- Pharo-dev - lists.pharo.org

ConfigurationiOfGitHelp is broken in catalog
by Esteban Lorenzano Sept. 6, 2016

Sept. 6, 2016

Please fix. most probably is because you do not have a correct config, which means: a baseline (even if not used), a version and a symbolic version #stable. Esteban

1 0

strange FFI behaviour? (or lack of understanding?)
by Ben Coman Sept. 5, 2016

Sept. 5, 2016

Is it a **requirement** that FFI calls have a method all to themselves like a primitive call? I have this definition... FFIExternalStructure subclass: #CXString CXString class>>fieldsDesc ^ #( void *data; uint private_flags; ) The instance returned by this... Xxxx>>getClangVersion ^ self ffiCall: #( CXString clang_getClangVersion () ) module: Libclang has private_flags=1 indicating the library allocated external memory for *data and to release this I need to call clang_disposeString( CXString string). However this doesn't guard against being called twice and double-free()'ing CXString crashing the VM, so I defined the following... CXString>>dispose self inform: 'debug_dispose1'. self private_flags = 1 ifTrue: [ self inform: 'debug_dispose2'. self private_flags: 0. self ffiCall: #( void clang_disposeString ( CXString self ) ) module: Libclang ] ifFalse: [ self inform: 'debug_dispose3'. Error signal: 'Cannot dispose twice' ]. However something strange, dispose will only execute once. For example for... s := Libclang getClangVersion. s dispose. "==>debug_dispose1, debug_dispose2" s dispose. "==>nothing at all" I would expect the second call to dispose to result it "==>debug_dispose1, debug_dispose3" but it does nothing at all. Debugging into "S dispose" displays the source of dispose and then returns without executing anything. The image seems fine and displays the updated zero in private_flags -- at least for a few minutes, then it might crash indicating the clang_disposeString() has been called twice. However it works if I separate out the FFI call to a separate method... CXString>>unsafeDispose self ffiCall: #( void clang_disposeString ( CXString self ) ) module: Libclang CXString>>dispose self inform: 'dispose1'. self private_flags = 1 ifTrue: [ self inform: 'dispose2'. self private_flags: 0. self unsafeDispose. ] ifFalse: [ self inform: 'dispose3'. Error signal: 'Cannot dispose twice' ]. s := Libclang getClangVersion. s dispose. "==>debug_dispose1, debug_dispose2" s dispose. "==>debug_dispose1, debug_dispose3" s dispose. "==>debug_dispose1, debug_dispose3" s dispose. "==>debug_dispose1, debug_dispose3" Can someone hazard a guess what is behind this behaviour? I would prefer not to require the unguarded #unsafeDispose. cheers -ben

1 1

Message browser strange new behavior
by stepharo Sept. 5, 2016

Sept. 5, 2016

Hi Often I look at senders and I modify them directly in the message browser and I do not one by one. Now when I recompile a method I get a new one and I lose the next one. Am I the only one to have this behavior. It is highly disrputing. Stef

3 5

Iceberg Question
by Alexandre Bergel Sept. 5, 2016

Sept. 5, 2016

Hi! I have a question regarding the window: Why if I want to retrieve the new versions, I need to press the button âFetch new versionsâ. I think I should do a pull ? But instead, i can do a load or a merge. A bit crypticâ¦ Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

3 2

about package naming
by stepharo Sept. 5, 2016

Sept. 5, 2016

Hi guys Sorry to be that dull and stupid but I do not get why (beside following the seaside convention - and I do not understand it either) why the test package for FileSystem-Core is named FileSystem-Tests-Core and not just FileSystem-Core-Tests like that we could have FileSystem-Core-Help FileSystem-Core-Examples And when we are looking for FileSystem-Core we can see automatically FileSystem-Core-Tests Now we cannot. Stef

7 12

ObjectVirus is renamed to ObjectMutation
by Denis Kudriashov Sept. 5, 2016

Sept. 5, 2016

Hi. Finally I made this change: ObjectVirus project is renamed to ObjectMutation. And all virus related words are now around word "mutate". Probably this metaphor is not good for some people but at least it has no negative impact which was mentioned here. Mocketry is adopted for new names too. Best regards, Denis

3 2

Got some crashes with recent versions
by stepharo Sept. 5, 2016

Sept. 5, 2016

Hi I got two vm crashes in a couple of hours and we are wondering if this is not due to the recent flushCache issues. So pay attention and do not hesitate to report if you have problems. Stef

1 0

Bizarre Text Selection in Pharo 5.0
by Sean P. DeNigris Sept. 5, 2016

Sept. 5, 2016

It seems impossible that no one has encountered this, but I didn't find anything on the mailing list. When I select text in Nautilus code pane and then e.g. press $[ to add brackets around the selection smartly, the selection shifts and one of the brackets ends up in the wrong place. Anyone else running into this? ----- Cheers, Sean -- View this message in context: http://forum.world.st/Bizarre-Text-Selection-in-Pharo-5-0-tp4908498.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

3 6

[pharo-project/pharo-core] c0db47: 60199
by GitHub Sept. 5, 2016

Sept. 5, 2016

Branch: refs/heads/6.0 Home: https://github.com/pharo-project/pharo-core Commit: c0db47aa711a34a5b3df11c9ba33ccab066fd891 https://github.com/pharo-project/pharo-core/commit/c0db47aa711a34a5b3df11c9… Author: Jenkins Build Server <board(a)pharo-project.org> Date: 2016-09-02 (Fri, 02 Sep 2016) Changed paths: M HudsonBuildTools20.package/HDTestCoverage.class/instance/actions/install.st M HudsonBuildTools20.package/HDTestCoverage.class/instance/actions/uninstall.st R HudsonBuildTools20.package/HDTestCoverage.class/instance/private/flushCache.st M Kernel.package/MethodDictionary.class/instance/accessing/at_put_.st M Kernel.package/MethodDictionary.class/instance/removing/removeKey_ifAbsent_.st M Refactoring-Core.package/RBNamespace.class/instance/changes/defineClass_.st M Refactoring-Core.package/RBNamespace.class/instance/changes/renameClass_to_around_.st A Refactoring-Core.package/RBNamespace.class/instance/private/markAsRemoved_.st A Refactoring-Core.package/RBNamespace.class/instance/private/unmarkAsRemoved_.st A Refactoring-Tests-Core.package/RBRenameClassTest.class/instance/tests/testUnmarkRemovedClassIfRenameTargetClass.st M Reflectivity.package/ReflectiveMethod.class/instance/invalidate/installMethod_.st R SUnit-UI.package/TestCoverage.class/instance/private/flushCache.st R ScriptLoader60.package/ScriptLoader.class/instance/pharo - scripts/script60198.st A ScriptLoader60.package/ScriptLoader.class/instance/pharo - scripts/script60199.st R ScriptLoader60.package/ScriptLoader.class/instance/pharo - updates/update60198.st A ScriptLoader60.package/ScriptLoader.class/instance/pharo - updates/update60199.st M ScriptLoader60.package/ScriptLoader.class/instance/public/commentForCurrentUpdate.st M Traits.package/TBehavior.class/instance/adding%2Fremoving methods/basicAddSelector_withMethod_.st M Traits.package/TBehavior.class/instance/adding%2Fremoving methods/basicRemoveSelector_.st Log Message: ----------- 60199 19021 RB forgets to unmark removedClasses https://pharo.fogbugz.com/f/cases/19021 10025 MethodDictionary at:put: and #flushCache https://pharo.fogbugz.com/f/cases/10025 http://files.pharo.org/image/60/60199.zip

2 1

could we discuss 19006
by stepharo Sept. 5, 2016

Sept. 5, 2016

Hi guys The change 19006 is adding a lot of streams AbstractBinaryStream and I do not really get the vision. Note that I'm not against. I just want to understand. Do we add these and remove some old ones? What is sync? Ideally I would like to throw away all the streams and use xtreams instead. So thanks for your time explaining us the idea. Stef

4 4