- Pharo-dev - lists.pharo.org

[Pharo-project] 64bit VMs some thoughts.
by John M McIntosh Dec. 3, 2009

Dec. 3, 2009

So you sit there smug about the fact you built a 64bit VM, likely for hosting on your 64bit Linux OS. {Or the unix one for Darwin, or that new fangled cocoa one} However it's possible that it's running 1/3 the performance of the 32bit VM. Did you check? Thought not... So let's talk. Are you using the gnuifed version of interp.c? If you don't know, well go check. Are you using GCC 4.1 or higher? The interpreter loop is highly tuned monster that suffers from compiler optimization issues. With careful tuning parms as found in the macintosh xcode build project for the carbon VM using gcc 4.0 you'll get the most optimum performance. GCC 4.2+ ? Michael Rueger and I spent a few days attempting to get good performance out of GCC 4.2 WITHOUT success. I think that can account for at least a 33% slowdown. So where does the other 33% slowdown come from? Well when we compile the VM in 64bit to use a 32bit image each reference to an oops requires us to add a 64bit memory start address to the 32bit oops number to resolve to a 64bit memory address. Unfortunately GCC 4.2 growls, and produces the lousiest code possible to do this. Maybe higher versions of GCC are better? Anyone care to test? So some solutions. (a) Ensure the squeak oops memory block loads within the 0-4GB address space. See pagezero size for Darwin. Then alter the logic a bit so that sqMemoryBase is zero and that the squeak memory accessors don't do the add of sqMemoryBase=0 to the oops address. Although you might have to use GCC 4.2 you'll run 100% faster. (b) Use the (non-free) Intel compiler -- =========================================================================== John M. McIntosh <johnmci(a)smalltalkconsulting.com> Twitter: squeaker68882 Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com ===========================================================================

1 0

[Pharo-project] On issue Issue 1155: "Renaming of a class do not changes title of a System Browser"
by csrabak＠bol.com.br Dec. 3, 2009

Dec. 3, 2009

Pertaining the bug #1155 http://code.google.com/p/pharo/issues/detail?id=1155, it seems to me that we need to register the change for the class name so the SystemWindow gets updated: O2PackageBrowser class>>registerForNotifications Smalltalk at: #SystemChangeNotifier ifPresent: [:cls | (cls uniqueInstance) noMoreNotificationsFor: self; notify: self ofSystemChangesOfItem: #class change: #Added using: #classModified:; notify: self ofSystemChangesOfItem: #class change: #Modified using: #classModified:; notify: self ofSystemChangesOfItem: #class change: #Renamed using: #classModified:; notify: self ofSystemChangesOfItem: #class change: #Commented using: #classModified:; notify: self ofSystemChangesOfItem: #class change: #Recategorized using: #classMoved:; notify: self ofSystemChangesOfItem: #class change: #Removed using: #classModified:; notify: self ofSystemChangesOfItem: #method change: #Added using: #methodModified:; notify: self ofSystemChangesOfItem: #method change: #Modified using: #methodModified:; notify: self ofSystemChangesOfItem: #method change: #Recategorized using: #methodModified:; notify: self ofSystemChangesOfItem: #method change: #Removed using: #methodModified: ]. However I could not find how to assemble a #notify:chnage:using: message in order to have the model (O2PackageBrowser) notified. . . If some more knowledgeable on this hierarchy can enlighten me I think we can solve this easily, no? -- Cesar Rabak

1 0

[Pharo-project] is this correct ?
by Mariano Martinez Peck Dec. 2, 2009

Dec. 2, 2009

Hi! Look this peace of code: | instance | instance := TestCase new. Transcript show: TestCase allInstances size; cr. instance become: 'aString'. Transcript show: TestCase allInstances size; cr. The size of TestCase instances is always the same. Although I used become: (and changed to an instance of another class) to one of its instances. So, my main question is, is this correct ? Or the message allInstances to TestCase should not take into account the instance I sent the message become: ? Thanks in advance. Mariano

4 4

[Pharo-project] Good article on font rendering and subpixel
by StÃ©phane Ducasse Dec. 2, 2009

Dec. 2, 2009

> From: Damien Pollet <damien.pollet(a)inria.fr> > > > http://www.antigrain.com/research/font_rasterization/index.html#FONT_RASTER… > -- > Damien Pollet > type less, do more [ | ] http://people.untyped.org/damien.pollet > >

2 1

[Pharo-project] Issue # 1498 "Collections should not accept non integer indexes"
by csrabak＠bol.com.br Dec. 2, 2009

Dec. 2, 2009

I'm trying to follow the directions given at: http://code.google.com/p/pharo/wiki/HowToContribute. However, after the SLICE asks for the first comment, it freezes the image and the small window "Writing definitions" stays at 50% progress and nothing happens then. So I'll resort to the alternative plan, which is to send the changeset to this list. On a core image update #11067 the Collection tests are all green. Running all tests show one error and two failures not related to the changed methods. -- Cesar Rabak

3 5

Re: [Pharo-project] [Moose-dev] Re: popupView: is now working
by Alexandre Bergel Dec. 2, 2009

Dec. 2, 2009

> The problem is that if you put it in the rendering code this will > basically get invoked every 30 ms. A more practical solution is to > use haltOnce. This is the same as halt, only that it is guarded by a > global flag. Just make sure that you enable it from the Debug world > menu. I am not expert in Morphic, but I feel there is a bug in how exceptions in Morph are handled. Create a subclass of Morph and put a "self halt" (or "1/0") in the drawOn: method. Open the morph and it will appear in red, telling that there is an error in the render loop. Now, for some reason, when blocks are heavily, the image freeze. I tried to isolate the error (i.e., without Mondrian), but I could not figure out how to do it. Alexandre > > Cheers, > Doru > > > On 2 Dec 2009, at 13:10, Johan Fabry wrote: > >> Hi Stef, >> this is the same behavior as what I reported in http://code.google.com/p/pharo/issues/detail?id=1525 >> and http://code.google.com/p/moose-technology/issues/detail?id=237 >> How to reproduce is also there. >> On 02 Dec 2009, at 08:30, StÃ©phane Ducasse wrote: >> >>> Do you have a example that we can reproduce? >>> >>> Stef >>> >>> On Dec 1, 2009, at 9:23 PM, Alexandre Bergel wrote: >>> >>>>> Regarding the problem with the number of arguments in the block. >>>>> Why is it crashing the image? >>>> >>>> The image is not crashing actually. It just hangs on forever >>>> without giving a chance to interrupt the process. >>>> >>>>> Anyway, you might want to use an approach similar to the >>>>> implementation of glamourValue: in Glamour. >>>> >>>> It could be a nice fix, but clearly not a general solution. The >>>> image hangs because of a bug in Squeak/Pharo. For some reason, >>>> the exception thrown by the virtual machine is not properly caught. >>>> This is a hard problem that will need to be fixed some days. >>>> >>>> Alexandre >> >> -- >> Johan Fabry >> jfabry(a)dcc.uchile.cl - http://dcc.uchile.cl/~jfabry >> PLEIAD Lab - Computer Science Department (DCC) - University of Chile >> >> >> >> >> _______________________________________________ >> Moose-dev mailing list >> Moose-dev(a)iam.unibe.ch >> https://www.iam.unibe.ch/mailman/listinfo/moose-dev > > -- > www.tudorgirba.com > > "Relationships are of two kinds: those we choose and those that > happen. They both matter." > > > > > > _______________________________________________ > Moose-dev mailing list > Moose-dev(a)iam.unibe.ch > https://www.iam.unibe.ch/mailman/listinfo/moose-dev > -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

1 0

[Pharo-project] [Pharo-Project] Monticello package corrupted
by Hernan Wilkinson Dec. 2, 2009

Dec. 2, 2009

Hi, I'm moving code from Pharo to GLASS using Monticello, but in some packages the code that is read by GLASS is not correct, it has 0s in it. For example, instead of 'to' it has 't o ' (where spaces are zeros) If I try to unzip the monticello package, I get a CRC error on the source file... Does anybody heard about this error before? Thanks, Hernan.

2 1

[Pharo-project] Time asMilliseconds...
by Hernan Wilkinson Dec. 2, 2009

Dec. 2, 2009

is not in the image anymore...is there a reason? Henan.

2 4

[Pharo-project] [update 1.1] #11068
by Marcus Denker Dec. 2, 2009

Dec. 2, 2009

11068 ----- Issue 1534: some more random cleans... Issue 1531: Integer>>#asFraction (only a typo in the comment) Issue 840: MorphViewHierarchyBrowser code Issue 1535: Time>>asMilliSeconds missing ======= brings a new useful tool for understanding morphic: World exploreViewHierarchy (a menu can "flash" the morph to show where it is on the screen). ========== ... removes lots of useless preferences.... "The paradox of Choice": Less is more. Preferences removePreference: #balloonHelpInMessageLists. Preferences removePreference: #projectViewsInWindows. Preferences removePreference: #projectZoom. Preferences removePreference: #automaticKeyGeneration. Preferences removePreference: #oliveHandleForScriptedObjects. Preferences removePreference: #canRecordWhilePlaying. Preferences removePreference: #fenceSoundEnabled. Preferences removePreference: #includeSoundControlInNavigator. Preferences removePreference: #projectsSentToDisk. Preferences removePreference: #menuButtonInToolPane.

2 1

[Pharo-project] Committing resource files with Monticello or along Monticello packages
by Simon Denier Dec. 2, 2009

Dec. 2, 2009

Hi there I would like to know if you people have some tricks or clever strategies to commit resource files (binary, icons, big text files) along monticello packages. And how do you load such files along monticello packages. My use case: I have a big text file with data, which I use as a model to run some tests. I don't want to store this information in the body of a method. -- Simon

2 1