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

Re: [Pharo-project] Fuel bug: infinite recursion in "analyzer when:substituteBy:"

by Martin Dias

Hi Yanni This is a feature or a known issue, I am not sure yet! In serialization, when you substitute an object by another, such substitution(*) is then serialized using the same rules, so you might need the guard condition to avoid the infinite loop. This was the simpler implementation, but maybe it's undesirable or it's too error prone. Do you think we should change this? *: note the substitution can be not just an object but a graph of objects, some of them able to be substituted too. MartÃn On Fri, Nov 25, 2011 at 6:20 PM, Yanni Chiu <yanni(a)rogers.com> wrote: > 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 resultOfSerializeAndMaterializ**e: > (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: >

Nov. 25, 2011

Re: [Pharo-project] class rename and Fuel

by Yanni Chiu

On 25/11/11 3:56 PM, Yanni Chiu wrote: > > Any ideas why Fuel is looking for OldClass, yet there are no instances > of OldClass in the de-serialized object graph. There are no instances, but the class object itself is being held in various places in the Pier kernel.

Nov. 25, 2011

Re: [Pharo-project] How turn on Halo's?

by Henrik Sperre Johansen

On 25.11.2011 22:18, Helene Bilbo wrote: > Denis Kudriashov wrote >> Can I turn on halos. And is rotating morphs supported in pharo? >> > I love the rotating Morphs in Squeak! I found the functionality still in > Pharo, but only in RectangleMorph. But it seems to work fine to just copy > the methods from the t-rotating category of RectangleMorph back into Morph. > > So Browsers can be rotated again, while you type text into them :) > > I think itâs a very important feature, even if it is of no professional use. > really. > > Best regards, > Helene. > Yes, it's possible, see Transform/Transformation morphs which are still there. Be aware though, if invalidating the morph often enough (say, resizing without fastDrag), it will generate enough garbage through temporary bitmaps that the garbage collector gets stressed out and the image can "hangs". http://forum.world.st/Canvas-transform-bug-when-rendering-text-td1754208.ht… I imagine this will be much simpler and easier to do in a path-based Canvas like Athens. Cheers, Henry

Nov. 25, 2011

[Pharo-project] Issue Template Pharo 1.3 Enhancement? (was: Font lookup using styleName?)

by Bernhard Pieber

Hi everyone, I have implemented a test for the missing method and the method itself. Before I can upload it I need to create an issue for the enhancement. In the tracker I am missing a template for a Pharo 1.3 Enhancement. Would you mind creating one? Or is it a policy that if someone who wants to contribute an enhancement has to use the unstable version, Pharo 1.4 in this case? Cheers, Bernhard Am 22.11.2011 um 08:19 schrieb StÃ©phane Ducasse: > On Nov 21, 2011, at 9:15 PM, Bernhard Pieber wrote: > >> Hi everyone, >> >> I am using Pharo 1.3 and I am quite a newbie when it comes to Pharo so please be kindâ¦ > > we always are :) > Seriously in fact :) > >> I saw in the font chooser that Pharo supports font families with multiple styles. On my Mac for instance the font family Helvetica has the styles Light, Light Oblique, Regular, Oblique, Bold, and Bold Oblique. In the code font families are subinstances of FontFamilyAbstract and styles are subinstances of FontFamilyMemberAbstract. >> >> However, for creating fonts programmatically there seems to be no message which takes the style into account. There is LogicalFont class>>#familyName:pointSize:. How do I know which of the styles this uses? >> >> Then there is LogicalFont class>>#familyName:fallbackFamilyNames:pointSize:stretchValue:weightValue:slantValue:. How should I know which parameters I have to pass in to get a certain style? >> >> I would have expected a method like #familyName:styleName:pointSize: but I did not find one. Any idea why? Is this an omission or a feature? > > probably an omission but I will wait to learn if somebody has the answer > >> Cheers, >> Bernhard

Nov. 25, 2011

Re: [Pharo-project] Morphic Cleanup Status

by StÃ©phane Ducasse

On Nov 25, 2011, at 11:08 AM, Fernando Olivero wrote: > Hi Sean, i got overwhelmed last year with the task of porting CUIS > Morphic, to Pharo. There's just to much differences on several common > classes, it's hard to fully understand how to merge both ui > frameworks. > > Even though i believe Juan has a did an excellent work in simplifying > Morphic, and adding simpler code, like for example his new layouts > (download CUIS and try them out). > > But my overall impression , was that we need to stop being backward > compatible in order to create a simple Morphic-like framework for > Pharo, and start from scratch. Fixing current Morphic is also a > humongous task, for example see Stef's nice work on cleaning the > event handling mechanism. BTW you should be able to use the new event hierarchy: it is clean (no first secondâ¦..) and a nice support for double dispatch. I hope that next week I can sit with igor and we finish our journey. How do you deal with event in Gaucho. Stef > > Fernando > > On Fri, Nov 25, 2011 at 9:38 AM, Alain Plantec <alain.plantec(a)yahoo.com> wrote: >> Le 25/11/2011 07:39, Sean P. DeNigris a Ã©crit : >>> What's currently in progress? SimpleMorphic? CUIS related? Other? >>> >>> I was pulling my hair out with Morphic layouts yesterday and started to >>> clean, but I don't want to duplicate efforts on things already being worked >>> on. There are a bunch of open issues and discussions on the list. What's >>> actively being done right now? >> Hi Sean, >> AFAIK, no deep cleaning action. nothing from my side for now. >> I think they have done some refactoring around layout managing in CUIS >> (SimpleMorphic). >> btw, **thanks a lot** for all you are doing. >> Cheers >> Alain >> >>> >>> Thanks. >>> Sean >>> >>> -- >>> View this message in context: http://forum.world.st/Morphic-Cleanup-Status-tp4106566p4106566.html >>> Sent from the Pharo Smalltalk mailing list archive at Nabble.com. >>> >> >> >

Nov. 25, 2011

Re: [Pharo-project] deprecated: in Deprecated14

by StÃ©phane Ducasse

+ 1 Given a date and a version is a nice way to educate people to traceability. > Marcus Denker-4 wrote >> >> I think the idea was to deprecate deprecated: so people use the one with >> more >> infos... (deprecated:on:in:) >> >> But for small external projects, deprecated: is just fine and there are >> users. So >> I suggest we just move it out of the deprecated package and have it as a >> simple >> to use version of deprecation tagging. >> > > It was quite a bit of work to clean the system of calls to #deprecated: > http://code.google.com/p/pharo/issues/detail?id=4718 > > If it's an important feature of other projects (I don't see the value over > the more informative version; who has a use case?), maybe the best thing is > for it to be loadable but not included by default? If it's left in the > image, I'm sure habit and inertia will have people continue to use it in the > core image, which is not what we want. > > Sean > > -- > View this message in context: http://forum.world.st/deprecated-in-Deprecated14-tp4107546p4108889.html > Sent from the Pharo Smalltalk mailing list archive at Nabble.com. >

Nov. 25, 2011

[Pharo-project] Fuel bug: infinite recursion in "analyzer when:substituteBy:"

by Yanni Chiu

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:

Nov. 25, 2011

Re: [Pharo-project] ProtoObject basicNew :)

by StÃ©phane Ducasse

On Nov 25, 2011, at 10:09 PM, Toon Verwaest wrote: > Ok, I guess I was confused by your initial question: "the methods that I should define in a subclass of ProtoObject so that I can use basicNew"... :) :) Yes it is more the methods that I should define to have a usable class (browse, doit, inspect it). I will probably have to do it by try and error. Stef > > On 11/25/2011 10:05 PM, StÃ©phane Ducasse wrote: >> I'm confused. >> I'm not creating a class but an instance. >> if it would be Behavior basicNew then yes I made sure that we could do Behavior new and not crashing the system back in Squeak. >> >> On Nov 25, 2011, at 9:45 PM, Toon Verwaest wrote: >> >>> With the magic number I mean the number that defines the object's structure. It's one of the first 3 instance variables of every class, encoding how many instance variables there are; what kind of object it is supposed to be (normal, array, bytes, words...). Every "class" basically needs 3 instance variables. 1 defines the structure, 1 the superclass, 1 the behavior (method dictionary). I don't know the other; but just look at some class to know what the order is. >> Yes format, methodDict, superclass (probably in a different order). >> And format is ugly because not explicit and coded in several place like instSpec (berk) >> >>> And yes, if you want to inspect the object, you'll need to support whatever protocol your inspector is using. >>> >>> The format is formed as follows: >>> >>> format >>> | fieldSize sizeHiBits format | >>> fieldSize := self fieldSize + 1. >>> sizeHiBits := fieldSize // 64. >>> format := sizeHiBits. >>> format := (format bitShift: 5) + compactClassIndex. >>> format := (format bitShift: 4) + self instanceSpecification. >>> format := (format bitShift: 6) + (fieldSize \\ 64). >>> format := (format bitShift: 1). >>> ^ format >>> >>> If I read it correctly (just glancing), instance specification specifies if it's arrayed and if it has fields. 0 -> no fields, not arrayed, 1 -> fields, not arrayed, 2 -> arrayed no fields, 3 -> arrayed fields. >>> >>> >>> HTH, >>> Toon >>> >>> On 11/25/2011 09:31 PM, StÃ©phane Ducasse wrote: >>>> Apparently I can do a doit I was naively doing an inspect ;) >>>> >>>> >>>>> Don't you just need the basicNew primitive, and make sure that the "magic number" is set in the object that has the basicNew installed to something that makes sense? >>>>> >>>> I do not know what is the magic number. >>>> I guess that my problem was that I have first to define what is the minimal behavior. >>>> >>>> >>>>> If you want to send messages to the object, probably also best install some dictionary as the method dict instvar. >>>>> >>>> But I guess that ProtoObject has some behavior since I can browse the class with a browser. >>>> >>>> >>>> >>>>> Toon >>>>> >>>>> On 11/25/2011 09:26 PM, StÃ©phane Ducasse wrote: >>>>> >>>>>> Hi >>>>>> I would like to know the methods that I should define in a subclass of ProtoObject so that I can use basicNew. >>>>>> Does any of you have an idea? >>>>>> >>>>>> Stef >>>>>> >> > >

Nov. 25, 2011

Re: [Pharo-project] deprecated: in Deprecated14

by Sean P. DeNigris

Marcus Denker-4 wrote > > I think the idea was to deprecate deprecated: so people use the one with > more > infos... (deprecated:on:in:) > > But for small external projects, deprecated: is just fine and there are > users. So > I suggest we just move it out of the deprecated package and have it as a > simple > to use version of deprecation tagging. > It was quite a bit of work to clean the system of calls to #deprecated: http://code.google.com/p/pharo/issues/detail?id=4718 If it's an important feature of other projects (I don't see the value over the more informative version; who has a use case?), maybe the best thing is for it to be loadable but not included by default? If it's left in the image, I'm sure habit and inertia will have people continue to use it in the core image, which is not what we want. Sean -- View this message in context: http://forum.world.st/deprecated-in-Deprecated14-tp4107546p4108889.html Sent from the Pharo Smalltalk mailing list archive at Nabble.com.

Nov. 25, 2011

Re: [Pharo-project] How turn on Halo's?

by Helene Bilbo

Denis Kudriashov wrote > > Can I turn on halos. And is rotating morphs supported in pharo? > I love the rotating Morphs in Squeak! I found the functionality still in Pharo, but only in RectangleMorph. But it seems to work fine to just copy the methods from the t-rotating category of RectangleMorph back into Morph. So Browsers can be rotated again, while you type text into them :) I think itâs a very important feature, even if it is of no professional use. really. Best regards, Helene. -- View this message in context: http://forum.world.st/How-turn-on-Halo-s-tp2227309p4108883.html Sent from the Pharo Smalltalk mailing list archive at Nabble.com.

Nov. 25, 2011