Squeak-dev November 2010

squeak-dev@lists.squeakfoundation.org

77 participants
254 discussions

The Trunk: Tools-ul.276.mcz
by commits＠source.squeak.org 06 Nov '10

06 Nov '10

Levente Uzonyi uploaded a new version of Tools to project The Trunk: http://source.squeak.org/trunk/Tools-ul.276.mcz ==================== Summary ==================== Name: Tools-ul.276 Author: ul Time: 6 November 2010, 4:10:08.889 am UUID: e7a95acb-6dee-e348-9725-7063ed5a72f7 Ancestors: Tools-ul.275 - use ClassDiffBuilder for diffing class definitions and comments in ChangeList - if MessageTally is not in memory, then don't load it in Debugger >> #process:controller:context: (from Cuis) =============== Diff against Tools-ul.275 =============== Item was changed: ----- Method: ChangeList>>contentsDiffedFromCurrent (in category 'viewing access') ----- contentsDiffedFromCurrent "Answer the contents diffed forward from current (in-memory) method version" | aChange aClass | listIndex = 0 ifTrue: [^ '']. aChange := changeList at: listIndex. (aChange type == #method and: [(aClass := aChange methodClass) notNil and: [aClass includesSelector: aChange methodSelector]]) ifTrue: [^Utilities methodDiffFor: aChange text class: aClass selector: aChange methodSelector prettyDiffs: self showingPrettyDiffs]. aChange type == #doIt ifTrue: [| tokens | tokens := Scanner new scanTokens: aChange string. ((tokens select: [:substr| #(subclass: variableByteSubclass: variableWordSubclass: instanceVariableNames: classVariableNames: ) includes: substr]) asSet size >= 3 and: [(aClass := Smalltalk at: tokens third ifAbsent: []) notNil and: [aClass isBehavior]]) ifTrue: + [^ClassDiffBuilder buildDisplayPatchFrom: aClass definition to: aChange string]. - [^TextDiffBuilder buildDisplayPatchFrom: aClass definition to: aChange string]. (tokens size = 4 and: [tokens second == #class and: [tokens third == #instanceVariableNames: and: [(aClass := Smalltalk at: tokens first ifAbsent: []) notNil and: [aClass isBehavior]]]]) ifTrue: + [^ClassDiffBuilder buildDisplayPatchFrom: aClass class definition to: aChange string]]. - [^TextDiffBuilder buildDisplayPatchFrom: aClass class definition to: aChange string]]. (aChange type == #classComment and: [(aClass := aChange commentClass) notNil]) ifTrue: + [^ClassDiffBuilder buildDisplayPatchFrom: aClass comment asString to: aChange string]. - [^TextDiffBuilder buildDisplayPatchFrom: aClass comment asString to: aChange string]. ^(changeList at: listIndex) text! Item was changed: ----- Method: Debugger>>process:controller:context: (in category 'private') ----- process: aProcess controller: aController context: aContext super initialize. + Smalltalk at: #MessageTally ifPresentAndInMemory: [ :tally | + tally terminateTimerProcess]. - Smalltalk at: #MessageTally ifPresent:[:tally| tally terminateTimerProcess]. contents := nil. interruptedProcess := aProcess. interruptedController := aController. contextStackTop := aContext. self newStack: (contextStackTop stackOfSize: 1). contextStackIndex := 1. externalInterrupt := false. selectingPC := true. Smalltalk isMorphic ifTrue: [errorWasInUIProcess := false]!

1 0

Bug (feature?) in SystemWindow>>addAllMorphs:aCollection
by Lawson English 05 Nov '10

05 Nov '10

OK, don't do this at home, but even so, this shouldn't permanently disable the system... Using Squeak 4.1 + Seaside 3.0: myWindow := (SystemWindow labelled: 'test') openInWorld. myArray := Array new: 10. "add zero or more morphs to myArray, but less than myArray size" myWindow addAllMorphs: myArray. "squeak stops working..." Lawson

2 3

building Cog / FFiPrims plgin on windows
by Alain_Rastoul 05 Nov '10

05 Nov '10

Hi I downloaded the Cog trunk , the last FFI version from Monticello, used the workspace script to generate the source and build Cog on Windows with cygwin (gcc 4.5.1). Every thing is ok, but the FFI plugin doesn't work. When I evaluate Win32Window class>>coloredEllipses I have SmallInteger(Object)>>doesNotUnderstand: #getHDCDuring: Win32Window getFocus evaluates to a integer handle, not a Win32Window. I did (Smalltalk at: #ExternalType) initialize. (Smalltalk at: #ExternalStructure) compileAllFields. and also tryed to build only the plugin with Mingw gcc 2.95.2 but I have allways the same problem Is the FFI plugin functional ? am I missing something ? What version of gcc is ok ? (the make CC=gcc-3 LD=gcc-3 DLLWRAP='dllwrap -mno-cygwin --driver-name gcc-3' doesn't work) Thanks, Regards, Alain

4 19

Daily Commit Log
by commits＠source.squeak.org 05 Nov '10

05 Nov '10

Changes to Trunk (http://source.squeak.org/trunk.html) in the last 24 hours: http://lists.squeakfoundation.org/pipermail/packages/2010-November/003883.h… Name: System-ul.390 Ancestors: System-ul.389 - be more explicit about attribute changes to gain another 5-10x speedup =============================================

1 0

Squeak hang at full cpu - help
by Alain_Rastoul 04 Nov '10

04 Nov '10

Hi Everybody, I have a standard squeak 4.1 with seaside 3.0 preloaded under Windows XP (seaside 3.0 rc OneClick Image) with FFI, ODBC packages installed and a very memory intensive application who load data from a MSSQL server and hash it into dictionaries. It is running in a process started at systemBackgroundPriority, and is launched with a MessageTally spyAllOn:. After running a few minutes it hangs up at full CPU, and I can't see what is going on, everything is hung (no Ctrl., no mouse) When I open the debug console and dump processes I see only few processes (MessageTally>spyEvery:on:, handleTimerEvent, userInterruptWatcher, lowSpaceWatcher) but no UI process and no app process. I tryed to put halts in code but it doesn't change anything. I noticed that it occurs when the memory usage is about 99% (I see it with a MemoryUsageMorph ticking), but not allways at the same time, it can load 10k, 20k or 30k rows before hanging. If I remove the MessageTally, it ran several times without hanging. I don't see any direct relation between MessageTally and hanging the VM (allocations / GC ?) I will try to generate and compile the vm with MSVC if I can (I can't test my app under linux) What is the way go with this kind of problem ? Any help, tip for debugging, advice, idea, remark (...) are welcome. Thanks in advance Regards, Alain

3 6

stdio support
by Levente Uzonyi 04 Nov '10

04 Nov '10

Hi, yesterday I added Eliot's stdio change set to the Trunk. There are some questions left to be answered: 1) Do we want to use CrLfFileStream? Currently it has only two users in the image, but the class is "patched" so #new will return an instance of MultiByteFileStream in those cases. The change set overrides this "patch" for the stdio streams. 2) Read buffering: currently it's enabled for the stdio streams. For stdout and stderr it doesn't make a difference. For stdin it's a problem if these streams should be really shared. For example: if there are 10 bytes readable from stdin and you evaluate [StandardFileStream stdin next], then it will fetch all 10 bytes. If you then evaluate [MultiByteFileStream stdin next], then you'll get nothing (nil). If we want these streams to be shared like this (accessible via both MultiByteFileStream and StandardFileStream), then this is a problem. 3) Should we set the line end convention of MultiByteFileStream for these streams? It's currently not set. Cheers, Levente

7 19

The Trunk: System-ul.390.mcz
by commits＠source.squeak.org 04 Nov '10

04 Nov '10

Levente Uzonyi uploaded a new version of System to project The Trunk: http://source.squeak.org/trunk/System-ul.390.mcz ==================== Summary ==================== Name: System-ul.390 Author: ul Time: 4 November 2010, 10:44:58.211 pm UUID: 57baf4d2-ec35-614b-875e-bfbca4d9cbf2 Ancestors: System-ul.389 - be more explicit about attribute changes to gain another 5-10x speedup =============== Diff against System-ul.389 =============== Item was changed: ----- Method: ClassDiffBuilder>>print:withAttributes:on: (in category 'private') ----- print: aString withAttributes: attributes on: stream stream + currentAttributes: attributes; + nextPutAll: aString! - withAttributes: attributes - do: [ stream nextPutAll: aString ]! Item was changed: ----- Method: TextDiffBuilder>>print:withAttributes:on: (in category 'private') ----- print: aString withAttributes: attributes on: stream stream + currentAttributes: attributes; + nextPutAll: aString. + (aString notEmpty and: [ + aString last = Character cr or: [ + aString endsWith: String crlf ] ]) + ifFalse: [ stream cr ]! - withAttributes: attributes - do: [ - stream nextPutAll: aString. - (aString notEmpty and: [ - aString last = Character cr or: [ - aString endsWith: String crlf ] ]) - ifFalse: [ stream cr ] ]!

1 0

How weak finalization affects the performance
by Igor Stasenko 04 Nov '10

04 Nov '10

Here the idea, which came to Chris, put a delay into #finalizationProcess loop: finalizationProcess [true] whileTrue: [ WeakFinalizationList initTestPair. FinalizationSemaphore wait. FinalizationLock critical: [ WeakFinalizationList checkTestPair. FinalizationDependents do: [:weakDependent | weakDependent ifNotNil: [weakDependent finalizeValues]]] ifError: [:msg :rcvr | rcvr error: msg]. 5 seconds asDelay wait. ]. And here a simple benchmark, which triggers GC often: [ Array new: 100 ] bench without delay: '2,450,000 per second.' '2,490,000 per second.' '2,490,000 per second.' '2,480,000 per second.' '2,530,000 per second.' with delay: '2,670,000 per second.' '2,680,000 per second.' '2,690,000 per second.' '2,730,000 per second.' roughly about ~8% faster :) But now lets put something big into weak array: | dict b | dict := WeakKeyDictionary new addAll: (( 1 to: 1000 ) collect: [:i | i->i] ); yourself. WeakArray addWeakDependent: dict. b := [ Array new: 100 ] bench. WeakArray removeWeakDependent: dict. b without delay: '1,840,000 per second.' '2,060,000 per second.' '2,130,000 per second.' with delay: '3,030,000 per second.' '2,880,000 per second.' '2,890,000 per second.' Do not forget to do: WeakArray restartFinalizationProcess when you changing the #finalizationProcess method, otherwise you won't see real numbers. So, i like the idea of putting delay there. Finalization is eventual, and there is no hard guarantees that it will happen in micro-second just after some object become garbage. So, be it 5 seconds or 1000 seconds not really matters. What is matters that with delay we win much more, by avoiding wasting time in finalization process too often. -- Best regards, Igor Stasenko AKA sig.

6 12

TheWorldMainDockingBar instance
by Hannes Hirzel 04 Nov '10

04 Nov '10

Hi Is there a particular reason that I have to access the sole occurence of TheWorldMainDockingBar with the the method #instance? Only this class uses #instance all the others use #current to access a singleton object? If there are no objections I put in an update into the inbox. -- Hannes

1 0

[ENH] Syntax extension for continuations
by Igor Stasenko 04 Nov '10

04 Nov '10

Hello, Out of curiosity, i tried to look how it would be hard to change parser to support extended continuation syntax. And it was quite simple, i had to change only single method! A new syntax using double semicolon ';;' to indicate a continuation. So, code like: self foo ;; bar ;; zork is equivalent to: self foo bar zork But when its going to binary, or keyword messages, it allows to write more clean code, because it not requires using parenthesis. For instance: (1/10) asFloat with new syntax can be written as: 1/10 ;; asFloat More complex example: ((self foo: bar) + 10 ) baz can be written as self foo: bar ;; + 10 ;; baz Note, that it is same number of characters to type.. but is much more clean, and less time expensive to code, because when you coding, you usually type first message: self foo: bar ` and then you realising that next message is binary and hence you need to go back to beginning of line and put open paren there: `( self foo: bar then again, go to the end of message, and continue typing: ( self foo: bar ` ) + 10 ` - is cursor position. so, it is much more keystrokes & navigation than just typing two semicolons :) P.S. don't burn me for my herecy, please :) -- Best regards, Igor Stasenko AKA sig.

4 11

← Newer
1
...
19
20
21
22
23
24
25
26
Older →

Jump to page: