- Pharo-dev - lists.pharo.org

adding temporary adds them outside the temp declaration?
by StÃ©phane Ducasse Aug. 11, 2013

Aug. 11, 2013

Hi guys Is it a know bug? | selector superClasses | selector := aMethodDefinition selector. superClasses := aMethodDefinition realClass allSuperclasses. result := OrderedCollection new. (superClasses select: [ :eachClass | eachClass hasProtocolFor: selector]) do: [ :eachClass | (eachClass includesSelector: selector) ifTrue: [ result add: (self whichCategoryIncludesSelector: selector)]]. ^ result asBag adds result result | selector superClasses | selector := aMethodDefinition selector. superClasses := aMethodDefinition realClass allSuperclasses. result := OrderedCollection new. (superClasses select: [ :eachClass | eachClass hasProtocolFor: selector]) do: [ :eachClass | (eachClass includesSelector: selector) ifTrue: [ result add: (self whichCategoryIncludesSelector: selector)]]. ^ result asBag stef

2 2

latest VM seems stable
by StÃ©phane Ducasse Aug. 11, 2013

Aug. 11, 2013

Hi esteban I did a massive amount of loads (basically loading Moose in Pharo20 all day long) and I used exclusively the latest vm with guille changes on top of latest of Eliot. And I got no smalltinteger problems so far. So this sounds good. Stef

1 0

pattern for configuration definition
by StÃ©phane Ducasse Aug. 10, 2013

Aug. 10, 2013

Hi camillo I like the pattern damien told me you are applying for your configuration. Could you explain them? I understood: baseline48: spec <version: '4.8-baseline'> spec repository: 'http://www.smalltalkhub.com/mc/Moose/Moose/main'. â¦ spec project: 'Metanool' with: [ spec className: 'ConfigurationOfMetanool'; version: #development; file: 'ConfigurationOfMetanool'; repository: 'http://www.smalltalkhub.com/mc/Moose/Metanool/main' ]. â¦ => baseline48: spec <version: '4.8-baseline'> spec repository: self repositoryUrlString â¦ spec project: 'Metanool' with: [ spec className: 'ConfigurationOfMetanool'; version: #development; file: 'ConfigurationOfMetanool'; repository: 'http://www.smalltalkhub.com/mc/Moose/Metanool/main' ]. â¦ repositoryUrlString ^ 'http://www.smalltalkhub.com/mc/Moose/Moose/main'. specMetanoolConfiguration: spec spec className: 'ConfigurationOfMetanool'; version: #development; file: 'ConfigurationOfMetanool'; repository: 'http://www.smalltalkhub.com/mc/Moose/Metanool/main' I like the idea to avoid duplication at the configuration level. Now damien mention a dev and stable baseline Stef

3 3

Fwd: Keyboard events is broken.
by Igor Stasenko Aug. 9, 2013

Aug. 9, 2013

I changed this method to see better what happens: HandMorph class>>showDebugEvent: evt ShowEvents == true ifTrue: [ | ofs| Display fill: (0@0 extent: 500@120) rule: Form over fillColor: Color white. ofs := (owner hands indexOf: self) - 1 * 60. evt printString displayAt: (0@ofs) + (evt isKeyboard ifTrue: [0@30] ifFalse: [0@0]). self keyboardFocus printString displayAt: (0@ofs)+(0@45). evt isKeyboard ifTrue: [ Transcript show: evt printString;cr ] ]. KeyboardEvent>> printOn: aStream "Print the receiver on a stream" aStream nextPut: $[. aStream nextPutAll: type; nextPutAll: ' '''. self printKeyStringOn: aStream. aStream nextPut: $'. aStream space; nextPutAll: 'keyValue: ', self keyValue asString. aStream nextPut: $] set HandMorph showEvents:true Now, pressing single space, gives me this: [keyDown ' ' keyValue: 49] [keystroke ' ' keyValue: 32] [keyUp ' ' keyValue: 49] [keyUp ' ' keyValue: 49] Pressing delete key (or fn-backspace , for those who having a lot of spare fingers to use bad keyboards): [keyDown ' ' keyValue: 117] [keystroke 'â¦' keyValue: 188] [keyUp ' ' keyValue: 117] [keyUp ' ' keyValue: 117] now, can someone tell me , why there is 2 keyUp events for each keyDown? and of course, main question is why key values are different for keydown and keystroke events? (and why delete key is not 127, but 117 or 188?) Editor expecting 127: cmdMap at: (127 + 1) put: #forwardDelete:. "del key" -- Best regards, Igor Stasenko.

5 22

Fwd: Consequence of NBExternalStructure refactorings
by Igor Stasenko Aug. 9, 2013

Aug. 9, 2013

Hello, lately we changed the object format of subject class from variableByte to normal one. All projects which using NBExternalStructure can meet certain problems with it, and first which comes is that it may load code cleanly but in fact result in bogus variable subclasses of NBExternalStructure , while it can't because superclass has instance variable. (i suspecting classbuilder bug, but need to verify first). To address that, you need to manually go over all subclasses and double-check. Another thing: since now instances NBExternalStructure can also point to external memory, special care should be taken for #copying. Sending #copy to instance which uses external memory, will create a copy on object memory (and copy all data from external memory into it). This is same behavior as in NBExternalArray. - there is a bugs in #initialize and #postCopy (fixed but not intergated yet) https://pharo.fogbugz.com/default.asp?11395#83020 will do that in nearest time. also note, that in callouts, you cannot pass an instance of NBExternalStructure anymore to function, which uses raw pointer (like void*) anymore. For passing a pointer to structure data for callouts, use myStructure address idiom (or use right struct type, which is preferable). -- Best regards, Igor Stasenko.

3 5

[update 3.0] #30329
by Esteban Lorenzano Aug. 9, 2013

Aug. 9, 2013

1 0

Mariano & Behavior>>methodDict
by ClÃ©ment Bera Aug. 9, 2013

Aug. 9, 2013

Hello Mariano and others, In Pharo there is a method: Behavior>>methodDict methodDict == nil ifTrue: [^ MethodDictionary new ]. ^ methodDict It seems (according to the latest author) that it is Mariano who changed it from: Behavior>>methodDict methodDict == nil ifTrue: [self error: 'MethodDict is nil']. ^ methodDict A comment is present: "The method dictionary of a class can be nil when we want to use the #cannotInterpret: hook. Indeed when a class dictionary is nil, the VM sends the message cannotInterpret: to the receiver but starting the look up in the superclass of the class whose method dictionary was nil. Now the system relies that when the message methodDict is sent to a class a method dictionary is returned. In order to prevent the complaints of tools and IDE unaware of this feature, we fool them by providing an empty MethodDictionary. This will hopefully work in most cases, but the tools will loose the ability to modify the behavior of this behavior. The user of #cannotInterpret: should be aware of this." Now my problem is when I do: MyClass methodDict: nil. MyClass new name. It crashes the image in latest Pharo 3. Mariano, do you have an idea on how to fix it ? I ask you because you are the latest author and it seems tat the bug comes from this method. Thanks for any answer.

3 2

Aug. 8, 2013

1 0

Changes in the repositories for the VM
by Esteban Lorenzano Aug. 8, 2013

Aug. 8, 2013

Hi, Ok, we finally decided to make some changes longtime postponed, to improve our way of building the VM. Squeak guys are still using the old mechanism, so this just applies to Pharo and the continuous integration under the Pharo umbrella... but please notice that ANY vm flavor can be built with this source repository. The first change was to put all sources together. No more MCZ files to keep VMMaker and plugins and platform sources as a GIT repository. We are using filetree to keep source files, so is just like a different kind of monticello repository. With this, now a commit is atomic and the platform sources will be in sync with VMMaker sources. Also, that simplifies a lot the ConfigurationOfCog file, since the commit number is the version number... now we just need the baseline (not version anymore). So now you can checkout a version of the repository and do: ConfigurationOfCog loadGit. and that's all :) (from our point of view, this simplification is an huge advantage over the older process). The second change is just a relocation. Now the VM source lives in GitHub :) The reason of this move is because GitHub has a lot more functionality than the old Gitorious, also most of the people using filetree are already using GitHub so is easier for them to fork, etc. Also, it is hyped so we wanted to be in that wave ;) The new repository can be found at: https://github.com/pharo-project/pharovm Cheers, Esteban

2 1

[regression reporter]regression occurred
by no-reply＠ci.inria.fr Aug. 8, 2013

Aug. 8, 2013

https://ci.inria.fr/pharo/job/Pharo-3.0-Update-Step-2-Validation/label=win/… 7 regressions found. KernelTests.Classes.BehaviorTest.testBehaviorRespectsPolymorphismWithTraitBehavior KernelTests.Classes.MetaclassTest.testMetaclassAndTraitClassRespectsPolymorphismRules KernelTests.Numbers.LargePositiveIntegerTest.testMethodsOfTheClassShouldNotBeRepeatedInItsSuperclasses Refactoring.Tests.Environment.RBBrowserEnvironmentTest.testNotEnvironment Refactoring.Tests.Environment.RBBrowserEnvironmentTest.testOrEnvironment Refactoring.Tests.Environment.RBBrowserEnvironmentTest.testRemoveSelectorByAndAndNot Tests.Release.ReleaseTest.testInstanceSideMethodsWithNilKeyInLastLiteral

1 0