ends in ProtoObject with no Behavior >>> involved. I think >>>> this to be is somewhat confusing for beginners especially >>> when they >>>> are trying to get into method lookup chain. >>> >>> Petr Vacha >>> >>> >>> <pharo-browsers.jpg> >> > > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

Oct. 7, 2009