August 2013 - Pharo-dev - lists.pharo.org

get an endless loop trying to get a menu on AbstractNautilusUI

by StÃ©phane Ducasse

buildAnalyzeSubMenu: aBuilder <contextMenu> <nautilusGlobalClassMenu> | target | target := aBuilder model. target selectedClass ifNil: [^target]. target selectedClass buildAnalyzeSubMenu: aBuilder. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Now I do not get much how to fix it (too tired) Stef

Aug. 29, 2013

Re: [Pharo-dev] [Moose-dev] Some FAMIX questions

by Tudor Girba

+100 Doru On Thu, Aug 29, 2013 at 9:58 PM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr > wrote: > nice analysis. > I really hope that using smarter class structure based on Slot we can > offer a solution in the future. > > > On Aug 29, 2013, at 6:24 PM, Stephan Eggermont <stephan(a)stack.nl> wrote: > > > Currently, quite a lot of space seem to be taken up by not-filled > instvars. > > In my Delphi based Whitestar model, 406085 out of 664371 > > are either nil or empty sets. In a JHotDraw model from inFusion > > it is 724408 out of 1351527 > > > > Stephan > > > > > > FAMeasurements>measure > > varSize := mClass instSize. > > instSize := mClass allInstances size. > > nrNil := 0. > > nrEmpty := 0. > > mClass allInstancesDo: [ :instance | > > 1 to: varSize do: [:varIndex | |value| > > value := instance instVarAt: varIndex. > > value ifNil: [ nrNil := nrNil+1 ]. > > value isCollection ifTrue: [ > > value ifEmpty: [ nrEmpty := nrEmpty+1 ] ] > ] ] > > > > MeasureFAMIX>initializeMeasurements > > FAMIXEntity withAllSubclassesDo: [ :each | > > self measurements add: (FAMeasurements new > > mClass: each; > > yourself). ] > > > > MeasureFAMIX>run > > self initializeMeasurements. > > measurements do: [ :each | each measure] > > > > MeasureFAMIX>sum > > nrOfVars := 0. > > nrEmpty := 0. > > measurements do: [ :measurement | > > nrOfVars := nrOfVars + (measurement varSize * measurement > instSize). > > nrEmpty := nrEmpty + measurement nrEmpty + measurement > nrNil]. > > > > In Famix-SelfMeasurements on smalltalkhub > > _______________________________________________ > > Moose-dev mailing list > > Moose-dev(a)iam.unibe.ch > > https://www.iam.unibe.ch/mailman/listinfo/moose-dev > > > -- www.tudorgirba.com "Every thing has its own flow"

Aug. 29, 2013

Re: [Pharo-dev] [Moose-dev] Some FAMIX questions

by StÃ©phane Ducasse

nice analysis. I really hope that using smarter class structure based on Slot we can offer a solution in the future. On Aug 29, 2013, at 6:24 PM, Stephan Eggermont <stephan(a)stack.nl> wrote: > Currently, quite a lot of space seem to be taken up by not-filled instvars. > In my Delphi based Whitestar model, 406085 out of 664371 > are either nil or empty sets. In a JHotDraw model from inFusion > it is 724408 out of 1351527 > > Stephan > > > FAMeasurements>measure > varSize := mClass instSize. > instSize := mClass allInstances size. > nrNil := 0. > nrEmpty := 0. > mClass allInstancesDo: [ :instance | > 1 to: varSize do: [:varIndex | |value| > value := instance instVarAt: varIndex. > value ifNil: [ nrNil := nrNil+1 ]. > value isCollection ifTrue: [ > value ifEmpty: [ nrEmpty := nrEmpty+1 ] ] ] ] > > MeasureFAMIX>initializeMeasurements > FAMIXEntity withAllSubclassesDo: [ :each | > self measurements add: (FAMeasurements new > mClass: each; > yourself). ] > > MeasureFAMIX>run > self initializeMeasurements. > measurements do: [ :each | each measure] > > MeasureFAMIX>sum > nrOfVars := 0. > nrEmpty := 0. > measurements do: [ :measurement | > nrOfVars := nrOfVars + (measurement varSize * measurement instSize). > nrEmpty := nrEmpty + measurement nrEmpty + measurement nrNil]. > > In Famix-SelfMeasurements on smalltalkhub > _______________________________________________ > Moose-dev mailing list > Moose-dev(a)iam.unibe.ch > https://www.iam.unibe.ch/mailman/listinfo/moose-dev

Aug. 29, 2013

Re: [Pharo-dev] It was a productive and nice sprint.

by Igor Stasenko

On 29 August 2013 09:54, Tudor Girba <tudor(a)tudorgirba.com> wrote: > Yes! I would love to have such a tool. > > i was thinking that too.. but as it appears it is too damn slow and (probably) unreliable. The algorithm conceptually quite simple.. but since the number of back references grow exponentially at each iteration, and you need to scan whole memory for every of them, it turns out that you can wait for result for days.. So, here the code, take a look and let me know what you thinking about it. > Doru > > > On Thu, Aug 29, 2013 at 9:32 AM, Igor Stasenko <siguctua(a)gmail.com> wrote: > >> >> >> >> On 29 August 2013 08:53, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>wrote: >> >>> in the bug tracker >>> >>> Igor spent the day chasing bad registration hanging around and fixing >>> the tools to let him >>> get there. >>> >>> >> unfortunately i didn't managed to chase them all. >> found couple leaks, but there are more which still escaping :( >> >> what i want to do is to write small tracer which will search for local >> roots of set of objects. >> >> >>> stef >>> >>> On Aug 29, 2013, at 7:00 AM, Tudor Girba <tudor(a)tudorgirba.com> wrote: >>> >>> Thank you! >>> >>> Where can I see the pointer explorer? >>> >>> Doru >>> >>> >>> >>> >>> On Wed, Aug 28, 2013 at 9:00 PM, StÃ©phane Ducasse < >>> stephane.ducasse(a)inria.fr> wrote: >>> >>>> Thanks all for your good energy. >>>> I will push some of the fixes in the stream too :) >>>> >>>> For the non participating guys we got >>>> - compiler fixes >>>> - faster tests >>>> - cleans in morphic >>>> - better pointer explorer >>>> - critics browser working >>>> - grease better configuration >>>> - and many moreâ¦ >>>> >>>> Stef >>>> >>> >>> >>> >>> -- >>> www.tudorgirba.com >>> >>> "Every thing has its own flow" >>> >>> >>> >> >> >> -- >> Best regards, >> Igor Stasenko. >> > > > > -- > www.tudorgirba.com > > "Every thing has its own flow" > -- Best regards, Igor Stasenko.

Aug. 29, 2013

[update 3.0] #30358

by StÃ©phane Ducasse

30358 ----- - issue 2489 Strange behavior of Rectangle (first part) https://pharo.fogbugz.com/default.asp?2489 - issue 11441 Use SmalllintManifestChecker in assertValidLintRule: https://pharo.fogbugz.com/default.asp?11441 Diff information: http://smalltalkhub.com/mc/Pharo/Pharo30/main/Tests-StephaneDucasse.618.diff http://smalltalkhub.com/mc/Pharo/Pharo30/main/SUnit-Core-StephaneDucasse.94… http://smalltalkhub.com/mc/Pharo/Pharo30/main/Refactoring-Environment-Steph… http://smalltalkhub.com/mc/Pharo/Pharo30/main/Refactoring-Critics-StephaneD… http://smalltalkhub.com/mc/Pharo/Pharo30/main/Manifest-Tests-StephaneDucass… http://smalltalkhub.com/mc/Pharo/Pharo30/main/Manifest-Core-StephaneDucasse… http://smalltalkhub.com/mc/Pharo/Pharo30/main/Kernel-StephaneDucasse.1582.d…

Aug. 29, 2013

Re: [Pharo-dev] method vs. selector consistency?

by Esteban Lorenzano

On Aug 29, 2013, at 3:43 PM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > > On Aug 29, 2013, at 2:28 PM, Denis Kudriashov <dionisiydk(a)gmail.com> wrote: > >> And now Package should be used instead Class category. Am I right? > > Kind of. > We would loveâ¦. and we will... but we are still in transition with this... probably for pharo3, but most sure for pharo4 :) > > >> 2013/8/29 Esteban Lorenzano <estebanlm(a)gmail.com> >> they are "protocols" >> method category is the old non-accurate name :) >> >> Esteban >> >> On Aug 29, 2013, at 1:44 PM, btc(a)openInWorld.com wrote: >> >> > StÃ©phane Ducasse wrote: >> >> methodsInCategory: aName "Answer a list of the methods of the receiver that are in category named aName" >> >> >> >> | aColl | >> >> aColl := Set withAll: (self organization listAtCategoryNamed: aName). >> >> ^ aColl asArray sort >> >> >> >> >> >> allMethodsInCategory: aName "Answer a list of all the methods of the receiver and all its superclasses that are in the category named aName" >> >> >> >> | aColl | >> >> aColl := OrderedCollection new. >> >> self withAllSuperclasses do: [:aClass | aColl addAll: (aClass organization listAtCategoryNamed: aName) ]. >> >> ^ aColl asSet asArray sort >> >> >> >> >> >> uncategorizedMethods >> >> ^ self methodsInCategory: Protocol unclassified >> >> >> >> all these methods returns selectors and not compiled methods. >> >> >> >> So I would like to rename them accordingly. >> >> What do you think? >> >> >> >> Stef >> >> >> >> >> > When writing about Smalltalk, I am often confused about whether I should call them Method Categories or Protocols ? >> > I actually prefer the latter, partly because it reads better in prose not having to write Class Categories and Method Categories every time, and just Categories and Protocols. >> > Are the terms interchangeable? What is the official stance? >> > >> > cheers -ben >> > >> > >> >> >> >

Aug. 29, 2013

Zinc-Server-Sent-Events

by Sven Van Caekenberghe

Hi, Server-Sent Events is a new HTML5 technology that is used to send data from a server to a client, over HTTP, on the initiative of the server. Here are some references: http://en.wikipedia.org/wiki/Server-sent_events http://dev.w3.org/html5/eventsource/ https://developer.mozilla.org/en-US/docs/Server-sent_events/Using_server-se… Zinc-Server-Sent-Events is a proof of concept implementation with a small demo of Server-Sent Events for Zinc HTTP Components in Pharo. Usage in quite simple: you specify a handler object that implements a #run: method for each outgoing feed, like this run: connection "Using Server-Sent Events is simple: keep a reference to the connection, instanciate ZnServerSentEvent objects and use #writeOn: to push them out." | event status | [ status := self statusString. event := ZnServerSentEvent type: 'status-update' data: status. event writeOn: connection. 2 seconds asDelay wait ] repeat On the client side, a standard JavaScript listener will receive the incoming events. The package can be found here http://www.smalltalkhub.com/#!/~SvenVanCaekenberghe/ZincHTTPComponents/pack… Make sure to update Zinc itself to its bleedingEdge first. Sven -- Sven Van Caekenberghe http://stfx.eu Smalltalk is the Red Pill

Aug. 29, 2013

Re: [Pharo-dev] method vs. selector consistency?

by StÃ©phane Ducasse

On Aug 29, 2013, at 2:28 PM, Denis Kudriashov <dionisiydk(a)gmail.com> wrote: > And now Package should be used instead Class category. Am I right? Kind of. We would loveâ¦. > 2013/8/29 Esteban Lorenzano <estebanlm(a)gmail.com> > they are "protocols" > method category is the old non-accurate name :) > > Esteban > > On Aug 29, 2013, at 1:44 PM, btc(a)openInWorld.com wrote: > > > StÃ©phane Ducasse wrote: > >> methodsInCategory: aName "Answer a list of the methods of the receiver that are in category named aName" > >> > >> | aColl | > >> aColl := Set withAll: (self organization listAtCategoryNamed: aName). > >> ^ aColl asArray sort > >> > >> > >> allMethodsInCategory: aName "Answer a list of all the methods of the receiver and all its superclasses that are in the category named aName" > >> > >> | aColl | > >> aColl := OrderedCollection new. > >> self withAllSuperclasses do: [:aClass | aColl addAll: (aClass organization listAtCategoryNamed: aName) ]. > >> ^ aColl asSet asArray sort > >> > >> > >> uncategorizedMethods > >> ^ self methodsInCategory: Protocol unclassified > >> > >> all these methods returns selectors and not compiled methods. > >> > >> So I would like to rename them accordingly. > >> What do you think? > >> > >> Stef > >> > >> > > When writing about Smalltalk, I am often confused about whether I should call them Method Categories or Protocols ? > > I actually prefer the latter, partly because it reads better in prose not having to write Class Categories and Method Categories every time, and just Categories and Protocols. > > Are the terms interchangeable? What is the official stance? > > > > cheers -ben > > > > > > >

Aug. 29, 2013

Re: [Pharo-dev] method vs. selector consistency?

by StÃ©phane Ducasse

Protocols we will rename every categories into protocols. then methods -> compiled Methods selectors -> symbol > StÃ©phane Ducasse wrote: >> methodsInCategory: aName "Answer a list of the methods of the receiver that are in category named aName" >> >> | aColl | >> aColl := Set withAll: (self organization listAtCategoryNamed: aName). >> ^ aColl asArray sort >> >> >> allMethodsInCategory: aName "Answer a list of all the methods of the receiver and all its superclasses that are in the category named aName" >> >> | aColl | >> aColl := OrderedCollection new. >> self withAllSuperclasses do: [:aClass | aColl addAll: (aClass organization listAtCategoryNamed: aName) ]. >> ^ aColl asSet asArray sort >> >> >> uncategorizedMethods >> ^ self methodsInCategory: Protocol unclassified >> >> all these methods returns selectors and not compiled methods. >> >> So I would like to rename them accordingly. >> What do you think? >> >> Stef >> >> > When writing about Smalltalk, I am often confused about whether I should call them Method Categories or Protocols ? > I actually prefer the latter, partly because it reads better in prose not having to write Class Categories and Method Categories every time, and just Categories and Protocols. > Are the terms interchangeable? What is the official stance? > > cheers -ben > >

Aug. 29, 2013

SHA1 changed ??

by Sven Van Caekenberghe

Max, Why was the contract of SHA1>>hashStream: changed ? It used to return a ByteArray like other HashFunction subclasses, now it returns an Integer. I see that you also changed the tests with this assumption. MD5 hashMessage: 'foo'. #[172 189 24 219 76 194 248 92 237 239 101 79 204 196 164 216] SHA1 hashMessage: 'foo'. 68123873083688143418383284816464454849230703155 It broke Zinc-WebSockets in 3.0 and now I will have to do an ugly hack to make the code work on multiple Pharo versions. Can you please explain ? Sven

Aug. 29, 2013