September 2011 - Pharo-dev - lists.pharo.org

Re: [Pharo-project] Zinc crashing image at start up

by Lukas Renggli

> I have now had two images rendered unusable because Zinc crashes them at > start up. Debug log attached. +1, I had a similar problem at some point from an image built on Jenkins. Lukas -- Lukas Renggli www.lukas-renggli.ch

Sept. 2, 2011

[Pharo-project] WriteStream strangeness

by Philippe Marschall

Hi I have run into two issues when #reset-ing WriteStreams. The first one is that #reset on WriteStream and ReadWriteStream don't have the same semantics [1] (although ReadWriteStream is a subclass of WriteStream). (WriteStream on: String new) nextPutAll: 'the content'; reset; contents gives you '' but (ReadWriteStream on: String new) nextPutAll: 'the content'; reset; contents gives you 'the content' The second is that #reset doesn't affect the answer of #size [2]. WriteStream >> #size has no comment (!) so my naÃ¯ve assumption of its semantics are: aStream size = aStream contents size However the following test fails: | stream | stream := String new writeStream. self assert: stream size = stream contents size. stream nextPutAll: 'abc'; reset. self assert: stream size = stream contents size [1] http://code.google.com/p/pharo/issues/detail?id=4706 [2] http://code.google.com/p/pharo/issues/detail?id=4707 Cheers Philippe

Sept. 2, 2011

[Pharo-project] Zinc crashing image at start up

by Philippe Marschall

Hi I have now had two images rendered unusable because Zinc crashes them at start up. Debug log attached. Cheers Philippe

Sept. 2, 2011

Re: [Pharo-project] [Pharo-users] [COTDC] 78 - RBSubclassResponsibilityNotDefinedRule

by Lukas Renggli

> For the RBRule what would be great (and this was on my overly long todo) was a improve the description. > May be we should have a short and long one. Because something I do not understand (well I do most of the times) but > I would love to get a nice description because we really want to push the rules usage. Longer descriptions for most rules can be found here: http://www.refactory.com/RefactoringBrowser/BrowserPagesLintChecks.html The #rational should remain a one sentence description for the UI. > Â Â Â Â - get rules automatically run with jenkins I've implemented that a long time ago with the rest of the build system. For example, on the Seaside build (http://jenkins.lukas-renggli.ch/job/Seaside%203.0/) there is a graph that shows the change in code critics issues over time ("Checkstyle Trend"). If you click on the graph or on "Checkstyle Warnings" (http://jenkins.lukas-renggli.ch/job/Seaside%203.0/715/checkstyleResult/?) you get detailed reports about all issues and how they change between the builds. The only thing to do is to tell Jenkins to run the report. The syntax is exactly the same as with running tests, except that you use HDLintReport instead of HDTestReport. It is all described in the README text of the Jenkins scripts. Lukas -- Lukas Renggli www.lukas-renggli.ch

Sept. 2, 2011

Re: [Pharo-project] [Pharo-users] [COTDC] 78 - RBSubclassResponsibilityNotDefinedRule

by StÃ©phane Ducasse

could be interesting. But it would just hides our poor comment quality. I prefer to see that I'm poor than pretend I'm rich :) For the RBRule what would be great (and this was on my overly long todo) was a improve the description. May be we should have a short and long one. Because something I do not understand (well I do most of the times) but I would love to get a nice description because we really want to push the rules usage. We got a postdoc with chili (yesterday - hello and welcome angel :) ) and we want him to help building better infrastructure around rules - book chapter on rule engine - storing metadata because it does not scale to add pragmas in method (we probably need a package manifesto). - analyzing the rules and their application to pharo and pharo projects - get rules automatically run with jenkins Stef On Sep 2, 2011, at 8:27 AM, laurent laffont wrote: > On Thu, Sep 1, 2011 at 10:08 PM, Lukas Renggli <renggli(a)gmail.com> wrote: > >> Today: RBSubclassResponsibilityNotDefinedRule > > > > This rules identifies when several methods in a set of subclasses are defined but not in the superclass. > > It identifies that an abstract method is missing in the superclass. > > Checks that all subclassResponsibility methods are defined in all > leaf classes. > > Note that each of the 120 subclasses of RBLintRule has a method called > #rationale that describes its intent. There are even tests that verify > this. Of course the string is not a class comment, but something to be > displayed in the code critics browser. Still I note a certain degree > of overlap ... > > > I wonder, we may just comment RBLintRule. Then wouldn't it be nice that class browser, when displaying a class without comment, goes up in the hierarchy up to Object (excluded) to get the first comment found ? > > > Laurent. > > > > Lukas > > -- > Lukas Renggli > www.lukas-renggli.ch > >

Sept. 2, 2011

Re: [Pharo-project] Testing Pharo OneClick 1.3

by StÃ©phane Ducasse

On Sep 2, 2011, at 7:16 AM, HernÃ¡n Morales Durand wrote: > Two more: > > ScriptLoader new installingInstaller should not work and should not be used. Nobody should use ScriptLoader :) > MNU possibly related with ZnNetworkingUtils url object. > > -Enable underscore assignment (some packages still needs this setting > for loading) > -Evaluate: > > | source mNode | > source := 'initialize > | array start stop | > array _ Array new: 100. > start _ 0. > stop _ 0.'. > mNode := UndefinedObject evaluatorClass new > compile: source > in: UndefinedObject > classified: nil > notifying: nil > ifFail: []. > SyntaxError open: ( SyntaxError new contents: mNode sourceText ). > > Hit right-button in the "UndefinedObject <none>" line in the > SyntaxError windows. > Select "Browse full" > MNU > > 2011/9/1 Damien Cassou <damien.cassou(a)gmail.com>: >> http://code.google.com/p/pharo/issues/detail?id=4751 >> >> On Thu, Sep 1, 2011 at 6:04 PM, HernÃ¡n Morales Durand >> <hernan.morales(a)gmail.com> wrote: >>> Generate #= and #hash >> >> >> >> -- >> Damien Cassou >> http://damiencassou.seasidehosting.st >> >> "Lambdas are relegated to relative obscurity until Java makes them >> popular by not having them." James Iry >> >> >

Sept. 2, 2011

Re: [Pharo-project] [Pharo-users] [COTDC] 78 - RBSubclassResponsibilityNotDefinedRule

by laurent laffont

On Thu, Sep 1, 2011 at 10:08 PM, Lukas Renggli <renggli(a)gmail.com> wrote: > >> Today: RBSubclassResponsibilityNotDefinedRule > > > > This rules identifies when several methods in a set of subclasses are > defined but not in the superclass. > > It identifies that an abstract method is missing in the superclass. > > Checks that all subclassResponsibility methods are defined in all > leaf classes. > > Note that each of the 120 subclasses of RBLintRule has a method called > #rationale that describes its intent. There are even tests that verify > this. Of course the string is not a class comment, but something to be > displayed in the code critics browser. Still I note a certain degree > of overlap ... > I wonder, we may just comment RBLintRule. Then wouldn't it be nice that class browser, when displaying a class without comment, goes up in the hierarchy up to Object (excluded) to get the first comment found ? Laurent. > > Lukas > > -- > Lukas Renggli > www.lukas-renggli.ch > >

Sept. 2, 2011

Re: [Pharo-project] Testing Pharo OneClick 1.3

by HernÃ¡n Morales Durand

Two more: ScriptLoader new installingInstaller MNU possibly related with ZnNetworkingUtils url object. -Enable underscore assignment (some packages still needs this setting for loading) -Evaluate: | source mNode | source := 'initialize | array start stop | array _ Array new: 100. start _ 0. stop _ 0.'. mNode := UndefinedObject evaluatorClass new compile: source in: UndefinedObject classified: nil notifying: nil ifFail: []. SyntaxError open: ( SyntaxError new contents: mNode sourceText ). Hit right-button in the "UndefinedObject <none>" line in the SyntaxError windows. Select "Browse full" MNU 2011/9/1 Damien Cassou <damien.cassou(a)gmail.com>: > http://code.google.com/p/pharo/issues/detail?id=4751 > > On Thu, Sep 1, 2011 at 6:04 PM, HernÃ¡n Morales Durand > <hernan.morales(a)gmail.com> wrote: >> Generate #= and #hash > > > > -- > Damien Cassou > http://damiencassou.seasidehosting.st > > "Lambdas are relegated to relative obscurity until Java makes them > popular by not having them." James Iry > >

Sept. 2, 2011

Re: [Pharo-project] [squeak-dev] Contexts must only be created with newForMethod: ...

by Eliot Miranda

On Thu, Sep 1, 2011 at 10:03 AM, Mariano Martinez Peck < marianopeck(a)gmail.com> wrote: > Since I didn't get feedback there I ask here....what do you think? I agree > with Chris. > you might write it as ContextPart class >> newFromFrameSize: aFrameSize self assert: (aFrameSize = SmallFrame or: [aFrameSize = LargeFrame]). ^ super basicNew: aFrameSize or ContextPart class >> newFromFrameSize: aFrameSize ^ super basicNew: (aFrameSize > SmallFrame ifTrue: [LargeFrame] ifFalse: [SmallFrame]) either after moving SmallFrame and LargeFrame to a shared pool, or by adding an accessor for LargeFrame in CompiledMethod class (there's already an accessor for SmallFrame, CompiledMethod class>>smallFrameSize). (i.e. : ContextPart class >> newFromFrameSize: aFrameSize self assert: (aFrameSize = CompiledMethod largeFrameSize or: [aFrameSize = CompiledMethod smallFrameSize]). ^ super basicNew: aFrameSize or ContextPart class >> newFromFrameSize: aFrameSize ^ super basicNew: (aFrameSize > CompiledMethod smallFrameSize ifTrue: [CompiledMethod largeFrameSize] ifFalse: [CompiledMethod smallFrameSize ]) > Thanks > > On Mon, Aug 1, 2011 at 8:34 PM, Mariano Martinez Peck < > marianopeck(a)gmail.com> wrote: > >> Hi Chris. Yes, I have this problem when materializing MethodContext >> instances in Fuel. >> I've solved it adding the method: >> >> ContextPart class >> newFromFrameSize: aFrameSize >> >> ^ super basicNew: aFrameSize >> >> but it is a hack. >> >> So yes, if there is no real reason to do the opposite, I would change >> #basicNew: >> >> Cheers >> >> >> On Mon, Aug 1, 2011 at 8:27 PM, Chris Muller <ma.chris.m(a)gmail.com>wrote: >> >>> .. but #newForMethod: requires more in the API than it really uses. >>> It only checks the methods #frameSize to determine what basicSized >>> MethodContext to create - it doesn't actually set its method from the >>> one passed in. >>> >>> newForMethod: aMethod >>> --- (comment snipped for brevity) --- >>> ^ super basicNew: aMethod frameSize >>> >>> So, would it be ok to relax MethodContext class>>basicNew: to something >>> like: >>> >>> basicNew: size >>> ^ (size = CompiledMethod smallFrameSize or: [ size = >>> CompiledMethod fullframeSize ]) >>> ifTrue: [ super basicNew: size ] >>> ifFalse: [ self error: 'Contexts must be ', >>> CompiledMethod smallFrameSize, ' or ', CompiledMethod fullframeSize, ' >>> bytes.' ] >>> >>> >> >> >> -- >> Mariano >> http://marianopeck.wordpress.com >> >> > > > -- > Mariano > http://marianopeck.wordpress.com > > -- best, Eliot

Sept. 2, 2011

Re: [Pharo-project] [Pharo-users] [COTDC] 78 - RBSubclassResponsibilityNotDefinedRule

by Lukas Renggli

>> Today: Â RBSubclassResponsibilityNotDefinedRule > > This rules identifies when several methods in a set of subclasses are defined but not in the superclass. > It identifies that an abstract method is missing in the superclass. Checks that all subclassResponsibility methods are defined in all leaf classes. Note that each of the 120 subclasses of RBLintRule has a method called #rationale that describes its intent. There are even tests that verify this. Of course the string is not a class comment, but something to be displayed in the code critics browser. Still I note a certain degree of overlap ... Lukas -- Lukas Renggli www.lukas-renggli.ch

Sept. 1, 2011