- Pharo-dev - lists.pharo.org

[Pharo-project] Hello everyone!
by Igor Stasenko Sept. 12, 2008

Sept. 12, 2008

I just subscribed and want to say hello to this list. :) At ESUG we discussed with Stephane and Markus some things in Squeak (and consequently in Pharo), which really need an overlook and better implementation. Stephane asked me, to think about overlooking Processes and concurrency in its current state. First, i must say that it is very fragile part of system because most things written w/o concurrency in mind. Last year i seen many issues related to concurrency: Semaphore , Delay, Process, Scheduler. Some of them resolved, some of them resolved in an ugly and hackish way - mostly as workarounds of lacking critical VM primitives which should guarantee atomicity for different purposes, and some of them is not resolved at all. Some things, which i noted and which i like/not like, can be found in squeak-dev list. But i will repeat them here, in single place: - Process should assume that it is running in ideal environment (e.g. in parallel to another processes w/o any preemption). Every bit of code, which requires synchronization should be based/take into account this assumption. a process should have three discrete states: - suspended - when initially created. - running - when you invoke resume. - terminated - when process no longer running and can't be resumed. It is wrong to consider a process which waiting on semaphore signal as suspended process. (see http://bugs.squeak.org/view.php?id=6822) Process which waits on semaphore is actually performing an action, while , by definition, a suspended process can't perform any action. Concerning process termination, it is also a can of worms to deal with: - process termination means stop doing anything in requested process, but this going into conflict with exception handling mechanism. The conflict comes from two things like #ensure: and #ifCurtailed: . Such mechanisms used to ensure that some code will have chance to run under any circumstances, regardless of the state of computation. It is interesting, where such code should run? In a process which initiated a terminate request, or in process which needs to be terminated by request. Most valid, of course - to run them in a terminating process. The problem, however, how to achieve that. Someone suggested to throw an exception, like ProcessTerminateException to indicate that the given process is going to be terminated. Such thing is also going into conflict with different points of view: a Process termination is usual functionality, provided for use, so why it should be considered as exceptional? Maybe, to solve the conflicts, and avoid using exceptions in an arguable way, it would be better to add support of process termination notification. So, each time process is going to terminate, it can notify all interesting parties of such event. Then, instead of using #ensure: to free critical system resources, one could use notification. As you can see, there is more questions than answers :) -- Best regards, Igor Stasenko AKA sig.

5 22

[Pharo-project] FileMan
by Alexandre Bergel Sept. 12, 2008

Sept. 12, 2008

Dear List, On SqueakMap there is a package called FileMan. Here is an example: Squeak's file libraries (Directory, DirectoryEntry, FileStream, etc.) are sometimes very confusing to use. FileMan provides a simple, refactored interfaces for manipulating files and directories by wrapping those classes. "Before" subDir := FileDirectory default directoryNamed: 'subDir'. subDir assureExistence. [str := subDir newFileNamed: 'file1'. str nextPutAll: 'Hello!'] ensure: [str close]. "After installation of FileMan" './subDir' asDirectoryEntry at: 'file2' put: 'Hello!' Maybe it worth having a look at it? Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

4 3

[Pharo-project] How to submit changes when SqueakSource is down ?
by Noury Bouraqadi Sept. 11, 2008

Sept. 11, 2008

Hi, SqueakSource is down again :-( Attached a monticello package with a serie of bugfixes and enhancements tested on pharo-dev of sept: *Bug Fixes -Transcript printing and self evaluating -Smalltalk printing and self evaluating -Instance Variable should never start by a capital *Enhancemnts -Monticello Package renaming -Prefix replacement for classes in a Monticello package Regards, Noury ------------------------------------------------------------------ Dr. Noury Bouraqadi - Enseignant/Chercheur Responsable de l'enseignement de l'informatique ARMINES - Ecole des Mines de Douai - Dept. I.A. http://vst.ensm-douai.fr/noury European Smalltalk Users Group Board http://www.esug.org ------------------------------------------------------------------

5 8

[Pharo-project] InterfaceDiscoverer
by Alexandre Bergel Sept. 11, 2008

Sept. 11, 2008

Dear List, I built a small extension to OB that display dependencies between class categories. Just load InterfaceDiscoverer from squeaksource, and click on a class category in an OB browser. Shall I add this package to the ScriptLoader>>loadOB? Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

4 11

Re: [Pharo-project] Your take on that: CollectionsTests-Ordered vs Collections-Ordered-Tests
by Bill Schwab Sept. 11, 2008

Sept. 11, 2008

I agree that packages should be promoted to real objects vs. the result of clever (ab)use of category names. That said, I frequently use a general-to-specific naming scheme for related classes on the grounds that it seems to help me when I get an alphabetical view of names. With that same logic, I suppose one could argue for just about any of the options described here :) Sorry. In addition to reified packages, I would like to float the idea of allowing multiple method categories to apply to any given method. That has been very useful in Dolphin. Bill Wilhelm K. Schwab, Ph.D. University of Florida Department of Anesthesiology PO Box 100254 Gainesville, FL 32610-0254 Email: bschwab(a)anest.ufl.edu Tel: (352) 273-6785 FAX: (352) 392-7029 >>> luc.fabresse(a)gmail.com 9/11/2008 3:09 AM >>> I understand the problem and I did not explain well but the solution of the Graphics package seems great : Graphics-Text Graphics-Files and GraphicsTests-Text GraphicsTests-Files Other packages like Graphics : CollectionTests-*, KernelTests-*, ... Packages with other naming conventions that do not seem great : File-Tests, Tests-*, Monticello-Tests Sure, a system that does not rely on string maching would be better. #Luc 2008/9/11 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> > > On Sep 11, 2008, at 9:35 AM, Luc Fabresse wrote: > > the convention <category>Tests[<-subcategory>] seems great IMO. >> e.g. Graphics-Text -> GraphicsTests-Text >> >> With this naming pattern it is possible to have 2 monticello packages : >> <category> and <category>Tests >> And it is possible to not load Tests packages in a deployed image. >> > > Yes it does not work if Graphics is composed of several subpackages > > With this I cannot have > > Collections-OrderedCollection > and > Collections-OrderedCollection-Tests > > separated from > > Collections-Set > and > Collections-Set-Tests > > For me I want to have packages that I can load independently and the tests > should be attached to the package (to be eventually not load with it). > >> >> > I really hate the fact that we feel everywhere category and string > matching. > > Stef > > > >> Luc >> >> 2008/9/11 Alexandre Bergel <alexandre(a)bergel.eu> >> Having a package name without any - will prevent Monticello from being >> confuse on some point. >> Stef and I already experienced that... >> >> I was never happy in inserting a - in a package name. >> >> Alexandre >> >> >> >> On 11 Sep 2008, at 07:23, Damien Cassou wrote: >> >> On Wed, Sep 10, 2008 at 9:32 PM, StÃ©phane Ducasse >> <stephane.ducasse(a)inria.fr> wrote: >> Hi >> >> I noticed that we have CollectionsTests-Ordered >> kind of package name and I would prefer to get Collections-Ordered-Tests >> What do you think. I bit of regularity would help. >> >> Collections-Test-* categories belong to the Collections-Test package >> on monticello. If you rename it to Collections-Ordered-Tests, you will >> put everything in the Collections package (because there is no >> Collections-Ordered monticello package). >> >> >> -- >> Damien Cassou >> Peter von der AhÃ©: Â«I'm beginning to see why Gilad wished us good >> luckÂ». (http://blogs.sun.com/ahe/entry/override_snafu) >> _______________________________________________ >> 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 >> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >> >> >> >> >> >> >> >> _______________________________________________ >> 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 >

1 0

[Pharo-project] Your take on that: CollectionsTests-Ordered vs Collections-Ordered-Tests
by StÃ©phane Ducasse Sept. 11, 2008

Sept. 11, 2008

Hi I noticed that we have CollectionsTests-Ordered kind of package name and I would prefer to get Collections-Ordered-Tests What do you think. I bit of regularity would help. Stef

4 7

[Pharo-project] Fwd: [squeak-dev] Re: FFI and FreeTypePlus problem?
by StÃ©phane Ducasse Sept. 10, 2008

Sept. 10, 2008

================================================================ Hi there, I'm using Installer to make a custom 3dot10 7159. The first things I do in the script is to load from universe FreeTypePlus (wich brings FFI). Then I've load LPF. If I do that it never ends due to Squeak not being able of resolving a name. NetNameResolver addressForName: 'installer.pbwiki.com/f/LPF.st' Before loading the FreeTypePlus, evaluating the previous sentence has an instant response of the address. But if I do load the FreeTypePlus package it never ends. Maybe an FFI issue? I'm in Win32 using SqueakFFIPrims.dll from 8/30/2007 Squeak.exe from 6/5/2007 any clue? thanks Sebastian Sastre PD: here is the beginning of the script: Utilities informUser: 'Upgrading universe packages' during: [ UUniverse systemUniverse updatePackagesViaWWW ]. (UUniverse systemUniverse newestPackageNamed: 'FreeType Plus') ifNotNilDo: [ :package| |inst| inst := UGlobalInstaller universe: UUniverse systemUniverse. inst planToInstallPackage: package. inst doInstall ]. (HTTPSocket httpGet: 'installer.pbwiki.com/f/LPF.st') readStream fileIn. ================================================================ I already created a bug report for this, Andreas may be able to tell you more: http://bugs.squeak.org/view.php?id=6980 As a workaround you can can use FFI-Kernel-ar6.mcz instead of FFI-Kernel-ar-8.mcz. Bye Torsten ================================================================ > From: Andreas Raab <andreas.raab(a)gmx.de> > Date: March 27, 2008 10:00:21 PM CEST > To: The general-purpose Squeak developers list <squeak-dev(a)lists.squeakfoundation.org > > > Subject: [squeak-dev] Re: FFI and FreeTypePlus problem? > Reply-To: The general-purpose Squeak developers list <squeak-dev(a)lists.squeakfoundation.org > > > > Yeah, I had forgotten about the problem. The issue comes from a call > to Smalltalk>>recreateSpecialObjectsArray (which is required for the > FFI to work correctly) and which accidentally nukes the external > semaphore table. The fix is simple: Replace the following line in > SystemDictionary>>recreateSpecialObjectsArray > > newArray at: 39 put: Array new. > > by > "Preserve external semaphores when recreating splObjs" > newArray at: 39 put: (self specialObjectsArray at: 39). > > Also, saving and restarting the image after loading the FFI is good > temporary workaround for this problem. > > Cheers, > - Andreas > > Torsten Bergmann wrote: >> I already created a bug report for this, Andreas may be able to >> tell you more: >> http://bugs.squeak.org/view.php?id=6980 >> As a workaround you can can use FFI-Kernel-ar6.mcz >> instead of FFI-Kernel-ar-8.mcz. >> Bye >> Torsten > > >

2 2

[Pharo-project] Wiki page rename
by Bill Schwab Sept. 9, 2008

Sept. 9, 2008

Hello, Anybody know how to rename a page on the Wiki? Is it as crude as add/copy/delete? Hopefully there is a better way?? Bill Wilhelm K. Schwab, Ph.D. University of Florida Department of Anesthesiology PO Box 100254 Gainesville, FL 32610-0254 Email: bschwab(a)anest.ufl.edu Tel: (352) 273-6785 FAX: (352) 392-7029

4 3

Re: [Pharo-project] Hello everyone!
by Bill Schwab Sept. 9, 2008

Sept. 9, 2008

Igor, THANKS for thinking about concurrency. Processes need to be sound. Re #ensure:, we might consider "unsuspending/unterminating" the process long enough to run the block. That way the block runs in the thread it expects, after which, there would hopefully be nothing else to do?? Re using exceptions to indicate impending termination, that sounds more like an event (potentially many consumers). If it is simply a system-level trick, then I suppose it's fine. Bill Wilhelm K. Schwab, Ph.D. University of Florida Department of Anesthesiology PO Box 100254 Gainesville, FL 32610-0254 Email: bschwab(a)anest.ufl.edu Tel: (352) 273-6785 FAX: (352) 392-7029

2 1

[Pharo-project] SCriptLoader>>InstallingInstaller
by Hilaire Fernandes Sept. 7, 2008

Sept. 7, 2008

With the lateset update. There is something uncorrect in the package name I guess ? SCriptLoader>>InstallingInstaller [...] fileToLoad := files detect: [:aFile | aFile beginsWith: 'InstallerAll.-md.1'] ifNone: []. -- http://blog.ofset.org/hilaire

3 3