January 2014 - Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Displaying labels with Athens on Linux?

by Igor Stasenko

it is stack alignment. see https://pharo.fogbugz.com/f/cases/12478/Crash-when-using-fonts-with-Athens-… for details On 14 January 2014 18:49, Igor Stasenko <siguctua(a)gmail.com> wrote: > so, yes on newer version of ubuntu it crashing. no clue why (yet) > > > On 14 January 2014 18:15, Alexandre Bergel <alexandre.bergel(a)me.com>wrote: > >> Ok, thanks Igor. >> >> This is really important for us to have Athens working well :-) >> I will soon start my lecture on software quality, and I will have about >> 10 or 20 students working with it. >> >> Alexandre >> >> >> On Jan 14, 2014, at 12:57 PM, Igor Stasenko <siguctua(a)gmail.com> wrote: >> >> > >> > >> > >> > On 14 January 2014 16:28, Alexandre Bergel <alexandre.bergel(a)me.com> >> wrote: >> > The following expression seems to reproduce the problem >> > AthensCairoSurfaceExamples exampleDrawText >> > >> > On windows and OSX it works well. On Linux it crashes >> > >> > yes.. more details: >> > we just tried with Damien on his machine, which has more recent version >> of Ubuntu, >> > and we got the crash.. >> > on my version everything works well.. >> > >> > in this regard i suspecting Cairo library.. i going to install fresh >> version of ubuntu and try things there. >> > >> > Alexandre >> > >> > On Jan 13, 2014, at 5:58 PM, Igor Stasenko <siguctua(a)gmail.com> wrote: >> > >> > > >> > > >> > > >> > > On 13 January 2014 14:53, Alexandre Bergel <alexandre.bergel(a)me.com> >> wrote: >> > > Hi! >> > > >> > > Sergio is experiencing some serious problem when displaying labels in >> Athens? >> > > The VM just crashes. >> > > >> > > Any idea what we can do? >> > > >> > > >> > > i still didn't had time to look at it. i will try to look at it with >> Ronie tomorrow, luckily he runs on linux box. >> > > it would be nice if you can give a short piece of code which >> (re)produces the crash. >> > > >> > > Alexandre >> > > -- >> > > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >> > > Alexandre Bergel http://www.bergel.eu >> > > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > -- >> > > Best regards, >> > > Igor Stasenko. >> > >> > -- >> > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >> > Alexandre Bergel http://www.bergel.eu >> > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >> > >> > >> > >> > >> > >> > >> > >> > -- >> > Best regards, >> > Igor Stasenko. >> >> -- >> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >> Alexandre Bergel http://www.bergel.eu >> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >> >> >> >> >> > > > -- > Best regards, > Igor Stasenko. > -- Best regards, Igor Stasenko.

Jan. 15, 2014

Constructing a patched Monticello package without installing it

by Eliot Miranda

Hi All, I'm trying to deploy Spur in our Newspeak bootstrap at Cadence. The starting point for the Newspeak bootstrap) is a Squeak4.3 image. I have a Squeak4.3 image transformed by the Spur bootstrap to Spur format, which means it has a few method modifications to three packages: Collections (changed Character methods now that Characters are immediate) Kernel (changed Behavior & ClassBuilder changes now that a class's format has changed, and instantiation interacts with the garbage collector differently) System (changed SystemDictionary/SmalltalkImage methods now that there's a generation scavenger) The Newspeak bootstrap loads versions of these packages that support Newspeak. So the problem is to load the Newspeak versions of these packages without undoing the Spur bootstrap modifications. As a quick hack I could patch the Monticello loader to avoid the modifications (they're marked by not having a source pointer but could be marked in some other way, e.g. a pragma). That's a fine simple approach. But... I just stumbled across Monticello's MCPatcher and the Backport button which claims to apply a partial delta to the ancestor of a package. Does anyone who understands this code know if the code supports loading a package, without installing it, applying a partial set of changes (i.e. the relevant changes form the bootstrap) and saving the package as an mcz? If it does then I could construct patched versions of the three packages and substitute them in the bootstrap instead of using the hack. -- best, Eliot

Jan. 15, 2014

Re: [Pharo-dev] problems with RPackage and cie...

by Esteban Lorenzano

but that is a hack, not a real solutionâ¦ we need to find the place where RPackage loses the update. On 15 Jan 2014, at 18:17, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > Ben had to define the method on the class RPackage: > > forceRefreshClassDefinedSelectors > | dict | > > dict := Dictionary new. > self definedClasses do: [ :e || set | > set := Set new. > e protocols > select: [ :p | (p beginsWith: '*') not ] > thenDo: [ :p | (e methodsInProtocol: p) do: [ :m | set add: m selector ] ]. > dict at: e name put:set ]. > classDefinedSelectors := dict > > Alexandre > > > On Jan 15, 2014, at 12:59 PM, Benjamin <benjamin.vanryseghem.pharo(a)gmail.com> wrote: > >> We are also facing some RPackage cache not in sync anymore with system leading to the impossibility to commit the code :( >> >> Ben >> >> On 15 Jan 2014, at 12:56, Goubier Thierry <thierry.goubier(a)cea.fr> wrote: >> >>> >>> >>> Le 15/01/2014 16:46, Alexandre Bergel a Ã©crit : >>>>> Do you have an issue on that? Or does it happens just when moving classes from any package to another? >>>> >>>> I do not know. I do not see on https://pharo.fogbugz.comhow I can list issues that contains the RPackage and that are still open. >>>> >>>> The problems we are facing usually involves having more than one browser opened, dragging classes from one browser to another browser. Hard to be more accurate because students are calling me to the rescue when they are already in the mess. We can put images online if this help. >>> >>> Ok, this could be in Nautilus. >>> >>> I was drag and dropping in AltBrowser and couldn't see anything wrong, but I'm not exercising the same code paths on RPackage than Nautilus does. >>> >>> As far as AltBrowser goes, recompiling the class with a new system category is picked up fine (apart from the disappearing tag). >>> >>> Thierry >>> >>>> Alexandre >>>> >>>> >>>>> >>>>> Thierry >>>>> >>>>> Le 15/01/2014 16:34, Alexandre Bergel a Ã©crit : >>>>>> Hi! >>>>>> >>>>>> We are constantly facing problems when moving classes between packages. This is highly discouraging. This involves tricky parts of the system that both difficult to understand and to debug. >>>>>> >>>>>> Alexandre >>>>>> >>>>> >>>>> -- >>>>> Thierry Goubier >>>>> CEA list >>>>> Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s >>>>> 91191 Gif sur Yvette Cedex >>>>> France >>>>> Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95 >>>> >>> >>> -- >>> Thierry Goubier >>> CEA list >>> Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s >>> 91191 Gif sur Yvette Cedex >>> France >>> Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95 >> > > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > >

Jan. 15, 2014

Re: [Pharo-dev] problems with RPackage and cie...

by Alexandre Bergel

Ben had to define the method on the class RPackage: forceRefreshClassDefinedSelectors | dict | dict := Dictionary new. self definedClasses do: [ :e || set | set := Set new. e protocols select: [ :p | (p beginsWith: '*') not ] thenDo: [ :p | (e methodsInProtocol: p) do: [ :m | set add: m selector ] ]. dict at: e name put:set ]. classDefinedSelectors := dict Alexandre On Jan 15, 2014, at 12:59 PM, Benjamin <benjamin.vanryseghem.pharo(a)gmail.com> wrote: > We are also facing some RPackage cache not in sync anymore with system leading to the impossibility to commit the code :( > > Ben > > On 15 Jan 2014, at 12:56, Goubier Thierry <thierry.goubier(a)cea.fr> wrote: > >> >> >> Le 15/01/2014 16:46, Alexandre Bergel a Ã©crit : >>>> Do you have an issue on that? Or does it happens just when moving classes from any package to another? >>> >>> I do not know. I do not see on https://pharo.fogbugz.comhow I can list issues that contains the RPackage and that are still open. >>> >>> The problems we are facing usually involves having more than one browser opened, dragging classes from one browser to another browser. Hard to be more accurate because students are calling me to the rescue when they are already in the mess. We can put images online if this help. >> >> Ok, this could be in Nautilus. >> >> I was drag and dropping in AltBrowser and couldn't see anything wrong, but I'm not exercising the same code paths on RPackage than Nautilus does. >> >> As far as AltBrowser goes, recompiling the class with a new system category is picked up fine (apart from the disappearing tag). >> >> Thierry >> >>> Alexandre >>> >>> >>>> >>>> Thierry >>>> >>>> Le 15/01/2014 16:34, Alexandre Bergel a Ã©crit : >>>>> Hi! >>>>> >>>>> We are constantly facing problems when moving classes between packages. This is highly discouraging. This involves tricky parts of the system that both difficult to understand and to debug. >>>>> >>>>> Alexandre >>>>> >>>> >>>> -- >>>> Thierry Goubier >>>> CEA list >>>> Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s >>>> 91191 Gif sur Yvette Cedex >>>> France >>>> Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95 >>> >> >> -- >> Thierry Goubier >> CEA list >> Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s >> 91191 Gif sur Yvette Cedex >> France >> Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95 > -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

Jan. 15, 2014

[ANN] Pillar release 0.6

by Damien Cassou

Dear lists, I'm proud to announce the 0.6 release of Pillar, a syntax and associated tools to write and generate documentation and books. Pillar is currently used to write the Enterprise Pharo book (https://ci.inria.fr/pharo-contribution/job/PharoForTheEnterprise/) and other projects. The attached screenshot shows a part of the Voyage documentation generated by Pillar (from https://github.com/SquareBracketAssociates/PharoForTheEnterprise-english/bl…) Pillar has many features: - simple markup-based syntax with references, tables, pictures, captions... (this is the syntax of Pier) - export to HTML, LaTeX and markdown (more to come) - customization of the export through a dedicated STON configuration file - support of templates using the Mustache templating engine - syntax-highlighting of generated code blocks (not yet in LaTeX) - configurable numbering of section titles and figures - ... Pillar has also: - a 5-minutes tutorial (https://github.com/DamienCassou/pillar-documentation/blob/master/README.md#…) - a documentation (https://github.com/DamienCassou/pillar-documentation) (in progress) - a good test coverage (91% with more than a 1000 executed tests) - a continuous integration job (https://ci.inria.fr/pharo-contribution/job/Pillar/) - a command-line interface (https://github.com/DamienCassou/pillar-documentation#5--command-line-interf…) - several existing use cases: https://github.com/DamienCassou/pillar-documentation#6--example-pillar-usage -- Damien Cassou http://damiencassou.seasidehosting.st "Success is the ability to go from one failure to another without losing enthusiasm." Winston Churchill

Jan. 15, 2014

Beware Issue 12657 Nautilus: Method DND explodes

by Sean P. DeNigris

Nasty one: Drag method from one open Nautilus to a class in another browser. 1. The method disappears from the original class, but is not listed as removed in MC changes 2. The method does not appear in the target class ----- Cheers, Sean -- View this message in context: http://forum.world.st/Beware-Issue-12657-Nautilus-Method-DND-explodes-tp473… Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

Jan. 15, 2014

Re: [Pharo-dev] how to debug NB?

by paolo.bernardi.67＠gmail.com

Il 13/01/2014 2.25, Igor Stasenko ha scritto: > > > > On 13 January 2014 00:08, paolo.bernardi.67(a)gmail.com > <mailto:paolo.bernardi.67@gmail.com> <paolo.bernardi.67(a)gmail.com > <mailto:paolo.bernardi.67@gmail.com>> wrote: > > Il 12/01/2014 19.41, Igor Stasenko ha scritto: > > All i can say that callbacks for handling window events need > special handling: > - you basically should prevent VM to do event handling during > handling of event by callback. > > > Do I achieve that when starting in headless mode? > (no ui, no messages...) > > nope. VM will still run event loop (in ioProcessEvents) and do more or > less the same things. > that is another reason, btw to bring event loop handling to image level. > > -- > Best regards, > Igor Stasenko. Could I prevent VM (as you say, Igor) do process events using primitiveEventProcessingControl? I tried hacking NBFFICallback >> pvtEnter:stackPointer:primitiveMethod: by disabling events before block execution and enabling them after callback termination. Pharo still works for a while, then crashes.

Jan. 15, 2014

Re: [Pharo-dev] Recordable bug reports

by Igor Stasenko

cool idea On 15 January 2014 17:40, Sean P. DeNigris <sean(a)clipperadams.com> wrote: > Kind of a cool use for EventRecorderMorphâ¦ > > I noticed this method comment: > "In case there is a chance of not regaining control to stop > recording and > save a file, the EventRecorder can write directly to file as it is > recording. This is useful for capturing a sequence that results in a nasty > crash." > > You could record the steps to reproduce a bug with EventRecorderMorph, save > it to a tape file, and attach it to the bug report. > > > > ----- > Cheers, > Sean > -- > View this message in context: > http://forum.world.st/Recordable-bug-reports-tp4736970.html > Sent from the Pharo Smalltalk Developers mailing list archive at > Nabble.com. > > -- Best regards, Igor Stasenko.

Jan. 15, 2014

Re: [Pharo-dev] problems with RPackage and cie...

by Goubier Thierry

Le 15/01/2014 17:34, Esteban Lorenzano a Ã©crit : > > yeah, I knowâ¦ I was just pointing how it is. Not sure how it has to beâ¦ AFAIR, this was a requested feature too :S Ok, I leave the requested "feature" :) For now, I have a test case which exhibit a bug. I'm pushing it on 12655 if you want to have a look. Thierry -- Thierry Goubier CEA list Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s 91191 Gif sur Yvette Cedex France Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95

Jan. 15, 2014

Recordable bug reports

by Sean P. DeNigris

Kind of a cool use for EventRecorderMorphâ¦ I noticed this method comment: "In case there is a chance of not regaining control to stop recording and save a file, the EventRecorder can write directly to file as it is recording. This is useful for capturing a sequence that results in a nasty crash." You could record the steps to reproduce a bug with EventRecorderMorph, save it to a tape file, and attach it to the bug report. ----- Cheers, Sean -- View this message in context: http://forum.world.st/Recordable-bug-reports-tp4736970.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

Jan. 15, 2014