November 2015 - Pharo-dev - lists.pharo.org

Re: [Pharo-dev] autocategorizing

by Stephan Eggermont

On 11-11-15 10:36, Peter UhnÃ¡k wrote: > How does autocategorizing decides what to put where? > > For example I have (on class-side) three methods > #example > #example2 > #exampleClusters > > after autocategorizing they end up being > > #example in 'example' protocol > #example2 in 'examples' protocol > #exampleClusters in 'as yet unclassified' protocol > > That seems really odd, shouldn't all of them be in the same protocol? Depends on superclass, your use of pragmas (and extension methods): MethodClassifier>>classify: (self classifyAccessor: aMethod) ifTrue: [ ^ aMethod category ]. (self classifyInSuperclassProtocol: aMethod) ifTrue: [ ^ aMethod category ]. (self classifyByKnownPrefix: aMethod) ifTrue: [ ^ aMethod category ]. (self classifyByOtherImplementors: aMethod) ifTrue: [ ^ aMethod category ]. (self classifyByKnownPragma: aMethod) ifTrue: [ ^ aMethod category ]. > Also is 'example' or 'examples' the normative protocol? examples https://pharo.fogbugz.com/f/cases/16841/MethodClassifier-should-classify-on… https://pharo.fogbugz.com/f/cases/16842/Recategorize-example-examples MethodClassifier>>buildPragmaDictionary pragmaMapping := Dictionary new. pragmaMapping at: 'example' put: 'examples'. 'example' is currently not in the list of known prefixes. Stephan

Nov. 11, 2015

Re: [Pharo-dev] [Pharo-users] Changing playground and monticello shortcuts

by Nicolai Hess

2015-11-11 10:45 GMT+01:00 Peter UhnÃ¡k <i.uhnak(a)gmail.com>: > Hi, > > currently to open a Playground via shortcut one has to type <cmd>+o+w, > that's because it used to open Workspace. > However since it now opens Playground the shortcut is no longer > appropriate. > > So can we change it to <cmd>+o+p? > (This shortcut is currently occupied by Monticello Browser, but that can > be movd to <cmd>+o+m which is free.) > cmd+o+m would be ctrl+o ctrl+m on windows, but ctrl+m does not work yet on windows. > > You could argue that people are used to this shortcut, however that will > not be true for newcommers, who will just see it as random naming (which it > currently is). > > Peter >

Nov. 11, 2015

Changing playground and monticello shortcuts

by Peter UhnÃ¡k

Hi, currently to open a Playground via shortcut one has to type <cmd>+o+w, that's because it used to open Workspace. However since it now opens Playground the shortcut is no longer appropriate. So can we change it to <cmd>+o+p? (This shortcut is currently occupied by Monticello Browser, but that can be movd to <cmd>+o+m which is free.) You could argue that people are used to this shortcut, however that will not be true for newcommers, who will just see it as random naming (which it currently is). Peter

Nov. 11, 2015

autocategorizing

by Peter UhnÃ¡k

How does autocategorizing decides what to put where? For example I have (on class-side) three methods #example #example2 #exampleClusters after autocategorizing they end up being #example in 'example' protocol #example2 in 'examples' protocol #exampleClusters in 'as yet unclassified' protocol That seems really odd, shouldn't all of them be in the same protocol? Also is 'example' or 'examples' the normative protocol? Thanks, Peter

Nov. 11, 2015

Re: [Pharo-dev] MNU RBBroweserEnvironment>>includesKey:

by Dimitris Chloupis

there is no code compiled just me trying to create a class in the image that is named ChrTimeInput and subclasses Textmorph, there is nothing more to it image: pharo 5 latest vm: one provided by pharolauncher 'NBCoInterpreter NativeBoost-CogPlugin-EstebanLorenzano.21 uuid: 4d9b9bdf-2dfa-4c0b-99eb-5b110dadc697 Oct 8 2015 NBCogit NativeBoost-CogPlugin-EstebanLorenzano.21 uuid: 4d9b9bdf-2dfa-4c0b-99eb-5b110dadc697 Oct 8 2015 https://github.com/pharo-project/pharo-vm.git Commit: c062708833feb9048c1bad48b01867020daa0939 Date: 2015-09-30 09:22:51 +0200 By: Esteban Lorenzano <estebanlm(a)gmail.com> Jenkins build #15008 OS: Yosemite with its latest update. If you need any more info and help please dont hesitate. I am prepared to file a bug report if other can reproduce this bug. On Wed, Nov 11, 2015 at 10:11 AM Stephan Eggermont <stephan(a)stack.nl> wrote: > On 11-11-15 08:53, Dimitris Chloupis wrote: > > And I am back with new problems with pharo 5 , I hope you missed me :D > > Compiling exactly what code? In what image & vm version/os? > > Stephan > > > >

Nov. 11, 2015

Re: [Pharo-dev] MNU RBBroweserEnvironment>>includesKey:

by Dimitris Chloupis

yes the message is certainly missing, it references Smalltalk. I was just testing a simple fix, I was no way prepared to accept as a fix even if it did work because I have no clue about Pharo internal and I dont want to mess with them. On Wed, Nov 11, 2015 at 10:10 AM Thierry Goubier <thierry.goubier(a)gmail.com> wrote: > Le 11/11/2015 08:53, Dimitris Chloupis a Ã©crit : > > And I am back with new problems with pharo 5 , I hope you missed me :D > > > > If I try to create a new class called ChrTimeInputM that subclasses > > TextMorph I am getting this MNU with the stack bellow > > Strange. > > Code failing is this: > > ((oldClass isNil or: [oldClass theNonMetaClass name asString ~= > newClassName]) > and: [envt includesKey: newClassName asSymbol]) ifTrue: > ["Attempting to define new class over existing one > when > not looking at the original one in this > browser..." > (self confirm: ((newClassName , ' is an existing > class in this system. > Redefining it might cause serious problems. > Is this really what you want to do?') asText makeBoldFrom: 1 to: > newClassName size)) > ifFalse: [^ nil]]. > > The strange thing is that, of anything which looks like an environment > in Pharo, very few do have an includesKey: and certainly not > RBBrowserEnvironment > > Moreover, in the case of scoped browsing, I don't think the test amount > to anything significant: if I try to redefine Object and if Object is > not in the browsed environment, I'll get no warning whatsoever. > > Thierry > > > I tried to replace includesKey: with includesClass: but then it alerts > > me falsely that there is already an existing class in the image, I tried > > this on the latest pharo 5 image with nothing installed in it. Here is > > the stack > > > > RBBrowserEnvironment(Object)>>doesNotUnderstand: #includesKey: > > > PackageTreeNautilusUI(AbstractTool)>>compileANewClassFrom:notifying:startingFrom: > > PackageTreeNautilusUI(AbstractNautilusUI)>>compileAClassFrom:notifying: > > ClassDefinitionAcceptor>>accept:notifying: > > PackageTreeNautilusUI(NautilusUI)>>compileSource:notifying: > > PackageTreeNautilusUI(AbstractNautilusUI)>>whenTextAcceptedInView: > > MessageSend>>value: MessageSend>>cull: MessageSend>>cull:cull: [ action > > cull: anAnnouncement cull: announcer ] in > > AnnouncementSubscription>>deliver: in Block: [ action cull: > > anAnnouncement cull: announcer ] BlockClosure>>on:do: > > BlockClosure>>on:fork: AnnouncementSubscription>>deliver: [ subscription > > deliver: anAnnouncement ] in > > SubscriptionRegistry>>deliver:to:startingAt: in Block: [ subscription > > deliver: anAnnouncement ] BlockClosure>>ifCurtailed: > > SubscriptionRegistry>>deliver:to:startingAt: > > SubscriptionRegistry>>deliver:to: SubscriptionRegistry>>deliver: > > Announcer>>announce: RubScrolledTextMorph>>acceptContents > > RubScrolledTextMorph>>whenTextAcceptRequest: MessageSend>>value: > > MessageSend>>cull: MessageSend>>cull:cull: [ action cull: anAnnouncement > > cull: announcer ] in AnnouncementSubscription>>deliver: in Block: [ > > action cull: anAnnouncement cull: announcer ] BlockClosure>>on:do: > > BlockClosure>>on:fork: AnnouncementSubscription>>deliver: [ subscription > > deliver: anAnnouncement ] in > > SubscriptionRegistry>>deliver:to:startingAt: in Block: [ subscription > > deliver: anAnnouncement ] BlockClosure>>ifCurtailed: > > >

Nov. 11, 2015

Re: [Pharo-dev] MNU RBBroweserEnvironment>>includesKey:

by Thierry Goubier

Le 11/11/2015 08:53, Dimitris Chloupis a Ã©crit : > And I am back with new problems with pharo 5 , I hope you missed me :D > > If I try to create a new class called ChrTimeInputM that subclasses > TextMorph I am getting this MNU with the stack bellow Strange. Code failing is this: ((oldClass isNil or: [oldClass theNonMetaClass name asString ~= newClassName]) and: [envt includesKey: newClassName asSymbol]) ifTrue: ["Attempting to define new class over existing one when not looking at the original one in this browser..." (self confirm: ((newClassName , ' is an existing class in this system. Redefining it might cause serious problems. Is this really what you want to do?') asText makeBoldFrom: 1 to: newClassName size)) ifFalse: [^ nil]]. The strange thing is that, of anything which looks like an environment in Pharo, very few do have an includesKey: and certainly not RBBrowserEnvironment Moreover, in the case of scoped browsing, I don't think the test amount to anything significant: if I try to redefine Object and if Object is not in the browsed environment, I'll get no warning whatsoever. Thierry > I tried to replace includesKey: with includesClass: but then it alerts > me falsely that there is already an existing class in the image, I tried > this on the latest pharo 5 image with nothing installed in it. Here is > the stack > > RBBrowserEnvironment(Object)>>doesNotUnderstand: #includesKey: > PackageTreeNautilusUI(AbstractTool)>>compileANewClassFrom:notifying:startingFrom: > PackageTreeNautilusUI(AbstractNautilusUI)>>compileAClassFrom:notifying: > ClassDefinitionAcceptor>>accept:notifying: > PackageTreeNautilusUI(NautilusUI)>>compileSource:notifying: > PackageTreeNautilusUI(AbstractNautilusUI)>>whenTextAcceptedInView: > MessageSend>>value: MessageSend>>cull: MessageSend>>cull:cull: [ action > cull: anAnnouncement cull: announcer ] in > AnnouncementSubscription>>deliver: in Block: [ action cull: > anAnnouncement cull: announcer ] BlockClosure>>on:do: > BlockClosure>>on:fork: AnnouncementSubscription>>deliver: [ subscription > deliver: anAnnouncement ] in > SubscriptionRegistry>>deliver:to:startingAt: in Block: [ subscription > deliver: anAnnouncement ] BlockClosure>>ifCurtailed: > SubscriptionRegistry>>deliver:to:startingAt: > SubscriptionRegistry>>deliver:to: SubscriptionRegistry>>deliver: > Announcer>>announce: RubScrolledTextMorph>>acceptContents > RubScrolledTextMorph>>whenTextAcceptRequest: MessageSend>>value: > MessageSend>>cull: MessageSend>>cull:cull: [ action cull: anAnnouncement > cull: announcer ] in AnnouncementSubscription>>deliver: in Block: [ > action cull: anAnnouncement cull: announcer ] BlockClosure>>on:do: > BlockClosure>>on:fork: AnnouncementSubscription>>deliver: [ subscription > deliver: anAnnouncement ] in > SubscriptionRegistry>>deliver:to:startingAt: in Block: [ subscription > deliver: anAnnouncement ] BlockClosure>>ifCurtailed:

Nov. 11, 2015

Re: [Pharo-dev] MNU RBBroweserEnvironment>>includesKey:

by Stephan Eggermont

On 11-11-15 08:53, Dimitris Chloupis wrote: > And I am back with new problems with pharo 5 , I hope you missed me :D Compiling exactly what code? In what image & vm version/os? Stephan

Nov. 11, 2015

MNU RBBroweserEnvironment>>includesKey:

by Dimitris Chloupis

And I am back with new problems with pharo 5 , I hope you missed me :D If I try to create a new class called ChrTimeInputM that subclasses TextMorph I am getting this MNU with the stack bellow I tried to replace includesKey: with includesClass: but then it alerts me falsely that there is already an existing class in the image, I tried this on the latest pharo 5 image with nothing installed in it. Here is the stack RBBrowserEnvironment(Object)>>doesNotUnderstand: #includesKey: PackageTreeNautilusUI(AbstractTool)>>compileANewClassFrom:notifying:startingFrom: PackageTreeNautilusUI(AbstractNautilusUI)>>compileAClassFrom:notifying: ClassDefinitionAcceptor>>accept:notifying: PackageTreeNautilusUI(NautilusUI)>>compileSource:notifying: PackageTreeNautilusUI(AbstractNautilusUI)>>whenTextAcceptedInView: MessageSend>>value: MessageSend>>cull: MessageSend>>cull:cull: [ action cull: anAnnouncement cull: announcer ] in AnnouncementSubscription>>deliver: in Block: [ action cull: anAnnouncement cull: announcer ] BlockClosure>>on:do: BlockClosure>>on:fork: AnnouncementSubscription>>deliver: [ subscription deliver: anAnnouncement ] in SubscriptionRegistry>>deliver:to:startingAt: in Block: [ subscription deliver: anAnnouncement ] BlockClosure>>ifCurtailed: SubscriptionRegistry>>deliver:to:startingAt: SubscriptionRegistry>>deliver:to: SubscriptionRegistry>>deliver: Announcer>>announce: RubScrolledTextMorph>>acceptContents RubScrolledTextMorph>>whenTextAcceptRequest: MessageSend>>value: MessageSend>>cull: MessageSend>>cull:cull: [ action cull: anAnnouncement cull: announcer ] in AnnouncementSubscription>>deliver: in Block: [ action cull: anAnnouncement cull: announcer ] BlockClosure>>on:do: BlockClosure>>on:fork: AnnouncementSubscription>>deliver: [ subscription deliver: anAnnouncement ] in SubscriptionRegistry>>deliver:to:startingAt: in Block: [ subscription deliver: anAnnouncement ] BlockClosure>>ifCurtailed:

Nov. 11, 2015

Re: [Pharo-dev] Raspberry Pi 2 - NativeBoost?

by ClÃ©ment Bera

In short NativeBoost has in production no support for ARM assembler. At some point a DSL was built to generate assembly code in an abstract way over ARM and x86 (it was called VirtualCPU) but it has never been released. To solve this issue you need to have NativeBoost generate ARM assembly code. This should be solved mainly inside the image. Alternatively Esteban is working (and has almost finished) on a FFI backend for Cairo and FFI already supports ARM and works fine on the Pie. Best, Clement 2015-11-10 21:38 GMT-03:00 Eagle Offshore <eagleoffshore(a)mac.com>: > I've got a nifty raspberry pi 2 - which is a pretty cool lil computer. I > am amazed at how powerful it is. > > I got the build from the ci server and wanted to play with open street map > stuff but, while I have the cairo2 library installed on raspbian, the > NativeBoost class cannot load it. > > Is this a Smalltalk issue or a VM level issue? I saw some messages from > early October - please keep me in mind if you need testing help. > > Thanks, > -Todd Blanchard >

Nov. 11, 2015