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

Re: [Pharo-project] how to remove World?

by Pavel Krivanek

Hi, you do not need only the DummyUIManager but some modifications to MorphicUIManager from Pharo-Kernel. You should start with the scripts: DummyUIManager.st MorphicUIManager.st patch.st unload.st - here the UIProcess is stopped There's updated version of Pharo-Kernel here: http://gforge.inria.fr/frs/download.php/26462/PharoKernel-1.1-11229-UNSTABL… However it's still very dirty and a lot of things will change soon. Cheers, -- Pavel On Thu, Feb 25, 2010 at 1:49 PM, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > Ok, I tried passing a .st by command line: > > UIManager default: DummyUIManager new. > > Project uiProcess terminate. > > SmalltalkImage current snapshot: true andQuit: true. > > but I fail miserably :'( > (it stops do anything after "Project uiProcess terminate"... and the > same using "Project uiProcess suspend") > > Any idea of how can I proceed from here? (I confess: I'm totally lost here) > > Cheers, > Esteban > > > On 2010-02-23 06:44:18 -0300, Mariano Martinez Peck > <marianopeck(a)gmail.com> said: > >> >> >> I think that maybe it is a good idea to use the DummyUIManager and the >> MorphicUIManager Â from Pavel MinimalImage. >> >> We can put those classes in a repository called PharoMinimal or something >> similar. Then, you will be able to load both classes, run your image >> headless and before, evaluate something to set those classes as the current= >> . >> >> Just an idea... >> >> On Tue, Feb 23, 2010 at 10:18 AM, St=E9phane Ducasse < >> stephane.ducasse(a)inria.fr> wrote: >> >>> this is a really good experiment >>> please continue >>> and let us know the results. >>> Stef >>> On Feb 22, 2010, at 8:21 PM, Esteban Lorenzano wrote: >>> >>>> I tried by starting pharo with a script: >>>> >>>> Project uiProcess suspend. >>>> Smalltalk >>>> snapshot: true >>>> andQuit: true. >>>> >>>> ...and the processing stop working after the #suspend (no >>>> snapshot:andQuit: happens :( ) >>>> >>>> On 2010-02-22 13:43:05 -0300, Lukas Renggli >>>> <renggli(a)gmail.com> said: >>>> >>>>>> Is that also necessary for a headless Pharo/Squeak? >>>>> >>>>> No, it is not necessary. It saves a few CPU cycles because the event >>>>> polling stops, the #step methods on the visible morphs are not called >>>>> anymore, and the display is not updated any longer. >>>>> >>>>> Programmatically you can still do everything (headless or >>>>> non-headless). For example you can open morphs and move them around, >>>>> you just won't see a change on the screen. Code that have some >>>>> interaction with the GUI thread (like the debugger) might deadlock (or >>>>> hang) the system. >>>>> >>>>> Lukas >>>> >>>> >>>> >>>> >>>> _______________________________________________ >>>> Pharo-project mailing list >>>> Pharo-project(a)lists.gforge.inria.fr >>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >>> >>> >>> _______________________________________________ >>> Pharo-project mailing list >>> Pharo-project(a)lists.gforge.inria.fr >>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >>> >> >> >> I think that maybe it is a good idea to use the DummyUIManager and the Morp= >> hicUIManager=A0 from Pavel MinimalImage. We can put those classes in= >> Â a repository called PharoMinimal or something similar. Then, you will be a= >> ble to load both classes, run your image headless and before, evaluate some= >> thing to set those classes as the current. >> Just an idea... <div class=3D"gmail_quote">On Tue, Feb 23, 2010 = >> at 10:18 AM, St=E9phane Ducasse <<a href=3D"mailto:ste= >> phane.ducasse(a)inria.fr">stephane.ducasse(a)inria.fr</a>> >> wrote: = >> <blockquote class=3D"gmail_quote" style=3D"border-left: 1px solid rgb(204, = >> 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"> >> this is a really good experiment >> please continue >> and let us know the results. >> Stef >> <div><div></div><div class=3D"h5">On Feb 22, 2010, at 8:21 PM, Esteban Lore= >> nzano wrote: >> >> > I tried by starting pharo with a script: >> > >> > Project uiProcess suspend. >> > Smalltalk >> > =A0 =A0 =A0 snapshot: true >> > =A0 =A0 =A0 andQuit: true. >> > >> > ...and the processing stop working after the #suspend (no >> > snapshot:andQuit: happens :( ) >> > >> > On 2010-02-22 13:43:05 -0300, Lukas Renggli >> > <<a >> href=3D"mailto:renggli@gmail.com">renggli(a)gmail.com</a>> >> sai= >> d: >> > >> >>> Is that also necessary for a headless Pharo/Squeak? >> >> >> >> No, it is not necessary. It saves a few CPU cycles because the eve= >> nt >> >> polling stops, the #step methods on the visible morphs are not cal= >> led >> >> anymore, and the display is not updated any longer. >> >> >> >> Programmatically you can still do everything (headless or >> >> non-headless). For example you can open morphs and move them aroun= >> d, >> >> you just won't see a change on the screen. Code that have some= >> >> >> interaction with the GUI thread (like the debugger) might deadlock= >> Â (or >> >> hang) the system. >> >> >> >> Lukas >> > >> > >> > >> > >> > _______________________________________________ >> > Pharo-project mailing list >> > <a >> href=3D"mailto:Pharo-project@lists.gforge.inria.fr">Pharo-project@l= > ists.gforge.inria.fr</a> > > >> >> <a href=3D"http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo= >> -project" target=3D"_blank">http://lists.gforge.inria.fr/cgi-bin/mailman/li= >> stinfo/pharo-project</a> >> >> >> _______________________________________________ >> Pharo-project mailing list >> <a >> href=3D"mailto:Pharo-project@lists.gforge.inria.fr">Pharo-project(a)lists.= > gforge.inria.fr</a> > <a >> >> href=3D"http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-proj= >> ect" target=3D"_blank">http://lists.gforge.inria.fr/cgi-bin/mailman/listinf= >> o/pharo-project</a> >> </div></div></blockquote></div> >> >> >> >> >> _______________________________________________ >> Pharo-project mailing list >> Pharo-project(a)lists.gforge.inria.fr >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >

Feb. 25, 2010

Re: [Pharo-project] Building Pharo on a Core 1.1

by Alexandre Bergel

Hi Dale, Here is the PharoDebug.log I would be pleased to give another try. Cheers, Alexandre On 24 Feb 2010, at 20:36, Dale Henrichs wrote: > > ----- "Alexandre Bergel" <alexandre(a)bergel.eu> wrote: > > | But I get an error raised by MetacelloPackageSpec>>instVarNames:put: > | > | apparently some reflective operations try to access a non-existing > | variable. > | > | I know that Pharo is meant to be built from a stable core. But I > would > | like to benefit from the snappy interface and the nice menu. Anyone > | tried ? > > Odd... there's one spot where Metacello uses instVarNamed:put: and > the names being used should correspond 1 to 1 with instance > variables of the specs. > > Do you have more detail about the error? The receiver's class and > the value and class of the args to instVarNamed:put: might help. > > Dale > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

Feb. 25, 2010

Re: [Pharo-project] how to remove World?

by Esteban Lorenzano

Ok, I tried passing a .st by command line: UIManager default: DummyUIManager new. Project uiProcess terminate. SmalltalkImage current snapshot: true andQuit: true. but I fail miserably :'( (it stops do anything after "Project uiProcess terminate"... and the same using "Project uiProcess suspend") Any idea of how can I proceed from here? (I confess: I'm totally lost here) Cheers, Esteban On 2010-02-23 06:44:18 -0300, Mariano Martinez Peck <marianopeck(a)gmail.com> said: > > > I think that maybe it is a good idea to use the DummyUIManager and the > MorphicUIManager from Pavel MinimalImage. > > We can put those classes in a repository called PharoMinimal or something > similar. Then, you will be able to load both classes, run your image > headless and before, evaluate something to set those classes as the current= > . > > Just an idea... > > On Tue, Feb 23, 2010 at 10:18 AM, St=E9phane Ducasse < > stephane.ducasse(a)inria.fr> wrote: > >> this is a really good experiment >> please continue >> and let us know the results. >> Stef >> On Feb 22, 2010, at 8:21 PM, Esteban Lorenzano wrote: >> >>> I tried by starting pharo with a script: >>> >>> Project uiProcess suspend. >>> Smalltalk >>> snapshot: true >>> andQuit: true. >>> >>> ...and the processing stop working after the #suspend (no >>> snapshot:andQuit: happens :( ) >>> >>> On 2010-02-22 13:43:05 -0300, Lukas Renggli >>> <renggli(a)gmail.com> said: >>> >>>>> Is that also necessary for a headless Pharo/Squeak? >>>> >>>> No, it is not necessary. It saves a few CPU cycles because the event >>>> polling stops, the #step methods on the visible morphs are not called >>>> anymore, and the display is not updated any longer. >>>> >>>> Programmatically you can still do everything (headless or >>>> non-headless). For example you can open morphs and move them around, >>>> you just won't see a change on the screen. Code that have some >>>> interaction with the GUI thread (like the debugger) might deadlock (or >>>> hang) the system. >>>> >>>> Lukas >>> >>> >>> >>> >>> _______________________________________________ >>> Pharo-project mailing list >>> Pharo-project(a)lists.gforge.inria.fr >>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> >> >> _______________________________________________ >> Pharo-project mailing list >> Pharo-project(a)lists.gforge.inria.fr >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> > > > I think that maybe it is a good idea to use the DummyUIManager and the Morp= > hicUIManager=A0 from Pavel MinimalImage. We can put those classes in= > a repository called PharoMinimal or something similar. Then, you will be a= > ble to load both classes, run your image headless and before, evaluate some= > thing to set those classes as the current. > Just an idea... <div class=3D"gmail_quote">On Tue, Feb 23, 2010 = > at 10:18 AM, St=E9phane Ducasse <<a href=3D"mailto:ste= > phane.ducasse(a)inria.fr">stephane.ducasse(a)inria.fr</a>> > wrote: = > <blockquote class=3D"gmail_quote" style=3D"border-left: 1px solid rgb(204, = > 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"> > this is a really good experiment > please continue > and let us know the results. > Stef > <div><div></div><div class=3D"h5">On Feb 22, 2010, at 8:21 PM, Esteban Lore= > nzano wrote: > > > I tried by starting pharo with a script: > > > > Project uiProcess suspend. > > Smalltalk > > =A0 =A0 =A0 snapshot: true > > =A0 =A0 =A0 andQuit: true. > > > > ...and the processing stop working after the #suspend (no > > snapshot:andQuit: happens :( ) > > > > On 2010-02-22 13:43:05 -0300, Lukas Renggli > > <<a > href=3D"mailto:renggli@gmail.com">renggli(a)gmail.com</a>> > sai= > d: > > > >>> Is that also necessary for a headless Pharo/Squeak? > >> > >> No, it is not necessary. It saves a few CPU cycles because the eve= > nt > >> polling stops, the #step methods on the visible morphs are not cal= > led > >> anymore, and the display is not updated any longer. > >> > >> Programmatically you can still do everything (headless or > >> non-headless). For example you can open morphs and move them aroun= > d, > >> you just won't see a change on the screen. Code that have some= > > >> interaction with the GUI thread (like the debugger) might deadlock= > (or > >> hang) the system. > >> > >> Lukas > > > > > > > > > > _______________________________________________ > > Pharo-project mailing list > > <a > href=3D"mailto:Pharo-project@lists.gforge.inria.fr">Pharo-project@l= ists.gforge.inria.fr</a> > > > <a href=3D"http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo= > -project" target=3D"_blank">http://lists.gforge.inria.fr/cgi-bin/mailman/li= > stinfo/pharo-project</a> > > > _______________________________________________ > Pharo-project mailing list > <a > href=3D"mailto:Pharo-project@lists.gforge.inria.fr">Pharo-project(a)lists.= gforge.inria.fr</a> <a > > href=3D"http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-proj= > ect" target=3D"_blank">http://lists.gforge.inria.fr/cgi-bin/mailman/listinf= > o/pharo-project</a> > </div></div></blockquote></div> > > > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Feb. 25, 2010

[Pharo-project] Delay startTimerInterruptWatcher in ScriptLoader >> cleanUpMethods -> DNU

by Mariano Martinez Peck

Try to evaluate ScriptLoader new cleanUpMethods and you will see that in ScriptLoader >> cleanUpMethods you do: Delay startTimerInterruptWatcher. and there is the DNU as startTimerInterruptWatcher is never implemented. Solutions ? 1) Remove Delay startTimerInterruptWatcher. from cleanUpMethods 2) Maybe the message should be stopTimerEventLoop ?? I have no idea The only thing I found in the issues was this: http://code.google.com/p/pharo/issues/detail?id=1424 In squeak 3.9 the method is implemented, and like this: startTimerInterruptWatcher "Reset the class variables that keep track of active Delays and re-start the timer interrupt watcher process. Any currently scheduled delays are forgotten." "Delay startTimerInterruptWatcher" | p | self primSignal: nil atMilliseconds: 0. TimingSemaphore == nil ifFalse: [TimingSemaphore terminateProcess]. TimingSemaphore _ Semaphore new. AccessProtect _ Semaphore forMutualExclusion. SuspendedDelays _ SortedCollection sortBlock: [:d1 :d2 | d1 resumptionTime <= d2 resumptionTime]. ActiveDelay _ nil. p _ [self timerInterruptWatcher] newProcess. p priority: Processor timingPriority. p resume.

Feb. 25, 2010

[Pharo-project] [update 1.1] #11233

by StÃ©phane Ducasse

11233 ----- - Issue 2055: Checking squeak fixes to SUnitGUI - Issue 2071: subclasses and withAllSubclasses return a array - Issue 2076: Clean Exceptions - Issue 2072: clean RealEstateAgent Stef

Feb. 25, 2010

[Pharo-project] We should run + fix the tests in 1.1

by StÃ©phane Ducasse

... testserver, where are you?

Feb. 25, 2010

Re: [Pharo-project] Building Pharo on a Core 1.1

by StÃ©phane Ducasse

On Feb 24, 2010, at 11:46 PM, Torsten Bergmann wrote: >> But I would like to benefit from the snappy interface and the nice menu. > > Yes - but since we changed more than just the nice menu I think > the whole config and anything that gets loaded need to be revised for 1.1. oh yesssssssssssss 1.1 is busy. Now alex can give a try. There are tests to run and propose fixes too. > > Bye > T. > > > -- > Sicherer, schneller und einfacher. Die aktuellen Internet-Browser - > jetzt kostenlos herunterladen! http://portal.gmx.net/de/go/atbrowser > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Feb. 25, 2010

Re: [Pharo-project] hex

by Nicolas Cellier

2010/2/25 <csrabak(a)bol.com.br>: > Em 24/02/2010 16:03, Nicolas Cellier < nicolas.cellier.aka.nice(a)gmail.com > escreveu: > 2010/2/24 Â : > >> > Stef, >> > Squeak Â prints Â in Â lowercase Â and Â reads Â *correctly* Â only Â in >> > uppercase. . . >> > >> > -- Cesar Rabak >> > >> Â I Â think recent Â trunk version Â read both Â correctly. Â Did Â you try >> recently? > > Nicolas, > > As I reported in an earlier post to this list, I used: In Squeak (3.10.2-7179), I've not been downloading newer images for Squeak for long time indeed. If newer Squeak versions work differently, I cannot ascertain that, others will have to perform the test as well. > > HTH, > > -- > Cesar Rabak > Ah, OK, so you should maybe conjugate with preterit :) Nicolas

Feb. 25, 2010

Re: [Pharo-project] hex

by csrabak＠bol.com.br

Stef, I've posted this yesterday, but here it is: In Pharo '16r7A2BD507' asNumber = 2049692935 '16r7a2bd507' asNumber = 7 In Squeak (3.10.2-7179) '16r7A2BD507' asNumber = 2049692935 '16r7a2bd507' asNumber = 7 Obviously I printed in the Workspace (Shout) -- Cesar Rabak Em 24/02/2010 19:41, StÃ©phane Ducasse < stephane.ducasse(a)inria.fr > escreveu: On Feb 24, 2010, at 6:47 PM, csrabak(a)bol.com.br wrote: > Stef, > > Squeak prints in lowercase and reads *correctly* only in uppercase. . . can you post the expression you use? Are you talking about byteArray? or plain number? > > -- > Cesar Rabak > > > Em 24/02/2010 04:05, StÃ©phane Ducasse < stephane.ducasse(a)inria.fr > escreveu: > >> The main reason for not using lowercase in squeak was Dan extension of >> number with radix plus exponent. >> 16r8e7 -> 16r8 * (16 raisedTo: 7) >> and the Float variations... >> >> If you decided to abandon this notation, then go with either upper or lower. >> But a bit of consistency could help... > > I'm confused what is the situation now in squeak? > Because it seems that hex in squeak now prints and read lowercase? > So did > 16r8e7 and float > so will someone fix squeak to be consistent. > I'm a bit confused by the decision process in squeak (even if I should not worried too much). > > Stef > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Feb. 25, 2010

Re: [Pharo-project] hex

by csrabak＠bol.com.br

Em 24/02/2010 16:03, Nicolas Cellier < nicolas.cellier.aka.nice(a)gmail.com > escreveu: 2010/2/24 : > > Stef, > > Squeak prints in lowercase and reads *correctly* only in > > uppercase. . . > > > > -- Cesar Rabak > > > I think recent trunk version read both correctly. Did you try > recently? Nicolas, As I reported in an earlier post to this list, I used: In Squeak (3.10.2-7179), I've not been downloading newer images for Squeak for long time indeed. If newer Squeak versions work differently, I cannot ascertain that, others will have to perform the test as well. HTH, -- Cesar Rabak

Feb. 25, 2010