[Pharo-dev] [regression reporter]regression occurred
https://ci.inria.fr/pharo/job/Pharo-3.0-Update-Step-2.1-Validation/label=win/531/ 2 regressions found. Zinc.Tests.ZnServerTests.testEntityTooLarge Zinc.Zodiac.ZnHTTPSTests.testGetPharoVersion
Re: [Pharo-dev] [update 3.0] #30425
On Sep 26, 2013, at 8:59 PM, Pavel Krivanek pavel.kriva...@gmail.com wrote: The update build failed on Linux slave a second run leads to the same problem. This is not good…. -- Pavel 2013/9/26 Marcus Denker marcus.den...@inria.fr: 30425 - 11687 Merge Opal with Opal Repository https://pharo.fogbugz.com/f/cases/11687 11688 Pharo Kernel shrinking update for 30424 https://pharo.fogbugz.com/f/cases/11688 Diff information: http://smalltalkhub.com/mc/Pharo/Pharo30/main/System-Support-MarcusDenker.916.diff http://smalltalkhub.com/mc/Pharo/Pharo30/main/OpalCompiler-Tests-MarcusDenker.213.diff http://smalltalkhub.com/mc/Pharo/Pharo30/main/OpalCompiler-Core-MarcusDenker.414.diff signature.asc Description: Message signed with OpenPGP using GPGMail
Re: [Pharo-dev] [update 3.0] #30425
It fails during Release test. Smalltalk stack dump: 0xbfee7364 M CompiledMethod class(Behavior)allInstances 0x776a94b4: a(n) CompiledMethod class 0xbfee7380 M SmalltalkImagefixObsoleteReferences 0x77916420: a(n) SmalltalkImage 0xbfee5234 M ReleaseTesttestObsoleteClasses 0x78e70cac: a(n) ReleaseTest I tried it on my local Linux machine but it worked well :-/ -- Pavel 2013/9/27 Marcus Denker marcus.den...@inria.fr: On Sep 26, 2013, at 8:59 PM, Pavel Krivanek pavel.kriva...@gmail.com wrote: The update build failed on Linux slave a second run leads to the same problem. This is not good…. -- Pavel 2013/9/26 Marcus Denker marcus.den...@inria.fr: 30425 - 11687 Merge Opal with Opal Repository https://pharo.fogbugz.com/f/cases/11687 11688 Pharo Kernel shrinking update for 30424 https://pharo.fogbugz.com/f/cases/11688 Diff information: http://smalltalkhub.com/mc/Pharo/Pharo30/main/System-Support-MarcusDenker.916.diff http://smalltalkhub.com/mc/Pharo/Pharo30/main/OpalCompiler-Tests-MarcusDenker.213.diff http://smalltalkhub.com/mc/Pharo/Pharo30/main/OpalCompiler-Core-MarcusDenker.414.diff
Re: [Pharo-dev] [update 3.0] #30425
And it's interesting that during the first pass on a dirty image it works well even on the slave too. -- Pavel 2013/9/27 Pavel Krivanek pavel.kriva...@gmail.com: It fails during Release test. Smalltalk stack dump: 0xbfee7364 M CompiledMethod class(Behavior)allInstances 0x776a94b4: a(n) CompiledMethod class 0xbfee7380 M SmalltalkImagefixObsoleteReferences 0x77916420: a(n) SmalltalkImage 0xbfee5234 M ReleaseTesttestObsoleteClasses 0x78e70cac: a(n) ReleaseTest I tried it on my local Linux machine but it worked well :-/ -- Pavel 2013/9/27 Marcus Denker marcus.den...@inria.fr: On Sep 26, 2013, at 8:59 PM, Pavel Krivanek pavel.kriva...@gmail.com wrote: The update build failed on Linux slave a second run leads to the same problem. This is not good…. -- Pavel 2013/9/26 Marcus Denker marcus.den...@inria.fr: 30425 - 11687 Merge Opal with Opal Repository https://pharo.fogbugz.com/f/cases/11687 11688 Pharo Kernel shrinking update for 30424 https://pharo.fogbugz.com/f/cases/11688 Diff information: http://smalltalkhub.com/mc/Pharo/Pharo30/main/System-Support-MarcusDenker.916.diff http://smalltalkhub.com/mc/Pharo/Pharo30/main/OpalCompiler-Tests-MarcusDenker.213.diff http://smalltalkhub.com/mc/Pharo/Pharo30/main/OpalCompiler-Core-MarcusDenker.414.diff
[Pharo-dev] [regression reporter]regression occurred
https://ci.inria.fr/pharo/job/Pharo-3.0-Update-Step-2.1-Validation/label=win/535/ 20 regressions found. CollectionsTests.Weak.WeakFinalizersTest.testNewFinalizationSupported CollectionsTests.Weak.WeakRegistryTest.testFinalization CollectionsTests.Weak.WeakRegistryTest.testFinalizationWithMultipleFinalizersPerObject CollectionsTests.Weak.WeakRegistryTest.testFinalizationWithOnFork CollectionsTests.Weak.WeakRegistryTest.testGrowingByAHighPriorityProcessDoesntLeak CollectionsTests.Weak.WeakRegistryTest.testGrowingDoesntLeak CollectionsTests.Weak.WeakRegistryTest.testRemovingByAHighPriorityProcessDoesntLeak CollectionsTests.Weak.WeakRegistryTest.testRemovingDoesntLeak Graphics.Tests.Files.PNGReadWriterTest.testBlue16 Graphics.Tests.Files.PNGReadWriterTest.testBlue32 Graphics.Tests.Files.PNGReadWriterTest.testBlue8 NativeBoost.Tests.NBCoreTests.testExternalResourceManagerFinalization NativeBoost.Tests.NBCoreTests.testFinalization NativeBoost.Tests.NBInterpreterProxyTest.testStackPointer NativeBoost.Tests.NBInterpreterProxyTest.testStackPointer2 Tests.Finalization.ObjectFinalizerTests.testFinalizationOfMultipleResources Tests.Monticello.MCDirectoryRepositoryTest.testAddAndLoad Tests.Monticello.MCDirectoryRepositoryTest.testIncludesName Tests.Monticello.MCDirectoryRepositoryTest.testLoadMissingNode Tests.Monticello.MCDirectoryRepositoryTest.testStoreAndLoad
Re: [Pharo-dev] UI Mess
For the theme idea, I allways have in mind the structure used in VisualWorks which is to have decorator wrappers (and a few specifics like scrollbars) for each theme, and the core GUI stays the same. Morphs aren't designed that way (except for scrollbars) so it looks harder. But maybe composing would work (a morph may have a border and delegates some behavior to the border: drawing, bounds, etc... Changing the theme would mean changing the border class. In a few cases, it would mean changing the morph class completely (i.e. scrollbars)). Designing and testing that sounds like long, hard work :( For the UIManager theme thing, I would couple that to Smalltalk, something as simple as : Smalltalk ui (command line or gui) and: Smalltalk ui theme (if someone want to write an ascii-art Morphic theme :)) (I used to write dialog boxes and menu bars in ascii for a VT100 terminal a long, long time ago). Thierry Le 26/09/2013 22:14, Stéphane Ducasse a écrit : If you remember a while ago I was thinking that the theme could be managed like settings (ie encapsulated in the widgets and that the widgets could encapsulate their default for theme), but sadly I was wrong because a theme oftn requires specific computation. So probably we should have a widget and some object that represents the theme strategy. Ideally I would like to get rid of the UIManager theme idea but this is difficult and it requires a bit of prototyping. So if one of you want to try to see on a typical case how a design could work. It would be really great. Setf -- Thierry Goubier CEA list Laboratoire des Fondations des Systèmes Temps Réel Embarqués 91191 Gif sur Yvette Cedex France Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95
Re: [Pharo-dev] [regression reporter]regression occurred
It gets strange by the minute… Are we sure that the VM did not change? On Sep 27, 2013, at 9:41 AM, no-re...@ci.inria.fr wrote: https://ci.inria.fr/pharo/job/Pharo-3.0-Update-Step-2.1-Validation/label=win/535/ 20 regressions found. CollectionsTests.Weak.WeakFinalizersTest.testNewFinalizationSupported CollectionsTests.Weak.WeakRegistryTest.testFinalization CollectionsTests.Weak.WeakRegistryTest.testFinalizationWithMultipleFinalizersPerObject CollectionsTests.Weak.WeakRegistryTest.testFinalizationWithOnFork CollectionsTests.Weak.WeakRegistryTest.testGrowingByAHighPriorityProcessDoesntLeak CollectionsTests.Weak.WeakRegistryTest.testGrowingDoesntLeak CollectionsTests.Weak.WeakRegistryTest.testRemovingByAHighPriorityProcessDoesntLeak CollectionsTests.Weak.WeakRegistryTest.testRemovingDoesntLeak Graphics.Tests.Files.PNGReadWriterTest.testBlue16 Graphics.Tests.Files.PNGReadWriterTest.testBlue32 Graphics.Tests.Files.PNGReadWriterTest.testBlue8 NativeBoost.Tests.NBCoreTests.testExternalResourceManagerFinalization NativeBoost.Tests.NBCoreTests.testFinalization NativeBoost.Tests.NBInterpreterProxyTest.testStackPointer NativeBoost.Tests.NBInterpreterProxyTest.testStackPointer2 Tests.Finalization.ObjectFinalizerTests.testFinalizationOfMultipleResources Tests.Monticello.MCDirectoryRepositoryTest.testAddAndLoad Tests.Monticello.MCDirectoryRepositoryTest.testIncludesName Tests.Monticello.MCDirectoryRepositoryTest.testLoadMissingNode Tests.Monticello.MCDirectoryRepositoryTest.testStoreAndLoad signature.asc Description: Message signed with OpenPGP using GPGMail
Re: [Pharo-dev] UI Mess
Le 26/09/2013 21:54, Sean P. DeNigris a écrit : Goubier Thierry wrote answer := UIManager default chooseFileReference: 'Select a file' path: aFileReference pattern: '*.png' name: nil I can't tell if that captures the full options of fileDialogWindow without a bit more investigation, but that would be a great improvement over the current mess. I feel that whatever we do should be coupled with moving the open/answer logic into that single entry point and out of UITheme. UITheme should only provide the morph. I'd do a change for a single entry point in UIManager, the theme also providing a single entry point for creating the dialog. Sounds doable without many changes to the dialog code itself. Thierry -- Thierry Goubier CEA list Laboratoire des Fondations des Systèmes Temps Réel Embarqués 91191 Gif sur Yvette Cedex France Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95
Re: [Pharo-dev] [update 3.0] #30425
On Sep 27, 2013, at 9:25 AM, Pavel Krivanek pavel.kriva...@gmail.com wrote: And it's interesting that during the first pass on a dirty image it works well even on the slave too. I will revert the compiler merge for now (it did even run the regression tests… but showing some strange - I think - trait-recompiling related bug..) So what we will do - revert this update (it is already auto-reverted due to the crash) - save Opal as it is in the image before this to to the Opal repo - get the Opal regression green (traits bug) - get image green (we do have some failing tests) - *then* on the Opal repo, merge step-by-the intermediate 5 small commits (Abstract compiler class, some trivial cleanups, clean block analysis that is not called) I do not see how these changes could lead to the problem, but it could be that there is some bug in the compiler that shows itself when recompiling certain methods. One interesting observation: we did not check the Opal-Regression tester for 4 weeks, and we are in the mess. This just shows how important continous integration + automatic regression testing *after every commit* is for compiler development… (or being more intelligent would maybe another solution ;-) Marcus signature.asc Description: Message signed with OpenPGP using GPGMail
Re: [Pharo-dev] Strange FileDoesNotExistException when loading code from Seaside...
On Sep 26, 2013, at 8:28 PM, Mariano Martinez Peck marianop...@gmail.com wrote: Hi guys, I have a link in a Seaside app that allow us to trigger a Metacello load to update the source code. It has a weird errorI get a FileDoesNotExistException with the .changes file, which indeed exists. The error is in #readOnlyFileNamed:. I attach an screenshot of the stack. What is funny is that the compiled method that has the problem is always (or at least most of the times) either WAAttribute#accept: or WAAttribute#isAdvanced ... More funny...if I open a debugger in the problem, and I re-execute readOnlyFileNamed: it works! (of course, because the .changes file does exists). Does anyone had a similar problem? This could be related to this bug: Plugins Primitives no longer work after a while https://pharo.fogbugz.com/f/cases/7559/ In 3 I think this does not happen anymore, and I though this was fixed in 2, too… but someone recently mentioned that it happened sometimes. Marcus signature.asc Description: Message signed with OpenPGP using GPGMail
Re: [Pharo-dev] [update 3.0] #30425
On Sep 27, 2013, at 10:29 AM, Marcus Denker marcus.den...@inria.fr wrote: On Sep 27, 2013, at 10:08 AM, Marcus Denker marcus.den...@inria.fr wrote: On Sep 27, 2013, at 9:25 AM, Pavel Krivanek pavel.kriva...@gmail.com wrote: And it's interesting that during the first pass on a dirty image it works well even on the slave too. I will revert the compiler merge for now (it did even run the regression tests… but showing some strange - I think - trait-recompiling related bug..) So what we will do - revert this update (it is already auto-reverted due to the crash) - save Opal as it is in the image before this to to the Opal repo - get the Opal regression green (traits bug) - get image green (we do have some failing tests) - *then* on the Opal repo, merge step-by-the intermediate 5 small commits (Abstract compiler class, some trivial cleanups, clean block analysis that is not called) I do not see how these changes could lead to the problem, but it could be that there is some bug in the compiler that shows itself when recompiling certain methods. One interesting observation: we did not check the Opal-Regression tester for 4 weeks, and we are in the mess. This just shows how important continous integration + automatic regression testing *after every commit* is for compiler development… (or being more intelligent would maybe another solution ;-) Result: it still fails, even after reverting. Are we sure that the VM did not change in the middle? It looks *really* strange! https://ci.inria.fr/pharo/job/Pharo-3.0-Update-Step-2.1-Validation/label=linux-stable-worker/ It seems someone changed the config to load the latest VM which has problems right now. I will revert that. Marcus signature.asc Description: Message signed with OpenPGP using GPGMail
Re: [Pharo-dev] [update 3.0] #30425
On Sep 27, 2013, at 10:32 AM, Marcus Denker marcus.den...@inria.fr wrote: On Sep 27, 2013, at 10:29 AM, Marcus Denker marcus.den...@inria.fr wrote: Result: it still fails, even after reverting. Are we sure that the VM did not change in the middle? It looks *really* strange! https://ci.inria.fr/pharo/job/Pharo-3.0-Update-Step-2.1-Validation/label=linux-stable-worker/ It seems someone changed the config to load the latest VM which has problems right now. I will revert that. Ok, now we can build the image again. Marcus signature.asc Description: Message signed with OpenPGP using GPGMail
Re: [Pharo-dev] [update 3.0] #30425
It was me. I tried to change the config to use the latest VM to see if it is a vm related problem. Sorry. -- Pavel 2013/9/27 Marcus Denker marcus.den...@inria.fr: On Sep 27, 2013, at 10:32 AM, Marcus Denker marcus.den...@inria.fr wrote: On Sep 27, 2013, at 10:29 AM, Marcus Denker marcus.den...@inria.fr wrote: Result: it still fails, even after reverting. Are we sure that the VM did not change in the middle? It looks *really* strange! https://ci.inria.fr/pharo/job/Pharo-3.0-Update-Step-2.1-Validation/label=linux-stable-worker/ It seems someone changed the config to load the latest VM which has problems right now. I will revert that. Ok, now we can build the image again. Marcus
[Pharo-dev] [update 3.0] #30425
30425 - 11688 Pharo Kernel shrinking update for 30424 https://pharo.fogbugz.com/f/cases/11688 11689 Spec Inspectors still let update process in some circumstances https://pharo.fogbugz.com/f/cases/11689 Diff information: http://smalltalkhub.com/mc/Pharo/Pharo30/main/System-Support-MarcusDenker.917.diff http://smalltalkhub.com/mc/Pharo/Pharo30/main/Spec-Widgets-MarcusDenker.251.diff http://smalltalkhub.com/mc/Pharo/Pharo30/main/Spec-Inspector-MarcusDenker.96.diff http://smalltalkhub.com/mc/Pharo/Pharo30/main/Spec-Debugger-MarcusDenker.125.diff http://smalltalkhub.com/mc/Pharo/Pharo30/main/Spec-Core-MarcusDenker.161.diff signature.asc Description: Message signed with OpenPGP using GPGMail
Re: [Pharo-dev] SmalltalkHub use of LESS stylesheets
Yes, it would speedup the load of the page. I kept everything on purpose including the Smalltalk source code in Amber, its IDE, etc. So that people can see how it is online and inspect everything. Now it's true that since we have an automated build for it, I could put it in deployment mode :) Nico 2013/9/27 Sebastian Sastre sebast...@flowingconcept.com I'm curions on how this develops. Looking forward for hear about every tactic to make amber apps loading snappishly :D On Sep 26, 2013, at 11:27 PM, Esteban A. Maringolo emaring...@gmail.com wrote: Hi all, The SmalltalkHub website seems to be using LESS stylesheets, which are downloaded in a sequential order, and it takes ~20sec to download the styles. Wouldn't it be better to compile the LESS stylesheets to regular CSS files (combined and minified in a single file would be great). LESS stylesheets are not meant to be used in production environments. I thinks that only with that the speedup will be significant. Regards, Esteban A. Maringolo
[Pharo-dev] Pharo (cog) Vm on Windows with 2GB virtual mem
Hi Guys, I read in this thread http://forum.world.st/out-of-memory-cog-on-windows-td3465621.html that 512 mb limit can be avoided in windows, but a re-compilation is needed. I was wondering if the situation changed, and now we can pass a parameter. Otherwise, I would be interested if any of you has a pre-compiled vm with more mem. I need to run some experiments and I need 2GB (or 1.5 GB at least ) on a cog VM for Windows. Thanks in advance. Oscar
Re: [Pharo-dev] Pharo (cog) Vm on Windows with 2GB virtual mem
I think latest Pharo VM supports to change the memory limit by only changing an ini file. Igor did that not long ago. Cheers, Guille On Fri, Sep 27, 2013 at 2:09 PM, Oscar E. A. Callaú oalva...@dcc.uchile.clwrote: Hi Guys, I read in this thread http://forum.world.st/out-of-memory-cog-on-windows-td3465621.html that 512 mb limit can be avoided in windows, but a re-compilation is needed. I was wondering if the situation changed, and now we can pass a parameter. Otherwise, I would be interested if any of you has a pre-compiled vm with more mem. I need to run some experiments and I need 2GB (or 1.5 GB at least ) on a cog VM for Windows. Thanks in advance. Oscar
[Pharo-dev] [update 3.0] #30426
30426 - 11691 Failing test: testLocalMethodsOfTheClassShouldNotBeRepeatedInItsTraits https://pharo.fogbugz.com/f/cases/11691 11693 Spec building is taking ages due to changes in PragmaCollector https://pharo.fogbugz.com/f/cases/11693 Diff information: http://smalltalkhub.com/mc/Pharo/Pharo30/main/Spec-Core-MarcusDenker.163.diff http://smalltalkhub.com/mc/Pharo/Pharo30/main/OpalCompiler-Core-MarcusDenker.415.diff signature.asc Description: Message signed with OpenPGP using GPGMail
[Pharo-dev] VM Serial plugin
Hi all, With Santiago, we made a small fix for the serial plugin on Linux and OsX. We worked on a cloned repository of the PharoVM on github. I just sent a pull request. I guess I should also fill a bug entry somewhere. on which bugtracker, the pharo one ? Tell me and I will do it. Thanks, #Luc
[Pharo-dev] Github VM Build Validation
I tried to setup a travis validation build [1] for the pharo-vm repository on github [2]. Of course as usual I missed some libraries, does anybody know what the corresponding libs are for the following errors? /usr/bin/ld: cannot find -lSM /usr/bin/ld: cannot find -lICE /usr/bin/ld: cannot find -lGL the currently installed dependencies can be found in .travis.yml file [3]. [1] https://travis-ci.org/pharo-project/pharo-vm/builds/11851007 [2] http://github.com/pharo-project/pharo-vm [3] https://github.com/pharo-project/pharo-vm/blob/master/.travis.yml signature.asc Description: Message signed with OpenPGP using GPGMail
Re: [Pharo-dev] Github VM Build Validation
On 27 September 2013 18:00, Camillo Bruni camillobr...@gmail.com wrote: I tried to setup a travis validation build [1] for the pharo-vm repository on github [2]. Of course as usual I missed some libraries, does anybody know what the corresponding libs are for the following errors? /usr/bin/ld: cannot find -lSM /usr/bin/ld: cannot find -lICE /usr/bin/ld: cannot find -lGL libICE ICE is the Inter Client Exchange protocol, part of X11 libSM session management library for X11 and GL opengl (install mesa). btw, i wonder if the former two is necessary.. but it just what i found when was digging into makefiles and translated them to cmake config the currently installed dependencies can be found in .travis.yml file [3]. [1] https://travis-ci.org/pharo-project/pharo-vm/builds/11851007 [2] http://github.com/pharo-project/pharo-vm [3] https://github.com/pharo-project/pharo-vm/blob/master/.travis.yml -- Best regards, Igor Stasenko.
Re: [Pharo-dev] Github VM Build Validation
On 2013-09-27, at 13:33, Igor Stasenko siguc...@gmail.com wrote: On 27 September 2013 18:00, Camillo Bruni camillobr...@gmail.com wrote: I tried to setup a travis validation build [1] for the pharo-vm repository on github [2]. Of course as usual I missed some libraries, does anybody know what the corresponding libs are for the following errors? /usr/bin/ld: cannot find -lSM /usr/bin/ld: cannot find -lICE /usr/bin/ld: cannot find -lGL libICE ICE is the Inter Client Exchange protocol, part of X11 libSM session management library for X11 and GL opengl (install mesa). btw, i wonder if the former two is necessary.. but it just what i found when was digging into makefiles and translated them to cmake config thanks, actually the first two were trivial to fix. the libgl part is quite a mess currently in ubuntu 12.04: https://bugs.launchpad.net/ubuntu/+source/mesa/+bug/949606 and for now I have to manually install the 64bit dev version and the 32bit library and set up manually a proper symlink ;) signature.asc Description: Message signed with OpenPGP using GPGMail
Re: [Pharo-dev] LDAP - licensing....
I will migrate it. We also have an early chapter that olivier wrote long long time ago. stef On Sep 27, 2013, at 6:15 PM, p...@highoctane.be wrote: Hello, I am looking for an LDAP integration for Pharo 2.0 After googling, I found LDAPlayer On Squeaksource: http://www.squeaksource.com/@O_tBkvZNK0GC3Hlz/Ghl6udGE I saw from the mailing lists that Olivier Auverlot did some work on that. It would be good to have such a thing on Smalltalkhub. Maybe I missed it but the query returned no results there. Thx Phil
Re: [Pharo-dev] LDAP - licensing....
BTW olivier uses it heavily for his daily work. stef On Sep 27, 2013, at 6:15 PM, p...@highoctane.be wrote: Hello, I am looking for an LDAP integration for Pharo 2.0 After googling, I found LDAPlayer On Squeaksource: http://www.squeaksource.com/@O_tBkvZNK0GC3Hlz/Ghl6udGE I saw from the mailing lists that Olivier Auverlot did some work on that. It would be good to have such a thing on Smalltalkhub. Maybe I missed it but the query returned no results there. Thx Phil
Re: [Pharo-dev] LDAP - licensing....
Hi phil I migrated it to SmalltalkHub under PharoExtras/LDAP I'm writing the configuration since I love doing that. Stef On Sep 27, 2013, at 6:15 PM, p...@highoctane.be wrote: Hello, I am looking for an LDAP integration for Pharo 2.0 After googling, I found LDAPlayer On Squeaksource: http://www.squeaksource.com/@O_tBkvZNK0GC3Hlz/Ghl6udGE I saw from the mailing lists that Olivier Auverlot did some work on that. It would be good to have such a thing on Smalltalkhub. Maybe I missed it but the query returned no results there. Thx Phil
Re: [Pharo-dev] LDAP - licensing....
Ah ah, thanks!! I am doing some Stamp with RabbitMQ over here. Works. /Phil