- Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Class vars and inst vars of a class

by Yuriy Tymchuk

Thank you all for your replies. I would say, that âClass Variableâ name is a bit misleading, but now I know what is going on behind it. :) Cheers! Uko On 19 Mar 2014, at 23:53, Ben Coman <btc(a)openinworld.com> wrote: > Yuriy Tymchuk wrote: >> Can anybody tell me what is the difference between the class variable on the "instance sideâ definition and instance variable on the âclass sideâ definition? >> >> Thanks in advance. >> Uko >> >> > instance-variable: A variable that holds the private state of an object. In Smalltalk these can only be accessed by instance-methods. > > class-variable: An attribute of a class that is shared by that class, its subclasses, and all instances of those classes. Can be accessed by both instance-methods and class-methods. > > class-instance-variable: An attribute of a class that is not shared outside that specific class. Only the definition is inherited by subclasses, not the value . Each subclass has its own private value. Accessible only by class-methods. > > > You will understand it best by experimenting. Here is one I made for myself... > > ivCounter - instance variable counter > CVCounter - class variable counter > CIVCounter - class instance variable counter > > "----------" > Object subclass: #TestCounter > instanceVariableNames: 'ivCounter iName' > classVariableNames: 'CVCounter' > poolDictionaries: '' > category: 'MyExamples' > > TestCounter class "accessed by clicking the <Class> button" > instanceVariableNames: 'CIVCounter' > > TestCounter subclass: #TestSubCounter > instanceVariableNames: '' > classVariableNames: '' > poolDictionaries: '' > category: 'MyExamples' > "----------" > TestCounter name: myInstanceName > iName := myInstanceName. > TestCounter >> next > ivCounter ifNil: [ ivCounter := 0 ]. > CVCounter ifNil: [ CVCounter := 0 ]. > ivCounter := ivCounter + 1. > CVCounter := CVCounter + 1. > TestCounter >> printOn: aStream > super printOn: aStream. > aStream nextPutAll: ' ['. aStream > print: iName. > aStream nextPutAll: '] ( '. aStream > nextPutAll: ' ivCounter='; > print: ivCounter. > aStream > nextPutAll: ', CIVCounter=n/a'. > aStream > nextPutAll: ', CVCounter='; > print: CVCounter. > aStream nextPutAll: ' )'. "----------" TestCounter class >> next2 > CIVCounter ifNil: [ CIVCounter := 0 ]. > CVCounter ifNil: [ CVCounter := 0 ]. > > CIVCounter := CIVCounter + 1. > CVCounter := CVCounter + 1. > TestCounter class >> printOn: aStream > super printOn: aStream. > aStream nextPutAll: ' ( '. aStream > nextPutAll: 'ivCounter=n/a'. > aStream > nextPutAll: ', CIVCounter='; > print: CIVCounter. > aStream > nextPutAll: ', CVCounter='; > print: CVCounter. > aStream nextPutAll: ' )'. TestCounter class >> reset > CVCounter := nil. > CIVCounter := nil. > > "---------" > Then from Workspace evaluate these... > > tc1 := TestCounter new. > tc2 := TestCounter new. > > tsc1 := TestSubCounter new. > tsc2 := TestSubCounter new. > > TestCounter reset. > TestSubCounter reset. > > tc1 next. tc2 next. > tc1 next. > tc2 next. > > tsc1 next. > tsc2 next. > tsc1 next. > tsc2 next. > > TestCounter next2. > TestCounter next2. > TestCounter next2. > > > TestSubCounter next2. > TestSubCounter next2. > TestSubCounter next2. > > cheers -ben > > > > >

March 20, 2014

keybindings problem?

by Tudor Girba

Hi, We (Andrei and I) are experiencing some strange behavior recently related to keybindings. In the example below: - pressing "Cmd+/" brings up the halt - pressing "Cmd+o" does not bring up the halt ((PluggableTextMorph on: 'haba' text: #asText accept: nil) bindKeyCombination: $o command toAction: [ 'does not halt' halt ]; bindKeyCombination: $/ command toAction: [ 'it halts' halt ]) openInWorld One difference between the two is that Cmd+o is already define inside the text morph, while Cmd+/ is not. However, this used to work in 30796 but does not work in 30798. I do not have time right now to look at it more deeply. Do you know why the behavior changed suddenly? What should we adapt? Doru -- www.tudorgirba.com "Every thing has its own flow"

March 20, 2014

An attempt to use interactive tools detected, while in non-interactive mode

by Sebastian Sastre

Hi there, my Pharo 3.0 staging builds started to fail today and this seems to be involved. Tomorrow Iâll investigate more so pointers will be appreciated muchly what is this really about? and why the headless load isnât ignoring the warning? Loading 1.5 of ConfigurationOfMongoTalk... Project: Grease stable [1.1.6] Fetched -> Grease-Core-JohanBrichau.86 --- http://smalltalkhub.com/mc/Seaside/Grease11/main/ --- http://smalltalkhub.com/mc/Seaside/Grease11/main/ Fetched -> Grease-Pharo30-Core-JohanBrichau.13 --- http://smalltalkhub.com/mc/Seaside/Grease11/main/ --- http://smalltalkhub.com/mc/Seaside/Grease11/main/ Fetched -> Grease-Tests-Core-JohanBrichau.95 --- http://smalltalkhub.com/mc/Seaside/Grease11/main/ --- http://smalltalkhub.com/mc/Seaside/Grease11/main/ Fetched -> Grease-Tests-Pharo20-Core-JohanBrichau.9 --- http://smalltalkhub.com/mc/Seaside/Grease11/main/ --- http://smalltalkhub.com/mc/Seaside/Grease11/main/ Fetched -> Grease-Slime-JohanBrichau.25 --- http://smalltalkhub.com/mc/Seaside/Grease11/main/ --- http://smalltalkhub.com/mc/Seaside/Grease11/main/ Fetched -> Mongo-Squeak-EstebanLorenzano.13 --- http://smalltalkhub.com/mc/francois/MongoTalk/main/ --- http://smalltalkhub.com/mc/francois/MongoTalk/main/ Fetched -> Mongo-BSON-EstebanLorenzano.31 --- http://smalltalkhub.com/mc/francois/MongoTalk/main/ --- http://smalltalkhub.com/mc/francois/MongoTalk/main/ Fetched -> Mongo-Core-DamienCassou.35 --- http://smalltalkhub.com/mc/francois/MongoTalk/main/ --- http://smalltalkhub.com/mc/francois/MongoTalk/main/ Fetched -> MongoQueries-NicolasPetton.6 --- http://smalltalkhub.com/mc/francois/MongoTalk/main/ --- http://smalltalkhub.com/mc/francois/MongoTalk/main/ Loaded -> Grease-Core-JohanBrichau.86 --- http://smalltalkhub.com/mc/Seaside/Grease11/main/ --- cache'Errors in script loaded from /var/lib/jenkins/workspace/stagingFlowingWorker/FlowingWorkerBuilder.st' [31mAn attempt to use interactive tools detected, while in non-interactive mode Interactive Request: You are about to load new versions of the following packages that have unsaved changes in the image: Grease-Pharo30-Core If you continue, you will lose these changes: [0mNonInteractiveUIManager>>nonInteractiveWarning: NonInteractiveUIManager>>nonInteractiveRequest:title: NonInteractiveUIManager>>nonInteractiveRequest: NonInteractiveUIManager>>confirm:trueChoice:falseChoice:cancelChoice:default: MCMergeOrLoadWarning>>defaultAction UndefinedObject>>handleSignal: MethodContext(ContextPart)>>handleSignal: MethodContext(ContextPart)>>handleSignal: MethodContext(ContextPart)>>handleSignal: MCMergeOrLoadWarning(Exception)>>signal MCMergeOrLoadWarning class>>signalFor: MCVersionLoader>>warnAboutLosingChangesTo:ifCancel:ifMerge: MCVersionLoader>>checkForModificationsIfCancel:ifMerge: MCVersionLoader>>loadWithNameLike: MCVersionLoader>>load in Block: [ ... BlockClosure>>ensure:

March 20, 2014

Re: [Pharo-dev] Class vars and inst vars of a class

by Ben Coman

Yuriy Tymchuk wrote: > Can anybody tell me what is the difference between the class variable on the "instance sideâ definition and instance variable on the âclass sideâ definition? > > Thanks in advance. > Uko > > instance-variable: A variable that holds the private state of an object. In Smalltalk these can only be accessed by instance-methods. class-variable: An attribute of a class that is shared by that class, its subclasses, and all instances of those classes. Can be accessed by both instance-methods and class-methods. class-instance-variable: An attribute of a class that is not shared outside that specific class. Only the definition is inherited by subclasses, not the value . Each subclass has its own private value. Accessible only by class-methods. You will understand it best by experimenting. Here is one I made for myself... ivCounter - instance variable counter CVCounter - class variable counter CIVCounter - class instance variable counter "----------" Object subclass: #TestCounter instanceVariableNames: 'ivCounter iName' classVariableNames: 'CVCounter' poolDictionaries: '' category: 'MyExamples' TestCounter class "accessed by clicking the <Class> button" instanceVariableNames: 'CIVCounter' TestCounter subclass: #TestSubCounter instanceVariableNames: '' classVariableNames: '' poolDictionaries: '' category: 'MyExamples' "----------" TestCounter name: myInstanceName iName := myInstanceName. TestCounter >> next ivCounter ifNil: [ ivCounter := 0 ]. CVCounter ifNil: [ CVCounter := 0 ]. ivCounter := ivCounter + 1. CVCounter := CVCounter + 1. TestCounter >> printOn: aStream super printOn: aStream. aStream nextPutAll: ' ['. aStream print: iName. aStream nextPutAll: '] ( '. aStream nextPutAll: ' ivCounter='; print: ivCounter. aStream nextPutAll: ', CIVCounter=n/a'. aStream nextPutAll: ', CVCounter='; print: CVCounter. aStream nextPutAll: ' )'. "----------" TestCounter class >> next2 CIVCounter ifNil: [ CIVCounter := 0 ]. CVCounter ifNil: [ CVCounter := 0 ]. CIVCounter := CIVCounter + 1. CVCounter := CVCounter + 1. TestCounter class >> printOn: aStream super printOn: aStream. aStream nextPutAll: ' ( '. aStream nextPutAll: 'ivCounter=n/a'. aStream nextPutAll: ', CIVCounter='; print: CIVCounter. aStream nextPutAll: ', CVCounter='; print: CVCounter. aStream nextPutAll: ' )'. TestCounter class >> reset CVCounter := nil. CIVCounter := nil. "---------" Then from Workspace evaluate these... tc1 := TestCounter new. tc2 := TestCounter new. tsc1 := TestSubCounter new. tsc2 := TestSubCounter new. TestCounter reset. TestSubCounter reset. tc1 next. tc2 next. tc1 next. tc2 next. tsc1 next. tsc2 next. tsc1 next. tsc2 next. TestCounter next2. TestCounter next2. TestCounter next2. TestSubCounter next2. TestSubCounter next2. TestSubCounter next2. cheers -ben

March 19, 2014

Re: [Pharo-dev] Class vars and inst vars of a class

by Nicolas Cellier

A class variable is shared between all the subclasses. A class side instance variable is like instance variable, a slot at class side: every subclass (class side) will have its own slot. 2014-03-19 23:26 GMT+01:00 Yuriy Tymchuk <yuriy.tymchuk(a)me.com>: > Can anybody tell me what is the difference between the class variable on > the "instance side" definition and instance variable on the "class side" > definition? > > Thanks in advance. > Uko >

March 19, 2014

Re: [Pharo-dev] Class vars and inst vars of a class

by Eliot Miranda

Hi Uko, On Wed, Mar 19, 2014 at 3:26 PM, Yuriy Tymchuk <yuriy.tymchuk(a)me.com> wrote: > Can anybody tell me what is the difference between the class variable on > the "instance side" definition and instance variable on the "class side" > definition? > > Thanks in advance. > Uko > Class variables, as in classVariableNames: 'AClassVariable' are global variables in scope in all methods of the class and its subclass, instannce methods of the class and instance methods if the subclass. In the example, there is only one AClassVariable in the class. Class instance variables are instance variables in the class object. The class object inherits from Behavior, ClassDescription and Class, and hence already has the superclass, methodDict, format, instanceVariables, organization, subclasses, name, classPool, sharedPools instance variables (details depending on dialect). The instanceVariableNames: on the class side add to these, but they are per-instance, and so the class and each of its subclasses has one of these instance variables and hence each class can have its own value for it, just as the class and each of its subclasses have their own method dictionaries. Further, the class instance variable is only in scope in class methods. HTH, Eliot

March 19, 2014

Class vars and inst vars of a class

by Yuriy Tymchuk

Can anybody tell me what is the difference between the class variable on the "instance sideâ definition and instance variable on the âclass sideâ definition? Thanks in advance. Uko

March 19, 2014

Re: [Pharo-dev] Font problem is still there

by Tudor Girba

There exists a issue for it: https://pharo.fogbugz.com/f/cases/12777/Athens-font-cacheing-bug Doru On Wed, Mar 19, 2014 at 9:23 PM, Tudor Girba <tudor(a)tudorgirba.com> wrote: > Does it mean that we should add the workaround back? > > At this moment, to get the Moose image to work, we load an older version > of Athens manually. > > Cheers, > Doru > > > > > On Wed, Mar 19, 2014 at 3:07 PM, Igor Stasenko <siguctua(a)gmail.com> wrote: > >> >> >> >> On 19 March 2014 13:43, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: >> >>> Hi! >>> >>> We have had experience a problem with font when using Athens to display >>> text. >>> >>> >>> >>> Apparently, this is due to some recent version of Athens. We have to >>> load the following packages to not experience the problem: >>> >>> Name: Athens-Core-MarcusDenker.34 >>> Author: MarcusDenker >>> Time: 5 July 2013, 10:59:47.206954 am >>> UUID: e844b2d4-b091-42a4-9be0-17101dcdbd30 >>> Ancestors: Athens-Core-MarcusDenker.33, Athens-Core-ErwanDouaille.33 >>> >>> Name: Athens-Cairo-MarcusDenker.51 >>> Author: MarcusDenker >>> Time: 26 August 2013, 4:03:06.190096 pm >>> UUID: 9e1bfddb-67f7-4a36-864d-11060b8b6881 >>> Ancestors: Athens-Cairo-SebastianTleye.50 >>> >>> This is a _major_ problem to us and apparently there is no simple >>> solution. These kind of problem may force us to move to seek for an >>> alternative to Athens. >>> >>> >> there's some kind of caching interference either in cairo or between >> cairo and freetype, >> when using identity transformation matrix and integral (integer) font >> sizes. >> and there is a simple workaround: >> >> canvas pathTransform restoreAfter: [ >> canvas pathTransform scaleBy: 1.00001. >> "draw text here" >> >> ] >> >> this workaround was in font-rendering code, but then i removed it, >> because i thought problem was solved. >> >> Alexandre >>> -- >>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >>> Alexandre Bergel http://www.bergel.eu >>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >>> >>> >>> >>> >> >> >> -- >> Best regards, >> Igor Stasenko. >> > > > > -- > www.tudorgirba.com > > "Every thing has its own flow" > -- www.tudorgirba.com "Every thing has its own flow"

March 19, 2014

Re: [Pharo-dev] Font problem is still there

by Tudor Girba

Does it mean that we should add the workaround back? At this moment, to get the Moose image to work, we load an older version of Athens manually. Cheers, Doru On Wed, Mar 19, 2014 at 3:07 PM, Igor Stasenko <siguctua(a)gmail.com> wrote: > > > > On 19 March 2014 13:43, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > >> Hi! >> >> We have had experience a problem with font when using Athens to display >> text. >> >> >> >> Apparently, this is due to some recent version of Athens. We have to load >> the following packages to not experience the problem: >> >> Name: Athens-Core-MarcusDenker.34 >> Author: MarcusDenker >> Time: 5 July 2013, 10:59:47.206954 am >> UUID: e844b2d4-b091-42a4-9be0-17101dcdbd30 >> Ancestors: Athens-Core-MarcusDenker.33, Athens-Core-ErwanDouaille.33 >> >> Name: Athens-Cairo-MarcusDenker.51 >> Author: MarcusDenker >> Time: 26 August 2013, 4:03:06.190096 pm >> UUID: 9e1bfddb-67f7-4a36-864d-11060b8b6881 >> Ancestors: Athens-Cairo-SebastianTleye.50 >> >> This is a _major_ problem to us and apparently there is no simple >> solution. These kind of problem may force us to move to seek for an >> alternative to Athens. >> >> > there's some kind of caching interference either in cairo or between cairo > and freetype, > when using identity transformation matrix and integral (integer) font > sizes. > and there is a simple workaround: > > canvas pathTransform restoreAfter: [ > canvas pathTransform scaleBy: 1.00001. > "draw text here" > > ] > > this workaround was in font-rendering code, but then i removed it, because > i thought problem was solved. > > Alexandre >> -- >> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >> Alexandre Bergel http://www.bergel.eu >> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >> >> >> >> > > > -- > Best regards, > Igor Stasenko. > -- www.tudorgirba.com "Every thing has its own flow"

March 19, 2014

Re: [Pharo-dev] [Pharo-users] [ANN] ArchLinux pharo-vm / pharo-launcher packages

by Laurent Laffont

Le mercredi 19 mars 2014, 15:05:56 Damien Cassou a Ã©crit : > On Tue, Mar 18, 2014 at 10:40 PM, Laurent Laffont > <laurent.laffont(a)gmail.com> wrote: > > you can now install pharo vm on Arch with: > > > that's brilliant. Thanks. Can I reference your work in > http://www.smalltalkhub.com/#!/~Pharo/PharoLauncher ? Sure ! Laurent > >

March 19, 2014