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

Re: [Pharo-project] Issue 3187 in pharo: Undeclared in SmartRefStream>>#nextPut:
by pharo＠googlecode.com Nov. 13, 2010

Nov. 13, 2010

Comment #1 on issue 3187 by guillermopolito: Undeclared in SmartRefStream>>#nextPut: http://code.google.com/p/pharo/issues/detail?id=3187 There's a line that's using that instance variable, but it was removed. Does it have something to do with the Uniclasses removal? There are no tests for SmartRefStream... :(

1 3

Re: [Pharo-project] Issue 3182 in pharo: UndoManagerTest>>#testGrouping failing
by pharo＠googlecode.com Nov. 13, 2010

Nov. 13, 2010

Comment #1 on issue 3182 by akellens: UndoManagerTest>>#testGrouping failing http://code.google.com/p/pharo/issues/detail?id=3182 UndoManagerTest>>testGrouping failed because no UndoRecords were added to the group. Proposed fix: System-Undo-AndyKellens.3

1 2

[Pharo-project] Issue 3249 in pharo: SourceReference throw an error when <historical> is answered
by pharo＠googlecode.com Nov. 13, 2010

Nov. 13, 2010

Status: Fixed Owner: Benjamin.VanRyseghem.Pharo New issue 3249 by Benjamin.VanRyseghem.Pharo: SourceReference throw an error when <historical> is answered http://code.google.com/p/pharo/issues/detail?id=3249 Attached: the cs Attachments: SourceReferenceHistoricalFix.1.cs 1.4 KB

1 1

Re: [Pharo-project] Issue 3185 in pharo: BitmapStreamTests have errors
by pharo＠googlecode.com Nov. 13, 2010

Nov. 13, 2010

Updates: Cc: alexandre.bergel Comment #2 on issue 3185 by guillermopolito: BitmapStreamTests have errors http://code.google.com/p/pharo/issues/detail?id=3185 As we noticed, here are two problems: 1) SmartStreamRef>>nextPut: has already an instance variable that (as we supose) was removed while removing UniClasses) 2) After removing the line using that variable, we had another issue: The test is Saving an ImageSegment in a file, and then it is loading it using (FileDirectory default readOnlyFileNamed: externalSegmentFilename) fileInObjectAndCode which throws an Invalid UTF8 input detected (because it is assuming the file as a text file).

1 4

[Pharo-project] Issue 3256 in pharo: matchesRegex on colon
by pharo＠googlecode.com Nov. 13, 2010

Nov. 13, 2010

Status: New Owner: ---- New issue 3256 by davide.v...(a)gmail.com: matchesRegex on colon http://code.google.com/p/pharo/issues/detail?id=3256 Pharo image: Pharo-core Pharo core version: 1.2-12146 Virtual machine used: Squeak 4.2.5beta1U for mac If you evaluate: 'whatever' matchesRegex: ':' it rises an error. if I escape the colon the error disappears, but is it right? I think it's a sort of coupling with method parsing. Tx Davide

1 1

Re: [Pharo-project] smalltalk: breaking out of a loop..
by Marcus Denker Nov. 13, 2010

Nov. 13, 2010

On Nov 7, 2010, at 5:50 PM, GÃ¶ran Krampe wrote: > On 11/07/2010 08:06 PM, Guillermo Polito wrote: >> #anySatisfy: does not iterate over the entire collection either. So if >> you only want to know if there is one that satisfies the condition... >> this is the message you're looking for. >> >> Cheers! >> Guille > > And finally, a generic "trick" for breaking out of loops (not just iterations over collections): > > Factor out the loop in a single method, then call it. When you want to break out, just do a normal return! This often even makes the code more readable. But speaking of tricks... testValueWithExitBreak | val | [ :break | 1 to: 10 do: [ :i | val := i. i = 4 ifTrue: [break value]. ] ] valueWithExit. self assert: val = 4. testValueWithExitContinue | val last | val := 0. 1 to: 10 do: [ :i | [ :continue | i = 4 ifTrue: [continue value]. val := val + 1. last := i ] valueWithExit. ]. self assert: val = 9. self assert: last = 10. with BlockClosure>>valueWithExit self value: [ ^nil ] -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

3 3

[Pharo-project] Issue 3257 in pharo: ImageSegment
by pharo＠googlecode.com Nov. 13, 2010

Nov. 13, 2010

Status: Accepted Owner: stephane.ducasse New issue 3257 by stephane.ducasse: ImageSegment http://code.google.com/p/pharo/issues/detail?id=3257 If you fill an issue for the first time, please read "How to report bugs" at http://www.pharo-project.org/community/issue-tracking Pharo image: <core, dev or web> Pharo core version: <copy from World/System/About> Virtual machine used: <ex: pharo-vm-0.15.2f-linux> Class browser used if applicable: <ex: O2PackageBrowserAdaptor. If you don't know, print "SystemBrowser default"> Steps to reproduce: 1. 2. 3. Paste or attach stack trace if applicable (look at the file PharoDebug.log located in the same directory as your image):

1 0

Re: [Pharo-project] Issue 3009 in pharo: Add the overlapping package checker in ScriptLoader
by pharo＠googlecode.com Nov. 13, 2010

Nov. 13, 2010

Updates: Labels: Milestone-1.2 Comment #2 on issue 3009 by stephane.ducasse: Add the overlapping package checker in ScriptLoader http://code.google.com/p/pharo/issues/detail?id=3009 2736 the code is in RPackage apparently. So I will add it.

1 0

[Pharo-project] Strange with Author
by Hilaire Fernandes Nov. 13, 2010

Nov. 13, 2010

I have a small script to build automatically DrGeo release. In this script, I need to set the Author fullName to the value 'DrGeoUserScript'. I have enclosed below the used script. Strangely, the author fullname is not persisted in the image, although I set the author name just before the snapshot of the prepared image. Any idea ? -- Dr. Geo, to discover geometry on Linux, Windows, MAC and XO http://community.ofset.org/index.php/DrGeo

1 0

[Pharo-project] Repackaging task force
by Torsten Bergmann Nov. 13, 2010

Nov. 13, 2010

Hi, after an IRC discussion with Alexandre and Stef yesterday we agreed that it would be good to start working on cleaning up/organizing the packages and get better structure into the system. Goals: - seperate tests, examples, help, ... from code in distinguishable packages - but also have tests, examples, help packages closer to the code - be able to have cleaned up and reproducable package dependencies in the future - provide quality checks to see if packages conform to cleanup rules The abusing of class categories for packaging will not be replaced with a real packaging system in the near future but that shouldnt be a problem so far. I dont know where we will end up, how much time we can devote and what problems we may find but if we wont try we will never find out ... Proposal: ========= 1. we should have an informal repackaging task force and I would like to invite the following people since they already worked on repackaging/loading/unloading before: Adrian, Pavel, Stef and any other who wants to help 2. since we have to use a naming convention we would like to start with the one from Seaside: http://code.google.com/p/seaside/wiki/PackageNaming since it is already used and may help us here. Lets see how this turns out. 3. we should start with simple packages and clean them up 4. well formed packages also means we can run quality checks per package. Maybe the hudson infrastructure can help us here to tell if already cleaned up packages continue to be clean. To wet your appetite a little bit I worked in the announcement framework packages. See http://code.google.com/p/pharo/issues/detail?id=3252 As a goodie we have a new help book documenting the framework a little bit. Feel free to try the changesets in a recent 1.2 image and comment. I hope this doesnt start a flamewar on the correct package naming (which we can always change) but help to improve the quality of each package one by one ... Thx T. -- GRATIS! Movie-FLAT mit Ã¼ber 300 Videos. Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome

3 2