Squeak-dev September 2009

squeak-dev@lists.squeakfoundation.org

119 participants
370 discussions

Text truncation in trunk
by Javier Diaz-Reinoso 23 Sep '09

23 Sep '09

Testing some code I wrote in Squeak-3.10.2-Closures (and works ok), in the trunk I find some text truncation. Example code is: font1 := (TextFontReference toFont: (StrikeFont familyName: 'Atlanta' size: 24)). msg := (3.83 asString, ' RPS (', 3 asString, ' pts), Vavg=', 2.35 asString) asText addAttribute: font1. nota:=TextMorph new. nota contents: msg. nota openInWorld

3 3

Mutation Testing
by John M McIntosh 22 Sep '09

22 Sep '09

Although I didn't get to see the presentation on Mutation Testing at ESUG, I did have a chance to see it the following week after my XtC talk in London. At the time I had promised to try it out. http://www.slideshare.net/esug/mutation-testing My Fractions Calculator has 105 Sunits to confirm the keyboard state machine works, the calculator grammar via Smacc is sane, and the math results are as defined by either Wolfram Alpha, or Microsoft Excel rules. Actually the number of tests are 210 because they are subclassed for Excel, but running just the Wolfram Alpha based ones were sufficient to test the code base. I had two problems, one where the RBParser choked on #(#foo:. #bar:.) Oops the code really should have been #(#foo: #bar:) or {#foo:. #bar:.}, still it shouldn't have given a walkback. The second was that I had to increase the memory ceiling of the macintosh squeak carbon VM from 512MB to 1.5GB via the info.plist so that a recursive error situation would not make the VM run out of memory before the Mutation testing runtime monitor would kill the runaway task. As a result over the 105 SUnits it reported: It made 874 mutants, 873 got killed, 1 was terminated. Which means I don't have any head scratching this afternoon. -- = = = ======================================================================== John M. McIntosh <johnmci(a)smalltalkconsulting.com> Twitter: squeaker68882 Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com = = = ========================================================================

3 2

The Trunk: Graphics-jmv.77.mcz
by commits＠source.squeak.org 22 Sep '09

22 Sep '09

Juan Vuletich uploaded a new version of Graphics to project The Trunk: http://source.squeak.org/trunk/Graphics-jmv.77.mcz ==================== Summary ==================== Name: Graphics-jmv.77 Author: jmv Time: 22 September 2009, 6:42:48 am UUID: decb62a3-d543-40fe-a036-82b2664b5b78 Ancestors: Graphics-ar.76 Made #baseKern answer zero on most cases. This avoids what seems to be a bug in CharacterScanner and friends that might compute incorrect pixel width for strings. =============== Diff against Graphics-ar.76 =============== Item was changed: ----- Method: StrikeFont>>baseKern (in category 'accessing') ----- baseKern "Return the base kern value to be used for all characters." + | italic | - | italic baseKern | italic := emphasis allMask: 2. + (self isSynthetic not and: [ glyphs depth > 1 ]) ifTrue: [ + ^(italic or: [ pointSize < 9 ]) - self isSynthetic ifFalse: [ - baseKern := (italic or: [ pointSize < 9 ]) ifTrue: [ 1 ] + ifFalse: [ 0] ]. - ifFalse: [ 0 ]. - (italic not and: [pointSize = 12]) ifTrue: [ - baseKern := baseKern -1 ]. - pointSize >= 13 ifTrue: [ - baseKern := baseKern -1 ]. - pointSize >= 20 ifTrue: [ - baseKern := baseKern -1 ]. - ^baseKern - ]. italic ifFalse: [^ 0]. ^ ((self height-1-self ascent+4)//4 max: 0) "See makeItalicGlyphs" + (((self ascent-5+4)//4 max: 0))!

1 0

Squeak on Dr. Dobb's
by Chris Cunnington 22 Sep '09

22 Sep '09

Hi All, I've had the privilege of getting Squeak mentioned on Dr. Dobb's. Please check it out. http://www.ddj.com/development-tools/220100617 Chris

1 0

Another interesting behavior of Delay
by Igor Stasenko 22 Sep '09

22 Sep '09

I just found that its risky to use same instance of Delay, to wait by more than single process. | d sema t1 t2 first | d := Delay forMilliseconds: 200. sema := Semaphore new. first := nil. [ t1 := [ d wait. first ifNil: [ first := true]. ] timeToRun. sema signal ] forkAt: Processor activePriority +1. [ (Delay forMilliseconds: 100) wait. t2 := [ d wait. first ifNil: [first := false]. ] timeToRun. sema signal ] forkAt: Processor activePriority +1. sema wait. sema wait. { t1. t2. first. } both values, t1, t2 should be around 200 ms , because both processes waiting for same delay, right? But code always returns #(300 200 true) Of course, delay not guarantees that time elapsed will be the same as requested, because some higher priority process could simply prevent from activating the process with elapsed delay. But in given code, there is no such process. The 'first' flag is true, which means that process, who created first, gaining control after waiting on delay before second process, which makes things even more odd, because scheduling the same delay for a second time should not change the awake time of first one. The main reason , why i would want to reuse same delay instance, is to conserve space and not produce the garbage by creating a new Delay instance each time i want to put delay: "at some initialization " MyDelayClassVar := Delay forMilliseconds: 100. and then somewhere in another method: someMethod x to: y do: [ self doSomeThing. MyDelayClassVar wait. ] and as i illustrated, if #someMethod used by multiple processes, it starts behaving interestingly :) -- Best regards, Igor Stasenko AKA sig.

2 5

Collections
by Levente Uzonyi 22 Sep '09

22 Sep '09

Hi! I uploaded 5 packages which give cleaner and faster implementations of Set and subclasses' #fixCollisionsFrom:. These changes also replace the copying of these objects with the more common #copy and #postCopy methods. The suggested load/merge order is: Collections-ul.141 Kernel-ul.249 Collections-ul.142 Kernel-ul.250 Collections-ul.143 While I was doing these changes I came to the conclusion that Set >> #findElementOrNil: should be removed. It checks for a case which never happens (normally). I'm about to update #scanFor: to raise an error instead of returning 0 and replace all sends of #findElementOrNil: with #scanFor:. If you know something that's against this change, let me know. Cheers, Levente

2 1

Polymorph in a trunk image
by Ronald Spengler 22 Sep '09

22 Sep '09

What's different about the way Monticello decides how big it's system windows should be when new ones are open from the way the system browser does it? With Polymorph in a trunk image, something is quite wrong with the system browser that isn't wrong with Monticello. I'm hoping this is a clue to the bug. If you're interested, see the affixed image. To be clear: I have loaded: Polymorph-Widgets-gvc.97.mcz Polymorph-ToolBuilder-gvc.10.mcz Gut says it's related to something in the ToolBuilder. I'm going to start looking at what changed there in the package. -- Ron

4 11

Squeak-ja link
by Serge Stinckwich 22 Sep '09

22 Sep '09

Hi all, the link to squeak-ja mailing list is broken in this page : http://lists.squeakfoundation.org/ The right link is : http://www.smalltalk.jp/mailman/listinfo/squeak-ja Thank you. -- Serge Stinckwich UMI UMMISCO 209 (IRD/UPMC), Hanoi, Vietnam Smalltalkers do: [:it | All with: Class, (And love: it)] http://doesnotunderstand.org/

1 0

The Trunk: Kernel-ar.249.mcz
by commits＠source.squeak.org 22 Sep '09

22 Sep '09

Andreas Raab uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-ar.249.mcz ==================== Summary ==================== Name: Kernel-ar.249 Author: ar Time: 21 September 2009, 10:20:18 am UUID: e21b2985-4367-5342-8fa3-87a10df7a4d1 Ancestors: Kernel-nice.248 Put the delay guard back which prevented scheduling the same delay multiple times. It got accidentally removed in a recent cleanup. =============== Diff against Kernel-nice.248 =============== Item was changed: ----- Method: Delay>>schedule (in category 'private') ----- schedule "Schedule this delay" + beingWaitedOn ifTrue: [^self error: 'This Delay has already been scheduled.']. resumptionTime := Time millisecondClockValue + delayDuration. AccessProtect critical:[ ScheduledDelay := self. TimingSemaphore signal. ].!

1 0

The Trunk: KernelTests-ar.95.mcz
by commits＠source.squeak.org 22 Sep '09

22 Sep '09

Andreas Raab uploaded a new version of KernelTests to project The Trunk: http://source.squeak.org/trunk/KernelTests-ar.95.mcz ==================== Summary ==================== Name: KernelTests-ar.95 Author: ar Time: 21 September 2009, 10:17:45 am UUID: 80a7b914-beda-c44b-8e13-7734e7c3fe2e Ancestors: KernelTests-nice.94 Add tests for ensuring that delays raise errors when they are scheduled multiple times. =============== Diff against KernelTests-nice.94 =============== Item was added: + ----- Method: DelayTest>>testMultiSchedule (in category 'testing-limits') ----- + testMultiSchedule + "Ensure that scheduling the same delay twice raises an error" + | delay | + delay := Delay forSeconds: 1. + delay schedule. + self should:[delay schedule] raise: Error. + ! Item was added: + ----- Method: DelayTest>>testMultiProcessWaitOnSameDelay (in category 'testing-limits') ----- + testMultiProcessWaitOnSameDelay + "Ensure that waiting on the same delay from multiple processes raises an error" + | delay p1 p2 wasRun | + delay := Delay forSeconds: 1. + wasRun := false. + p1 := [delay wait] forkAt: Processor activePriority+1. + p2 := [ + self should:[delay wait] raise: Error. + wasRun := true. + ] forkAt: Processor activePriority+1. + p1 terminate. + p2 terminate. + self assert: wasRun. + + !

1 0

← Newer
1
...
7
8
9
10
11
12
13
...
37
Older →

Jump to page: