November 2010 - Pharo-dev - lists.pharo.org

[Pharo-project] Loading shout in 1.2
by Guillermo Polito Nov. 14, 2010

Nov. 14, 2010

Hi! I was trying to load dev packages in a 1.2Core. And I realized that shout uses Preferences. I changed the dependency in my image, but I think it's not right to upload the change now. With Mariano we talked some days ago about creating a PharoTemp squeaksource repo to save all this "transitory" changes... Any suggestions? Thanks!

2 2

Re: [Pharo-project] Issue 439 in pharo: Change autocompletion shortcut from cmd-q to ...?
by pharo＠googlecode.com Nov. 14, 2010

Nov. 14, 2010

Comment #13 on issue 439 by aplantec: Change autocompletion shortcut from cmd-q to ...? http://code.google.com/p/pharo/issues/detail?id=439 the probleme is that squeak/pharo is certainly the only ide that use this key combination. think of newbies. we should make pharo compliant with others regarding this particular point.

1 1

[Pharo-project] Monthly Pharo IRC or Chat meeting
by StÃ©phane Ducasse Nov. 14, 2010

Nov. 14, 2010

Hi guys I would really like that we get a monthly IRC or Chat discussions to discuss a bit the directions, efforts..... What do you think about it? What do you think is a good set up? Stef

4 3

[Pharo-project] Issue 3258 in pharo: Debugger hang when no source file
by pharo＠googlecode.com Nov. 14, 2010

Nov. 14, 2010

Status: Accepted Owner: hilaire.fernandes Labels: Milestone-1.2 Type-ReportDefect Importance-High New issue 3258 by hilaire.fernandes: Debugger hang when no source file http://code.google.com/p/pharo/issues/detail?id=3258 With Pharo 1.2 #12220 image, the image hangs when the debugger is invocated and there is no Source file. Easy to reproduce, remove the Source file and execute 'self halt'

1 2

[Pharo-project] ExternalIconFamily framework finally decoupled from Gaucho
by Fernando Olivero Nov. 14, 2010

Nov. 14, 2010

Hi, i set up the repo here http://www.squeaksource.com/ExternalIconFamily "This framework allows importing and managing any external icon family. Once imported , the icon family is reified into the image, As an example it comes with the Oxygen icon family imported, to load others see the example named OxygenIconFamilyBuilder. Subclasses of ExternalIconFamily act as caches for all the forms, that were requested previously. To load a form send, ExternalIconFamily>>iconNamed: aString size: aNumber " Now the integration to UITheme is missing, i will try to have a working example soon. Fernando

2 5

[Pharo-project] Issue 3254 in pharo: [Repackage] "Regular expressions" should be well structured
by pharo＠googlecode.com Nov. 14, 2010

Nov. 14, 2010

Status: Fixed Owner: Torsten....(a)astares.de CC: stephane.ducasse Labels: Milestone-1.2 New issue 3254 by Torsten....(a)astares.de: [Repackage] "Regular expressions" should be well structured http://code.google.com/p/pharo/issues/detail?id=3254 Pharo 1.2a 12235: I now reorganized the regular expression package: - tests are now in 'Regex-Tests-Core' (closer to the code than in the general Test package) - categories now start with 'Regex-...' instead of 'VB-Regex-...' (so nobody should think its a port from VisualBasic ;) - there was an RxParser "Documentation" method category => this has been converted to a help book - missing class comments were added and the note regarding the specific Documentation method category was removed Install the two changesets. After that you should have: - a new "Regex-Core" package - a changed package "Tests" since tests are now repackaged close to the code - a new "Regex-Help" package => this has to be SAVED - a "Regex-Tests-Help" package => this has to be SAVED - the package "VB-Regex" is obsolete then Attachments: 01 Reorg Regex.1.cs 9.6 KB 02 Enhance Regex.1.cs 44.1 KB

1 9

[Pharo-project] Issue 3247 in pharo: Remodel Text Constants
by pharo＠googlecode.com Nov. 14, 2010

Nov. 14, 2010

Status: New Owner: ---- New issue 3247 by Benjamin.VanRyseghem.Pharo: Remodel Text Constants http://code.google.com/p/pharo/issues/detail?id=3247 Gofer new squeaksource: 'PharoTaskForces'; package: 'KillTextConstants'; load. KillTextConstants new script. It remodel TextConstants as a SharedPool, and create a global dictionary SharedDictionary to allow classes to share constants. One step is missing, the one which check all methods invoking SharedDictionary to know if they really use SharedDictionary or pool variables from TextConstants

1 15

[Pharo-project] Issue 3263 in pharo: Add a followingBytecode to InstructionStream
by pharo＠googlecode.com Nov. 13, 2010

Nov. 13, 2010

Status: Accepted Owner: stephane.ducasse Labels: Milestone-1.2 New issue 3263 by stephane.ducasse: Add a followingBytecode to InstructionStream http://code.google.com/p/pharo/issues/detail?id=3263 Eliot Miranda uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-eem.477.mcz ==================== Summary ==================== Name: Kernel-eem.477 Author: eem Time: 14 August 2010, 7:05:27.885 pm UUID: 56dc3c2a-e134-4477-98c4-653096627951 Ancestors: Kernel-dtl.476 Add a followingBytecode to InstructionStream, e.g. SystemNavigation default browseAllSelect: [:m| | is | m isQuick not and: [is := InstructionStream on: m. is scanFor: [:b| b = 16r8F and: [is secondByte ~= 0 and: [is followingBytecode = 115]]]]] =============== Diff against Kernel-dtl.476 =============== Item was added: + ----- Method: InstructionStream>>nextPc: (in category 'private') ----- + nextPc: currentByte + "Answer the pc of the next bytecode following the current one, given the current bytecode.." + + | type | + type := currentByte // 16. + ^type = 8 "extensions" + ifTrue: [pc + (#(2 2 2 2 3 2 2 1 1 1 2 1 3 3 3 4) at: currentByte \\ 16 + 1)] + ifFalse: [type = 10 "long jumps" + ifTrue: [pc + 2] + ifFalse: [pc + 1]]! Item was changed: ----- Method: InstructionStream>>scanFor: (in category 'scanning') ----- scanFor: scanBlock + "Answer the index of the first bytecode for which scanBlock + answers true when supplied with that bytecode." - "Answer the index of the first bytecode for which scanBlock answer true - when supplied with that bytecode." + | method end byte | - | method end byte type | method := self method. end := method endPC. [pc <= end] whileTrue: [(scanBlock value: (byte := method at: pc)) ifTrue: [^true]. + pc := self nextPc: byte]. - type := byte // 16. - pc := type = 8 "extensions" - ifTrue: [pc + (#(2 2 2 2 3 2 2 1 1 1 2 1 3 3 3 4) at: byte \\ 16 + 1)] - ifFalse: [type = 10 "long jumps" - ifTrue: [pc + 2] - ifFalse: [pc + 1]]]. ^false! Item was added: + ----- Method: InstructionStream>>followingBytecode (in category 'scanning') ----- + followingBytecode + "Answer the bytecode of the following bytecode (different to nextByte)." + + ^self method at: self followingPc! Item was added: + ----- Method: InstructionStream>>followingPc (in category 'scanning') ----- + followingPc + "Answer the pc of the following bytecode." + + ^self nextPc: (self method at: pc)!

1 1

[Pharo-project] Issue 3262 in pharo: Don't cache the debugger's debuggerMap, it's not safe.
by pharo＠googlecode.com Nov. 13, 2010

Nov. 13, 2010

Status: Accepted Owner: stephane.ducasse Labels: Milestone-1.2 New issue 3262 by stephane.ducasse: Don't cache the debugger's debuggerMap, it's not safe. http://code.google.com/p/pharo/issues/detail?id=3262 http://lists.squeakfoundation.org/pipermail/packages/2010-August/003602.html Name: Tools-eem.258 Ancestors: Tools-eem.257 Don't cache the debugger's debuggerMap, its not safe. Add preference for and support of opening a debugger for an error in a background process when the UI process is busy. ==================================================== Eliot Miranda uploaded a new version of Tools to project The Trunk: http://source.squeak.org/trunk/Tools-eem.258.mcz ==================== Summary ==================== Name: Tools-eem.258 Author: eem Time: 14 August 2010, 6:45:20.412 pm UUID: 261ba269-1565-45b1-9f7b-b6ce8cee312c Ancestors: Tools-eem.257 Don't cache the debugger's debuggerMap, its not safe. Add preference for and support of opening a debugger for an error in a background process when the UI process is busy. =============== Diff against Tools-eem.257 =============== Item was added: + ----- Method: Debugger class>>interruptUIProcessIfBlockedOnErrorInBackgroundProcess: (in category 'preferences') ----- + interruptUIProcessIfBlockedOnErrorInBackgroundProcess: aBoolean + InterruptUIProcessIfBlockedOnErrorInBackgroundProcess := aBoolean! Item was changed: CodeHolder subclass: #Debugger + instanceVariableNames: 'interruptedProcess interruptedController contextStack contextStackTop contextStackIndex contextStackList receiverInspector contextVariablesInspector externalInterrupt proceedValue selectingPC savedCursor isolationHead failedProject errorWasInUIProcess labelString message' + classVariableNames: 'ContextStackKeystrokes ErrorRecursion InterruptUIProcessIfBlockedOnErrorInBackgroundProcess' - instanceVariableNames: 'interruptedProcess interruptedController contextStack contextStackTop contextStackIndex contextStackList receiverInspector contextVariablesInspector externalInterrupt proceedValue selectingPC debuggerMap savedCursor errorWasInUIProcess labelString message' - classVariableNames: 'ContextStackKeystrokes ErrorRecursion' poolDictionaries: '' category: 'Tools-Debugger'! !Debugger commentStamp: '<historical>' prior: 0! I represent the machine state at the time of an interrupted process. I also represent a query path into the state of the process. The debugger is typically viewed through a window that views the stack of suspended contexts, the code for, and execution point in, the currently selected message, and inspectors on both the receiver of the currently selected message, and the variables in the current context. Special note on recursive errors: Some errors affect Squeak's ability to present a debugger. This is normally an unrecoverable situation. However, if such an error occurs in an isolation layer, Squeak will attempt to exit from the isolation layer and then present a debugger. Here is the chain of events in such a recovery. * A recursive error is detected. * The current project is queried for an isolationHead * Changes in the isolationHead are revoked * The parent project of isolated project is returned to * The debugger is opened there and execution resumes. If the user closes that debugger, execution continues in the outer project and layer. If, after repairing some damage, the user proceeds from the debugger, then the isolationHead is re-invoked, the failed project is re-entered, and execution resumes in that world. ! Item was added: + ----- Method: Debugger class>>ifPreferredInterruptUIProcessIfBlocked: (in category 'private') ----- + ifPreferredInterruptUIProcessIfBlocked: errorWasInUIProcess + | sema | + errorWasInUIProcess ifTrue: + [^self]. + self interruptUIProcessIfBlockedOnErrorInBackgroundProcess ifFalse: + [^self]. + "Only interrupt the UI if it is unresponsive (and so is doing something that needs + interrupting). Test using addDeferredUIMessage: to see if it is running UI activities." + sema := Semaphore new. + Project current addDeferredUIMessage: [sema signal]. + (sema waitTimeoutMSecs: 100) ifTrue: + [[Project current interruptName: 'Interrupt from Background Error'] fork] + + "| s | + s := Semaphore new. + [self assert: 0 > 1000. s signal] fork. + s wait"! Item was changed: ----- Method: Debugger class>>openOn:context:label:contents:fullView: (in category 'opening') ----- openOn: process context: context label: title contents: contentsStringOrNil fullView: bool "Open a notifier in response to an error, halt, or notify. A notifier view just shows a short view of the sender stack and provides a menu that lets the user open a full debugger." + | controller errorWasInUIProcess block | - | controller errorWasInUIProcess | Smalltalk isMorphic ifTrue: [errorWasInUIProcess := Project spawnNewProcessIfThisIsUI: process] ifFalse: [controller := ScheduledControllers activeControllerProcess == process ifTrue: [ScheduledControllers activeController]]. + block := [ - WorldState addDeferredUIMessage: [ [ | debugger | debugger := self new process: process controller: controller context: context. Smalltalk isMorphic ifTrue: ["schedule debugger in deferred UI message to address redraw problems after opening a debugger e.g. from the testrunner." "WorldState addDeferredUIMessage: ["bool ifTrue: [debugger openFullNoSuspendLabel: title] ifFalse: [debugger openNotifierContents: contentsStringOrNil label: title]] ifFalse: ["deferred UI message would require special controller in MVC" bool ifTrue: [debugger openFullNoSuspendLabel: title] ifFalse: [debugger openNotifierContents: contentsStringOrNil label: title]]. debugger errorWasInUIProcess: errorWasInUIProcess. Preferences logDebuggerStackToFile ifTrue: [ Smalltalk logError: title inContext: context to: 'SqueakDebug.log']. Smalltalk isMorphic ifFalse: [ScheduledControllers searchForActiveController "needed since openNoTerminate (see debugger #open...) does not set up activeControllerProcess if activeProcess (this fork) is not the current activeControllerProcess (see #scheduled:from:)"]. ] on: Error do: [:ex | self primitiveError: 'Orginal error: ', title asString, '. Debugger error: ', ([ex description] on: Error do: ['a ', ex class printString]), ':' ] ]. + self ifPreferredInterruptUIProcessIfBlocked: errorWasInUIProcess. + Project current addDeferredUIMessage: block. + process suspend! - process suspend. - ! Item was changed: ----- Method: Debugger>>debuggerMap (in category 'accessing') ----- debuggerMap + ^self selectedContext debuggerMap! - ^debuggerMap ifNil: - [debuggerMap := self selectedContext debuggerMap].! Item was added: + ----- Method: Debugger class>>interruptUIProcessIfBlockedOnErrorInBackgroundProcess (in category 'preferences') ----- + interruptUIProcessIfBlockedOnErrorInBackgroundProcess + <preference: 'Interrupt UI process on background error' + category: 'debug' + description: 'When enabled, the debugger will interrupt the UI process if an error occurs in a background process and the UI process is blocked.' + type: #Boolean> + ^InterruptUIProcessIfBlockedOnErrorInBackgroundProcess ifNil: [false]! ====================================================

1 1

Re: [Pharo-project] Issue 2818 in pharo: Eliot and friends compiler fixes
by pharo＠googlecode.com Nov. 13, 2010

Nov. 13, 2010

Comment #3 on issue 2818 by stephane.ducasse: Eliot and friends compiler fixes http://code.google.com/p/pharo/issues/detail?id=2818 Split into parts FIrst issue done in Issue 3259

1 2