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

[Pharo-project] Fuel bug: infinite recursion in "analyzer when:substituteBy:"
by Yanni Chiu Nov. 27, 2011

Nov. 27, 2011

The attached test case, reproduced below: | result | self analyzer when: [:x | FLPair = x class and: [x left isNil not] ] substituteBy: [:x | x copy left: $A ]. result := self resultOfSerializeAndMaterialize: (FLPair new left: $a; right: $b; yourself). self assert: result left = $A. causes an infinite loop with IdentityDictionary (see stack trace at end). The similar test case #testTransientPairLeft, works because the guard condition is negated by the replacement (i.e. "x left isNil not" is made false by the replacement). The attached test case does not make "x left isNil not" false. It causes an infinite recursion, so use "ctrl-." to stop it before too long. Is this a bug, or the intention of the #when:substituteBy: feature? I've worked around it, by making sure to always have the substituted value fail the guard clause. -- Yanni ==== stack trace ==== IdentityDictionary>>scanFor: IdentityDictionary(HashedCollection)>>findElementOrNil: IdentityDictionary(Dictionary)>>at:ifAbsent: FLPluggableSubstitutionCluster(FLSubstitutionCluster)>>add:traceWith: FLAnalysis>>mapAndTrace: FLAnalysis>>run [:anObject | (FLAnalysis newWith: self firstInMapperChain root: anObject) run; yourself] in FLAnalyzer>>setDefaultAnalysis FLAnalyzer>>analysisFor: FLSerialization>>analysisStep FLSerialization>>run [:anObject :aStream | (FLSerialization root: anObject on: aStream analyzer: self analyzer) run; yourself] in FLSerializer>>defaultSerialization [self nextPutSignatureOn: aStream. self nextPutVersionOn: aStream. ^ self serializationFactory value: anObject value: aStream] in FLSerializer>>serialize:on: BlockClosure>>ensure: FLSerializer>>serialize:on: [:aStream | self serializer serialize: anObject on: aStream] in FLPluggableSubstitutionTest(FLSerializationTest)>>serialize: [:aStream | aValuable value: aStream binary] in FLFileStreamStrategy>>writeStreamDo: [anotherBlock value: file] in MultiByteFileStream class(FileStream class)>>detectFile:do: BlockClosure>>ensure: MultiByteFileStream class(FileStream class)>>detectFile:do: MultiByteFileStream class(FileStream class)>>forceNewFileNamed:do:

3 5

[Pharo-project] class rename and Fuel
by Yanni Chiu Nov. 27, 2011

Nov. 27, 2011

Here's my scenario: - a Pier kernel was previously serialized with Fuel-1.7 - a class was re-named (happens to be a subclass of PRCommand) - get a walkback when trying to de-serialize with the same Fuel-1.7 - define the missing class, then de-serialize works I was hoping that, after serializing the Pier kernel again (with both classes present, to be able to load), then on the next re-start, the old class would no longer be needed. That did not work - same problem of FLClassNotFound. To debug this, I de-serialized in a development image. To my surprise, evaluating "OldClass allInstances" found 0 instances of the OldClass. So I can't figure out how to track down why Fuel is still looking for that old class. BTW, since Pier looks for all subclasses of PRCommand, I made the NewClass a subclass of Object. That did not fix the problem. Any ideas why Fuel is looking for OldClass, yet there are no instances of OldClass in the de-serialized object graph. The stack trace is included at the end. -- Yanni ===== stack trace ===== [FLClassNotFound new signal: 'Class ' , className , ' not found at materialization'] in FLGlobalClassCluster(FLGlobalCluster)>>materializeGlobalClassFrom: SystemDictionary(Dictionary)>>at:ifAbsent: FLGlobalClassCluster(FLGlobalCluster)>>materializeGlobalClassFrom: FLGlobalClassCluster>>materializeFrom: FLGlobalClassCluster(FLPrimitiveCluster)>>materializeInstanceWith: [objects add: (self materializeInstanceWith: aMaterialization)] in FLGlobalClassCluster(FLIteratingCluster)>>materializeInstancesStepWith: SmallInteger(Integer)>>timesRepeat: FLGlobalClassCluster(FLIteratingCluster)>>materializeInstancesStepWith: FLMaterialization>>clusterInstancesStep [self clusterInstancesStep] in FLMaterialization>>instancesStep SmallInteger(Integer)>>timesRepeat: FLMaterialization>>instancesStep FLMaterialization>>run [:aStream | (FLMaterialization on: aStream) run; yourself] in FLMaterializer>>setDefaultMaterialization [self verifySignatureFrom: aStream. self verifyVersionFrom: aStream. ^ materializationFactory value: aStream] in FLMaterializer>>materializeFrom: BlockClosure>>ensure: FLMaterializer>>materializeFrom: [:aFileStream | self newDefault materializeFrom: aFileStream binary] in FLMaterializer class>>materializationFromFileNamed: [anotherBlock value: file] in StandardFileStream class(FileStream class)>>detectFile:do: BlockClosure>>ensure: StandardFileStream class(FileStream class)>>detectFile:do: StandardFileStream class(FileStream class)>>oldFileNamed:do: FLMaterializer class>>materializationFromFileNamed: UndefinedObject>>DoIt

3 7

[Pharo-project] gettext implementation
by CÃ©dric Ternon Nov. 26, 2011

Nov. 26, 2011

Hi, I'm trying to use the gettexts packages under seaside, I firstly posted to the seaside mailingList but Philippe Marschallrecommended me to try to reach mister Hilaire on this ml. I have success to use the gettexts packages creating a .pot file, translate the website in poedit, creating the .mo and having the website in different langages. The problem occurs when I try to create a .po file to continue an incomplete translation, The previous messages follows : Hi, I'm trying to use the gettexts packages for seaSide. I finally achieve to create a .pot file, translate it with poedit and make the website translated. But now, I'm trying to create a .po file, that is a .pot file with the strings already translated and the strings remaining to translate. (a .pot file beeing an empty template to begin without any translations) We followed this tutorial : http://code.google.com/p/seaside/wiki/Gettext The versions used are : Seaside-Gettext-Core-NorbertHartl.10 Seaside-Pharo-Gettext-Core-pmm.1 Name: Gettext-HilaireFernandes.22 When I simply try : GetTextExporter new exportTranslator: NaturalLanguageTranslator current. I got : http://pix.toile-libre.org/?img=1322152172.jpg I've got the same result with any GetTextTranslator I have tryed to look for an implementors of the method "translations" but no one is implementing it, is a pkged missing? I've also tryed to replace it with 'class translators' but this sound wrong and give me another unrelated error. Thanks in advance for your help, ----- Philippe answer : It looks like you're using GetTextExporter instead of WAGetTextExporter. Can you load Seaside-Gettext-Examples and evaluate WAGettextExample export and see if that works? Cheers Philippe ----- Hello, Thanks to help. It works fine but it creates a .pot file. My goal is to create a .po file : when a site is already half translated, I'd like to create a .po file a translator can use in poedit. So the translator only has to translate the half not yet translated (and maybe bring some little corrections to the half already translated.) WAGettextExample export calls WAGetTextExporter new exportTemplate wich calls self exportTranslator: nil and we go there exportTranslator: translator "Export translation files. the file extention is 'po', or 'pot' if translator is nil " when the translator is nil, everything works well. when I try with a translator it breaks here : translationFor: aKey in: translator translator ifNil: [^'']. translator translations at: aKey ifPresent: [:s | ^s]. "If we have old camelCase translations, make space-separated words" translator translations at: aKey toCamelCase ifPresent: [:s | (s includes: Character space) ifTrue: [^s]. ^s fromCamelCase]. ^'' --------------------------- Now, I have tried to change the code of translationFor: aKey in: translator in this way : translationFor: aKey in: translator |result| translator ifNil: [^'']. result := (WALocalizationContext new translate: 'aKey'). (aKey equal result) ifTrue: [^'']. ^result. That is maybe horrible to use the WA things at this level but I tried to make it works and the only way I achieve to translate things is at the WA level.

2 3

[Pharo-project] the screenshots
by CÃ©dric Ternon Nov. 26, 2011

Nov. 26, 2011

http://pix.toile-libre.org/?img=1322339454.png http://pix.toile-libre.org/?img=1322339771.png

1 0

[Pharo-project] Added to OmniBrowser info in online book
by Sean P. DeNigris Nov. 26, 2011

Nov. 26, 2011

http://book.pharo-project.org/book/PharoTools/OmniBrowser Anyone have anything to add? -- View this message in context: http://forum.world.st/Added-to-OmniBrowser-info-in-online-book-tp4110664p41… Sent from the Pharo Smalltalk mailing list archive at Nabble.com.

3 3

Re: [Pharo-project] Morphic replacement
by Fernando Olivero Nov. 26, 2011

Nov. 26, 2011

Glad to be able to say: yes! Each object speaks in its own coordinate system, its parent already internalized the point. Suppose we have a click, announced by the mouse at a given point, given in Display coordinates. >From now on we go, from the display, to the view , to the scene, to the shape, and on each step the parent deals with a point in its own coordinate system, and passes on the internalized click to its children. Views present a scene, consisting of shapes, at a given viewport of the display. Makes it a lot easier when you are programming the interaction of custom widgets, to always deal with local coordinates, and greatly simplifies the event handling code as well. On Fri, Nov 25, 2011 at 5:00 PM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > did you kill the absolute coordinates? > > say yes please :) > > Stef > > On Nov 25, 2011, at 11:00 AM, Fernando Olivero wrote: > >> Dear all, >> >> I've been working on a hole replacement of Morphic myself, because if >> was tired of fixing Morphic (or porting CUIS Morphic into Pharo), and >> found much more productive the approach of rewriting it from scratch. > > excellent! > >> >> In my expierience, i was 100% more productive by not trying to be >> backward compatible with Morphic, Polymorph , etc. >> >> I found it easier to Â come up with a new set of widgets (EntryField, >> AutoCompletionEntryField, Button, Label, Column, Row, >> ScrollableColumns) from scratch. They are all fairly tested also. >> >> Maybe it's time to move on, and develop from the ground up a >> Morphic-like successor, instead of fixing the current Morphic. There's >> no problem with preserving the current UI for all the apps that are >> out there, and provide a cleaner, simpler UI framework for the new >> apps. >> >> Regarding the new "morphic", I only assume that i have a scalable >> canvas, and the vm's notifying me of the user events. On top of that i >> recreated a World, that can have multiple views. There's only one >> World at a given time. Views are presented within the world in a given >> viewport, and present a scene. Scene are were the shapes live in. >> No World as a subclass of PasteUpMorph, no globals such as World, >> ActiveWorld, Hand. ActiveHand, etcs.. >> >> The world has a mouse and a keyboard, which take care of events >> announced by the world. For example GMouse>>down: aDisplayPosition, >> would take care of transforming the display position into a world >> position, into a view position, into a scene position, and finally >> notifying the proper shape under the mouse that it has been clicked at >> a certain local position. GShape>>clickedAt: , >> GShape>>secondaryClickedAt:, etc.. >> No Events hierarchy, or the awful switch currently in HandMorph ( >> generateMouseEvent: or generateKeyboardEvent:). >> >> I have a "Gaucho" system working now, Â with GShape (instead of >> Morphs), using the simple new layouts developed by Juan in CUIS, >> uniform usage of keybindings across the whole GShape hierarchy, >> amongst other goodies. It's still tailored to be a Morphic replacement >> for Gaucho, so first i need to polish a few things prior to releasing >> it. >> >> For example, i'm going to ditch my current scalable canvas >> implementation, in favor of Athens. >> >> Hope that very soon i can provide more details, but my plans are to >> release it as part of the next Gaucho version, to provide a working >> example. >> >> Fernando >> > >

4 3

[Pharo-project] Added brief OSProcess info to online book
by Sean P. DeNigris Nov. 26, 2011

Nov. 26, 2011

http://book.pharo-project.org/book/PharoTools/OSProcess -- View this message in context: http://forum.world.st/Added-brief-OSProcess-info-to-online-book-tp4110571p4… Sent from the Pharo Smalltalk mailing list archive at Nabble.com.

2 1

[Pharo-project] Accessing the OS Environment Variables
by Alberto Bacchelli Nov. 26, 2011

Nov. 26, 2011

Hello, How can I access the OS Environment Variables from Pharo? Thanks, Alberto -- View this message in context: http://forum.world.st/Accessing-the-OS-Environment-Variables-tp4024106p4024… Sent from the Pharo Smalltalk mailing list archive at Nabble.com.

3 4

[Pharo-project] RB API
by Sean P. DeNigris Nov. 26, 2011

Nov. 26, 2011

Is there any documentation on using AST & RB? I didn't find anything enlightening via google, the tests, or the class/method comments. In trying to replace all calls to #deprecate: and #deprecate:in: to #deprecate:on:in: (Issue 4718), it took me hours to step through OB's refactoring and learn the syntax. In case anyone else can benefit, I eventually ended up with: RBParseTreeRewriter new replace: '``@object deprecated: ``@arg1 ' with: '``@object deprecated: ``@arg1 on:(''', dateString, ''') in:(''Pharo', versionNumber asString, ''')'. Sean -- View this message in context: http://forum.world.st/RB-API-tp3781019p3781019.html Sent from the Pharo Smalltalk mailing list archive at Nabble.com.

4 12

[Pharo-project] INRIA Postdoc position: security in dynamic languages
by stephane ducasse Nov. 26, 2011

Nov. 26, 2011

Postdoc Position â¢ Duration: 3 years â¢ Starting date: end 2011 â¢ Where: RMOD team at INRIA Lille Nord Europe http://rmod.lille.inria.fr â¢ Salary: 55 Keuros net (before taxes) Keywords: dynamic languages, security, information flow, Smalltalk, Javascript The position will take place in the context of the Systematic Resilience project. We will focus on the following main tasks: â¢ Identification and understanding of security problems in Javascript and other dynamic languages â¢ Study and analyze existing approaches towards securing Javascript (Caja, isolates...) â¢ Comparison of different approaches to secure dynamic applications (Zope model, ZTK), Ruby â¢ Study of Mirrors/Capabilities as a basis for security in Javascript â¢ Proposition of an safe architecture for Resilience Please contact Stephane Ducasse (Stephane.Ducasse(a)inria.fr) for more information. http://rmod.lille.inria.fr/ http://stephane.ducasse.free.fr

1 0