July 2009 - Pharo-dev - lists.pharo.org

[Pharo-project] speedup to window activation
by Torsten Bergmann July 23, 2009

July 23, 2009

comment from Morphic-Windows-edc.2.mcz in http://source.squeak.org/trunk "Adds a nice speedup to window activation by removing a bit of code intended as optimization but actually causing significnt slowdowns when opening new windows. Speeds up the overall browser benchmark by 20-25%." Maybe someone with a little bit time is able to have a look if this can be applied for pharo too... -- GRATIS fÃ¼r alle GMX-Mitglieder: Die maxdome Movie-FLAT! Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome01

3 5

[Pharo-project] Author class
by Miguel CobÃ¡ July 23, 2009

July 23, 2009

I'm fixing a bug in the world menu -> System -> Preferences -> "set author initials". Questions? The package it is Morphic, the class is TheWorldMenu. This package, in the last core image is modified but not commited so when following the steps from the HowToContribute wiki page, in the part of creating a mcz package with the change the mcz generate will include the other uncommited changes. Should I use a changeset instead in order to not commit changes that I don't want? What is the goal with the Author class? I can see that the Utilities methods that used to be use to set author name and initials are being replaced with calls to Author. Also, they are deprecated in Utilities. Questions: Author has name and initials, what will be the intended use of them? By previous mails in the list I remember that Pharo doesn't want to use initials (like mcm, lr, avi) but full name as hinted by the message prompt in the Author requestName. But then initials will have exactly the same that name in the author class? I can unify all the message sends to Author in the core image, also I can unify the initials and name usage, but I need to know the intended goal for this class. What is the goal? Cheers, Miguel CobÃ¡

7 20

[Pharo-project] FastSets bug fix
by Ralph Boland July 23, 2009

July 23, 2009

For some reason one of my bug fixes didn't make it into the version of FastSets I released. The fix is there now. The bug is in MethodDictionary so it's a must have for anybody using FastSets. Ralph Boland

1 0

[Pharo-project] Seaside question
by Schwab,Wilhelm K July 23, 2009

July 23, 2009

Hello all, Have there been any changes to the way Seaside behaves? I recall two things: (1) using it caused my image to bloat up by an additional 30+MB; (2) I could safely use #startOn: and #stop to control whether or not it was running. Now I am finding the bloat is not nearly so bad as before[*], and every image save seems to restart the Seaside process, whether I had used #stop or not. A side effect of the latter is that #startOn: ends up complaining about a service already being on 8080 (and the process is running), even though WAKom port returns nil. I, rather un-deftly, tweaked the way I register my Seaside apps, so I'm willing to entertain the idea that I confused something. Bottom line: have there been any relevant changes to expected behavior, and if not, any thoughts on how to undo any confusion I might have caused? Bill [*] Perhaps I am stumbling less with time, maybe leaving a reduced amount of challenging junk for the GC??

2 1

Re: [Pharo-project] [ANN] Better profiler thanks to Juan (of CUIS)
by StÃ©phane Ducasse July 23, 2009

July 23, 2009

I imagine that this is this one spyEvery: millisecs at: spyPriority on: aBlock "Create a spy and spy on the given block at the specified rate." | myDelay startTime time0 observedProcess | (aBlock isMemberOf: BlockClosure) ifFalse: [self error: 'spy needs a block here']. self class: aBlock receiver class method: aBlock method. "set up the probe" observedProcess := Processor activeProcess. myDelay := Delay forMilliseconds: millisecs. time0 := Time millisecondClockValue. gcStats := SmalltalkImage current getVMParameters. Timer := [ [true] whileTrue: [ startTime := Time millisecondClockValue. myDelay wait. self tally: Processor preemptedProcess suspendedContext in: (ShowProcesses ifTrue: [ observedProcess == Processor preemptedProcess ifTrue: [observedProcess] ifFalse: [nil]]) "tally can be > 1 if ran a long primitive" by: (Time millisecondClockValue - startTime) // millisecs]. nil] newProcess. Timer priority: spyPriority. "activate the probe and evaluate the block" Timer resume. ^ aBlock ensure: [ "Collect gc statistics" SmalltalkImage current getVMParameters keysAndValuesDo: [ :idx :gcVal | gcStats at: idx put: (gcVal - (gcStats at: idx))]. "cancel the probe and return the value" Timer terminate. time := Time millisecondClockValue - time0]

1 0

[Pharo-project] PackageOrganizer seems to be full of crap....
by StÃ©phane Ducasse July 23, 2009

July 23, 2009

I will check how to clean that but there is something wrong somewhere. Probably the class is not that used too. PackageOrganizer default packageNames a Set('System-FilePackage' 'Kernel' 'Flash' 'Collections-Stack' 'OmniBrowser' 'System-Object Events' 'Network-MIME' 'ToolBuilder- Specs' 'ToolBuilder-Kernel' 'Tests-PrimCallController' 'Tests-System- Settings' 'Network-Protocols' 'Network-Kernel' 'System-Clipboard' 'System-FileRegistry' 'FFI' 'Tweak' 'Traits' 'Files' 'SystemChangeNotification-Tests' 'Network-RFC822' 'Network- RemoteDirectory' 'System-Download' 'Polymorph-ToolBuilder' 'Tests- Bugs' 'System-Applications' 'HostMenus' 'Services-Base' 'VersionNumber' 'Tests-System-Digital Signatures' 'BalloonMMFlash' 'Tests' 'Polymorph-Widgets' 'KernelTests' 'CollectionsTests' '39Deprecated' 'ToolsTest' 'MultilingualTests' 'MonticelloGUI' 'CollectionExtensions' 'Graphics' 'System' 'Collections-Support' 'Collections-Arrayed' 'NetworkTests' 'FixUnderscores' 'Monticello' 'Morphic' 'Protocols' 'Network-MailSending' 'PlusTools' 'MonticelloConfigurations' 'Collections-Sequenceable' 'TrueType' 'PinesoftEnhancementsForFreetype' 'Tests-System-Object Events' 'System- Support' 'Polymorph-Geometry' 'MonticelloMocks' 'MorphicExtras' 'Balloon' 'Network-UUID' '38Deprecated' 'Collections-Unordered' 'StarSqueak' 'Collections' 'Collections-Abstract' 'FlexibleVocabularies' 'System-Localization' 'System-Finalization' 'ToolBuilder-Morphic' 'System-Changes' 'ReleaseBuilder' 'Collections- Streams' 'Splijoin' 'EventRewrite2' 'Closure' 'Sixx' 'System-Settings' 'Collections-Weak' 'ToolBuilder-Examples' 'FreeTypeTests' 'VB-Regex' 'System-Pools' 'Multilingual' 'Polymorph-EventEnhancements' 'Nebraska' 'Speech' 'Network-URI' 'Tests-SystemChangeNotification' 'Pinesoft- Removals-EToys' 'ScriptLoader' 'Collections-Strings' 'System-Change Notification' 'OB-Standard' 'MorphicTests' #'Monticello-Mocks' 'Tools' 'Tests-VM' 'System-Serial Port' 'DeleteMe' 'Polymorph-Tools-Diff' 'SUnit' 'Compression' 'Announcements' 'Network-Url' 'EToys' 'System- Platforms' 'System-Digital Signatures' 'System-Tools' 'Sound' 'Tests- ObjectsAsMethods' 'SMLoader' 'FixInvisible' 'FlexibleVocabularies- Info' 'PackageInfo' 'Collections-Text' 'Collections-SkipLists' 'GraphicsTests' 'SUnitGUI' 'SmaCC' 'FreeTypeSubPixelAntiAliasing' 'Morphic-TrueType' 'System-Object Storage' 'SMBase' 'PreferenceBrowser' 'ToolBuilder-SUnit' 'ThreadSafeTranscript' 'Installer-Core' 'System-Hashing' 'MethodFinder' 'ST80' 'Exceptions' 'SplitJoin' 'Morphic-Balloon' 'Movies' 'Tests-Utilities' 'ToolBuilder- MVC' 'BlockClosureTests' 'Tests-Release' 'Compiler' 'InstallerAll' 'FreeType' 'Installer-Test' 'CompilerTests') Stef

2 3

[Pharo-project] [ANN] 10388 new core image released
by StÃ©phane Ducasse July 23, 2009

July 23, 2009

Issue 637: Undeclared ProjectNavigationMorph FFI updated in ScriptLoader Better comment of the process - isFontAvailable fix - Remove check for higherPerformance in doOneCycleFor: (preference to be removed in a second step) Issue 943: fix polymorph cursor Polymorph Fix for colouring of submenu markers. W2K Improvements. Fix for drop-list button label in W2K theme. Support for adjustment of label inset for buttons when pressed. Added #setColor: to ColorPresenterMorph to enable use without a model. Quit gets back its normal behavior Issue 944: SUnit assert:equals SUnit assert equals + better feedback message Issue 947: Deprecation is broken after sprint http://code.google.com/p/pharo/issues/detail?id=947 Deprecate some more Utilities methods - Gofer script loader :) - removing some methods of Project and friends (the eternal battle) - Issue 948: Mantis 7361: bug 0.5 >= (1/2) - remove PackageInfo from MessageSet methodsFor: methodDisplayStringForClass: class selector: sel - Henrik speedup - TextMorph handleInteraction uses self editor Issue 348: UIManager enhacements 2 - Part one Issue 348: UIManager enhacements 2 - Part Two - ClosureTests expectedFailures - MethodFinder, SelectorBrowser, ThreadSafeTranscript - Move JoystickMorph to MorphicExtras package - Remove a couple of references to Squeak in strings (like SqueakDebug.log -> PharoDebug.log) - ScriptLoader>>#cleanUpForProduction - delete unused bitmaps in ScriptingSystem Issue 443: Adrian and I have spent some time on fixing the bug. A fix has been produced: - Kernel-adrian_lienhard.358 in PharoInbox - Traits-Alexandre_Bergel.285 contains a test Missing part added a test that pinpoint why the checkSanity fails. Pair programmed with Andreas L.'. - Issue 962: remove BlockArgsNode KeyWordNode - Issue 463: SmallDictionary should not be a subclass of Set (and Dictionary) cleaner SmallDictionary + more tests - removing RBSmallDictionaryTests and MethodDictionaryTestBis - Issue 964: Better Message Tally Thanks to Juan Vuletich and CUIS - Detangle MorphicExtras and extract package to external SqueakSource project (PharoMorphicExtras). -Fixing fixing fixing collection tests now they are all greeeeeeen stef and cyrille -Fixing message tally ObservedProcess - fix Host* tests - integrate hash functions from Cryptography (MIT licensed) - remove VersionHistory and VersionNumber'. Issue 937: [fix] remove some unimplemented calls from BalloonEngine Issue 888: LongTestCase is unpackaged Issue 962: remove BlockArgsNode KeyWordNode Issue 894: Browser>>copyClass is lost Issue 960: unused methods in MethodDictionary Issue 898: MessageNotUnderstood: UndefinedObject>>something

6 10

[Pharo-project] [ANN] Dependency Browser update
by HernÃ¡n Morales Durand July 23, 2009

July 23, 2009

Dear all, I've updated the Dependency Browser package and documentation, take a look here: http://cs.hernanmorales.com.ar/projects/dependencyBrowser/DBrowser-en.php Quick install if you have OmniBrowser and Installer packages loaded: | dwInstaller | dwInstaller := Installer ss project: 'DependencyWalker'. dwInstaller install: 'DependencyWalker'; install: 'OBDependencyWalker'; install: 'DWTraitsFix'. ( Smalltalk version beginsWith: 'Squeak' ) ifTrue: [ dwInstaller install: 'OBDWMethodNodeFix' ] Cheers. HernÃ¡n

3 3

July 23, 2009

Juan I fixed it in the subsequent release but could you have a look and if necessary I will remove it. Stef On Jul 20, 2009, at 8:54 PM, Juan Vuletich wrote: > stephane ducasse wrote: >> Hi >> >> I spent some time looking at the enh of juan and the results are >> really cool. >> Thanks Juan for making that available. (BTW if you can sign and >> send/fax me the pharo license it will be easier for the future) >> I updated the class comments: >> >> You can also see the results using the Debug menu :) >> > > Hi Stef, > > I'm glad to be of help. Just a comment. I did not include > #spyAt:on: . I don't think it is good. With it, if I remember > correctly, in your example, > > > Here we do not see factorial since it is run at a high > priority > > [1000 timesRepeat: [ > > 100 timesRepeat: [120 factorial]. > > (Delay forMilliseconds: 10) wait > > ]] forkAt: 45 named: '45'. > > MessageTally spyAt: 40 on: [10000 timesRepeat: > [1.23 printString]] > > the time used by the factorial will be assigned to some random lower > priority process. I think this is a bug, and I recommend removing > #spyAt:on: > > BTW, I'll send you the license agreement. Is a digital photo of the > signed paper ok? > > Cheers, > Juan Vuletich >

1 0

[Pharo-project] Updating 10387 failed
by Nicolas Cellier July 23, 2009

July 23, 2009

This is what I get during the update: ThirtyTwoBitRegister doesNotUnderstand: #load: This is triggered by: MD5NonPrimitive>>initialize "MD5 initialize" "Obscure fact: those magic hex numbers that are hard to type in correctly are actually the result of a simple trigonometric function and are therefore easier to compute than proofread. Laziness is sometimes a virtue." | c | c _ 2 raisedTo: 32. SinTable _ Array new: 64. 1 to: 64 do: [ :i | SinTable at: i put: (ThirtyTwoBitRegister new load: (c * i sin abs) truncated)]. SNIP... Is this related to load order or something? Also mind the previously banned underscores... Nicolas

2 1