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

Re: [Pharo-project] Arg assignement [Fwd: how can we sync]

by Nicolas Cellier

No, after a full review of 'Closure Usage Fixes.1.cs', I verified this file only fix minor case of blocks writing in outside temps, not the case of blocks writing their own arguments. Nicolas 2009/4/16 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: > Thanks nicolas > You are talking about > Â Â Â Â http://code.google.com/p/pharo/issues/detail?id=742 > > Stef > > On Apr 15, 2009, at 11:10 PM, Nicolas Cellier wrote: > >> I published very trivials cleanups from Eliots abount Sounds package >> in PharoInbox >> I will continue reviewing tomorrow. >> Bye >> >> 2009/4/15 Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com>: >>> Oops, I just miss the attachments :( >>> >>> 2009/4/15 Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com>: >>>> >>>> Is there a more recent version? >>>> >>>> Nicolas >>>> >>> >> >> _______________________________________________ >> 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 >

April 16, 2009

Re: [Pharo-project] The System Package

by StÃ©phane Ducasse

Same here ScriptLoader new deletePackage: 'System' -> error no On Apr 16, 2009, at 10:16 AM, Lukas Renggli wrote: >> It seems that the class category 'System' is registered as a package >> and each System-XXX category is both a package and a class category. >> As a consequence, System is a package that contains sub-packages, and >> 'System-Object Storage' is listed as a package. > > I don't have a package registered called 'System', however there are > some other packaging related problems in the latest Pharo. The classes > BlockClosuresTestCase, ClosureTests, CollectionCombinator, and > DecompilerTests are in no package. Ah this is another problem. I will add an entry when the web site works again. > > > I used the following script, that is part of code critics: > > | packages | > packages := MCWorkingCopy allManagers > collect: [ :each | each packageInfo ]. > SystemNavigation default allClassesAndTraits > do: [ :class | > (packages count: [ :package | package includesSystemCategory: class > category ]) ~= 1 > ifTrue: [ Transcript show: class name; show: ' ('; show: class > category; show: ')'; cr ] > ifFalse: [ [ :mclass | > mclass selectors do: [ :selector | > (packages count: [ :package | package includesMethod: selector > ofClass: mclass ]) ~= 1 > ifTrue: [ Transcript show: mclass name; show: '>>'; show: > selector; cr ] ] ] value: class; value: class class ] ] > displayingProgress: 'Checking Packages' > >> I suggest that package name should never contain a '-'. This require >> quite some discipline to not mess up the thing. > > That does not solve the problem of overlapping packages. See > <http://code.google.com/p/seaside/wiki/PackageNaming> for a large > scale packaging scheme that works. > > 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 >

April 16, 2009

Re: [Pharo-project] WeakArray class>>#finalizationProcess

by StÃ©phane Ducasse

We are always interested in better solutions. Do you have code and got good feedback? Stef > Btw, if anyone interested, i explored what the finalizationProcess > does in the past.. > and found some things i don't like. > Maybe now in Pharo, you can cosider about fixing/improving this stuff. > > http://lists.squeakfoundation.org/pipermail/squeak-dev/2007-July/119052.html > http://lists.squeakfoundation.org/pipermail/squeak-dev/2007-July/119054.html > > 2009/4/16 Markus Fritsche <fritsche.markus(a)gmx.net>: >> StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: >> >>> Strange I fixed those ones. Can you check if you have the same >>> problem >>> in the latest? >> >> It could have happened that I did overwrite those by loading new >> PolyMorph >> packages. >> >>> See the other email :) >> >> Merci! >> >>> Torsten sent us a loadFFI now available in scriptLoader in one the >>> latest updates. >>> Can you check? >> >> Seems to work well. Thanks for the pointer. >> >>> If you can reproduce that with the latest (updated image). >>> Let us know. We are chasing bugs now. >> >> Nope, working now I think. >> >> >> _______________________________________________ >> Pharo-project mailing list >> Pharo-project(a)lists.gforge.inria.fr >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > > > -- > Best regards, > Igor Stasenko AKA sig. > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >

April 16, 2009

Re: [Pharo-project] WeakArray class>>#finalizationProcess

by StÃ©phane Ducasse

>> Strange I fixed those ones. Can you check if you have the same >> problem >> in the latest? > > It could have happened that I did overwrite those by loading new > PolyMorph > packages. for the progressbar probably. I ask gary to check that. > > >> See the other email :) > > Merci! > >> Torsten sent us a loadFFI now available in scriptLoader in one the >> latest updates. >> Can you check? > > Seems to work well. Thanks for the pointer. > >> If you can reproduce that with the latest (updated image). >> Let us know. We are chasing bugs now. > > Nope, working now I think. Cool I recompile the image again and fixed some points. stef

April 16, 2009

Re: [Pharo-project] Arg assignement [Fwd: how can we sync]

by StÃ©phane Ducasse

Thanks nicolas You are talking about http://code.google.com/p/pharo/issues/detail?id=742 Stef On Apr 15, 2009, at 11:10 PM, Nicolas Cellier wrote: > I published very trivials cleanups from Eliots abount Sounds package > in PharoInbox > I will continue reviewing tomorrow. > Bye > > 2009/4/15 Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com>: >> Oops, I just miss the attachments :( >> >> 2009/4/15 Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com>: >>> >>> Is there a more recent version? >>> >>> Nicolas >>> >> > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >

April 16, 2009

Re: [Pharo-project] Arg assignement [Fwd: how can we sync]

by Nicolas Cellier

Balloon-nice.17 trivial BlockClosure cleanup from Eliot's 'Closure Usage Fixes.1.cs' ---- Collections-Arrayed-nice.16 trivial BlockClosure cleanup from Eliot's 'Closure Usage Fixes.1.cs' I also cleaned evalStrings a little further ---- Collections-Unordered-nice.13 trivial BlockClosure cleanup from Eliot's 'Closure Usage Fixes.1.cs' ---- Compression-nice.19 trivial BlockClosure cleanup from Eliot's 'Closure Usage Fixes.1.cs' ---- Compiler-nice.99 trivial BlockClosure cleanup from Eliot's 'Closure Usage Fixes.1.cs' ---- Files-nice.ducasse.46 trivial BlockClosure cleanup from Eliot's 'Closure Usage Fixes.1.cs' ---- System-Change Notification-nice.3 trivial BlockClosure cleanup from Eliot's 'Closure Usage Fixes.1.cs' ---- System-Object Storage-nice.7 trivial BlockClosure cleanup from Eliot's 'Closure Usage Fixes.1.cs' ---- System-Changes-nice.3 trivial BlockClosure cleanup from Eliot's 'Closure Usage Fixes.1.cs' ---- Tools-nice.149 trivial BlockClosure cleanup from Eliot's 'Closure Usage Fixes.1.cs' ---- Kernel-nice.renggli.290 trivial BlockClosure cleanup from Eliot's 'Closure Usage Fixes.1.cs' ---- Traits-nice.271 trivial BlockClosure cleanup from Eliot's 'Closure Usage Fixes.1.cs' ---- Graphics-nice.86 trivial BlockClosure cleanup from Eliot's 'Closure Usage Fixes.1.cs' Nicolas 2009/4/15 Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com>: > I published very trivials cleanups from Eliots abount Sounds package > in PharoInbox > I will continue reviewing tomorrow. > Bye > > 2009/4/15 Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com>: >> Oops, I just miss the attachments :( >> >> 2009/4/15 Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com>: >>> >>> Is there a more recent version? >>> >>> Nicolas >>> >> >

April 16, 2009

Re: [Pharo-project] About literalsDo: deprecation

by StÃ©phane Ducasse

In fact I do not know why it has been deprecated. Marcus probably knows. We should certainly undeprecated it :) Stef On Apr 15, 2009, at 10:53 PM, Nicolas Cellier wrote: > I notice while scanning Eliot changes that #literalsDo: is a > deprecated message... > But (literals do:) replacement is not neutral concerning speed. > It creates an Array where literalsDo: did not. > This is not a detail because this is called on each and every method > when scanning code: > > For example, I tried this in Pharo image: > > Time millisecondsToRun: [SystemNavigation default allGlobalRefs] > ->5356 with Eliot's literalsDo: > ->8947 pharo version calling literals do: > > This change spoils a few minutes out of our every day work ;) > > Nicolas > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >

April 16, 2009

Re: [Pharo-project] The System Package

by Lukas Renggli

> It seems that the class category 'System' is registered as a package > and each System-XXX category is both a package and a class category. > As a consequence, System is a package that contains sub-packages, and > 'System-Object Storage' is listed as a package. I don't have a package registered called 'System', however there are some other packaging related problems in the latest Pharo. The classes BlockClosuresTestCase, ClosureTests, CollectionCombinator, and DecompilerTests are in no package. I used the following script, that is part of code critics: | packages | packages := MCWorkingCopy allManagers collect: [ :each | each packageInfo ]. SystemNavigation default allClassesAndTraits do: [ :class | (packages count: [ :package | package includesSystemCategory: class category ]) ~= 1 ifTrue: [ Transcript show: class name; show: ' ('; show: class category; show: ')'; cr ] ifFalse: [ [ :mclass | mclass selectors do: [ :selector | (packages count: [ :package | package includesMethod: selector ofClass: mclass ]) ~= 1 ifTrue: [ Transcript show: mclass name; show: '>>'; show: selector; cr ] ] ] value: class; value: class class ] ] displayingProgress: 'Checking Packages' > I suggest that package name should never contain a Â '-'. This require > quite some discipline to not mess up the thing. That does not solve the problem of overlapping packages. See <http://code.google.com/p/seaside/wiki/PackageNaming> for a large scale packaging scheme that works. Lukas -- Lukas Renggli http://www.lukas-renggli.ch

April 16, 2009

[Pharo-project] The System Package

by Alexandre Bergel

Dear All, It seems that the class category 'System' is registered as a package and each System-XXX category is both a package and a class category. As a consequence, System is a package that contains sub-packages, and 'System-Object Storage' is listed as a package. I suggest that package name should never contain a '-'. This require quite some discipline to not mess up the thing. Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

April 16, 2009

Re: [Pharo-project] WeakArray class>>#finalizationProcess

by Igor Stasenko

Btw, if anyone interested, i explored what the finalizationProcess does in the past.. and found some things i don't like. Maybe now in Pharo, you can cosider about fixing/improving this stuff. http://lists.squeakfoundation.org/pipermail/squeak-dev/2007-July/119052.html http://lists.squeakfoundation.org/pipermail/squeak-dev/2007-July/119054.html 2009/4/16 Markus Fritsche <fritsche.markus(a)gmx.net>: > StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > >> Strange I fixed those ones. Can you check if you have the same problem >> in the latest? > > It could have happened that I did overwrite those by loading new PolyMorph > packages. > >> See the other email :) > > Merci! > >> Torsten sent us a loadFFI now available in scriptLoader in one the >> latest updates. >> Can you check? > > Seems to work well. Thanks for the pointer. > >> If you can reproduce that with the latest (updated image). >> Let us know. We are chasing bugs now. > > Nope, working now I think. > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project -- Best regards, Igor Stasenko AKA sig.

April 15, 2009