- Pharo-dev - lists.pharo.org

[Pharo-project] KernelTests green
by Oscar Nierstrasz June 12, 2008

June 12, 2008

I have posted some patches to the inbox to make the Kernel Number tests run, and in the Issues page I propose to delete a bunch of nonsensical testHash tests. With these changes the Kernel tests should all be green. - on

4 11

[Pharo-project] Fwd: Comment on BrokenTests in pharo
by Pharo Smalltalk June 12, 2008

June 12, 2008

My apologies if this message is forwarded to the list twice. Victor Rodriguez. ---------- Forwarded message ---------- From: <codesite-noreply(a)google.com> Date: Wed, Jun 11, 2008 at 12:05 PM Subject: Comment on BrokenTests in pharo To: pharo.smalltalk(a)gmail.com Comment by oscar.nierstrasz: FloatTests>>#testNaN5 currently reads as follows: {{{ testNaN5 "self assert: (Float nan asIEEE32BitWord printPaddedWith: $0 to: 32 base: 2) = '01111111110000000000000000000000'." ... }}} I believe this is over-specified. Instead it should just read as follows: {{{ testNaN5 self assert: ((Float nan asIEEE32BitWord printPaddedWith: $0 to: 32 base: 2) copyFrom: 2 to: 9) = '11111111'. ... }}} This would be consistent with the NaN spec (according to wikipedia). However it does not explain why these tests pass in 3.9 and 3.10. Have there been changes in the VM to the way that Floats are represented? For more information: http://code.google.com/p/pharo/wiki/BrokenTests

1 0

[Pharo-project] New pharo-dev based on latest release
by Damien Cassou June 11, 2008

June 11, 2008

I've just published a new pharo-dev based on 10036: https://gforge.inria.fr/frs/download.php/4978/pharo0.1-10036dev08.06.2.zip -- 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)

3 6

[Pharo-project] HaveALookAt
by Norbert Hartl June 11, 2008

June 11, 2008

I think the link to the packages issue should point to http://www.squeaksource.com/PackagesForSqueak/ Norbert

4 3

[Pharo-project] MVC removal step2 investigations
by Alain Plantec June 11, 2008

June 11, 2008

Hi all, below is given a minute that briefly describes what I suggest for MVC removal. It's an aggressive update especially because a lot of classes are removed. (see all *ST80* system categories before and after having applied the patch and see Controller hierarchy also) I've attached the .cs. Maybe you can make a try and give your opinion. thanks alain MVC removal step2 proposition 1 : * Object>>couldOpenInMorphic : senders updated and method set as deprecated (as for Project>>isMorphic and SystemDictionary>>isMorphic) * Remove method openAsMVC and openInMVC * ST80-Editors system category seems removeable o BitEditor, FormEditor and al : are strongly related to MVC and seems not used * ScheduledControllers : o Remove all methods which are referencing it, this leads to remove class ControlManager; o Remove ControlManager refs in SystemDictionary>>initialize and o cleanup StartUpList and ShutDownList SystemDictionary class variables with a script. * ST80-Framework system category seems removeable o View class seems removable (see ST80-Views system category point) o Then, ModalController, NoController are not referenced anymore, they can be removed o WindowingTransformation is only used by GraphicSymbolInstance which is removable (see ST80-Symbols system category point) * ST80-Views system category can be removed: o StandardSystemView>>Comment: "I represent a view that has a label above its top left corner. The text in the label identifies the kind of view. In addition to a label, I add control over the maximum and minimum size of the display box of my instance. My default controller is StandardSystemController. The elements of ScheduledControllers, the sole instance of ControlManager, are usually controllers for instances of me". o StandardSystemView hierarchy seems removable, (subclasses are ColorSystemView, ModalSystemWindowView, PluggableFileListView and ProjectView ) o MorphWorldView class removed * ST80-Pluggable Views system category o PluggableListView removed * ST80-Symbols system category seems removable o GraphicSymbol and GraphicSymbolInstance are not refed and have no instance, seems removable * WorldViewModel, MVCWiWPasteUpMorph, MorphWorldController and CautiousModel: all these classes are not used, to be removed o MorphicExtras-ST80 system category is then removed * ST80-Support system category o StandardSystemController class>>initialize removed: it only initialize unused class variables HBorderCursor, ScheduledBlueButtonMenu ScheduledBlueButtonMessages and VBorderCursor (these class variables are removed) o StandardSystemController: remove all methods that are related to view (mvc view) o StandardSystemController: class seems to be unused, therefore removed o FillInTheBlankController: class removed o ColorSystemView: class removed * ST80-Kernel-Remnants o Controller and its subclasses: remove all methods that are related to view (mvc view) o Controler: MinActivityLapse class variable seems unused o ControlManager seems removeable o ScrollController seems removeable o MouseMenuController seems removeable * Remove all view classes refs: o SelectorBrowser>>classListIndex: depends on PluggableListView o Browser>>couldBrowseAnyClass depends on PluggableListView * Kernel-ST80 Remnants o ParagraphEditor and TextMorphEditor: kept but have to be cleaned up * ST80-Pluggable Views system category removed * Preferences o removed #mvcProjectsAllowed from flags dictionary * MorphicExtras-ST80 o MorphWorldController class removed * ParagraphEditor o made direct subclass of Controller (get rid of MouseMenuController) o renoved view instance variable o TextMorphEditor is the unique ParagraphEditor subclass, maybe we should only have one class now (ParagraphEditor was more for mvc ?)

1 0

[Pharo-project] Harvesting some fixes
by StÃ©phane Ducasse June 11, 2008

June 11, 2008

Hi all I will harvest the MVC cleaning step one. I harvested some changes. I suggest you to reload 10034. because of some manipulations I did with the scriptloader

2 2

[Pharo-project] self halt in RxParserTest>>testMatchesInwW
by Oscar Nierstrasz June 11, 2008

June 11, 2008

Stef, RxParserTest>>testMatchesInwW has a self halt in it dating from 2006, apparently added by you. I guess you should remove it ... The test runs fine otherwise. - on

3 5

Re: [Pharo-project] Who is working on which tests? Why are there new, broken tests?
by Bill Schwab June 10, 2008

June 10, 2008

Is your concern that the new tests are broken (which seems normal to me that they would start out that way), or are you objecting to their being included in the image? I can see arguments either way on the latter. If code in the image is wrong and a clearly correct but currently failing test highlights it, I'd rather be alerted to it. You are also correct that having them included draws attention that could easily encourage duplication of effort. Comments? 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) 846-1285 FAX: (352) 392-7029 >>> oscar(a)iam.unibe.ch 06/10/08 8:31 AM >>> I looked at the wiki page http://code.google.com/p/pharo/wiki/CurrentlyWorkingOn but it does not say much. I would like to work through some tests, but I cannot find an overview. The tests CharacterTest>>testHex and IntegerTest>>testHex are broken but seem to be very new. Stef, are you working on these? You seem to have written the tests ... - on _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

3 10

June 10, 2008

Hi guys I harvested the MVC step one changes of Alain Plantec Next the on: of damien. PS: I'm checking the hex problem by rebuilding my image from the latest image available. Stef

1 1

Re: [Pharo-project] Who is working on which tests? Why are there new, broken tests?
by Bill Schwab June 10, 2008

June 10, 2008

But if the test is revealing a true flaw in the system, better to keep it than to remove it and pat ourselves on the back about "all tests passing." 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) 846-1285 FAX: (352) 392-7029 >>> oscar(a)iam.unibe.ch 06/10/08 10:00 AM >>> Well, I thought we would start to fix the existing broken tests before adding new broken tests. I would also be very happy to see some comments when someone adds a new broken test, so I know what the plan is. Now I added an implementation of Integer>>hex, for example, but I am not so sure why we need this. Ain't too many senders. An easier fix would have been to remove the broken test. Cheers, - on On Jun 10, 2008, at 15:53, Bill Schwab wrote: > Is your concern that the new tests are broken (which seems normal to > me > that they would start out that way), or are you objecting to their > being > included in the image? > > I can see arguments either way on the latter. If code in the image is > wrong and a clearly correct but currently failing test highlights it, > I'd rather be alerted to it. You are also correct that having them > included draws attention that could easily encourage duplication of > effort. > > Comments? > > 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) 846-1285 > FAX: (352) 392-7029 > >>>> oscar(a)iam.unibe.ch 06/10/08 8:31 AM >>> > > I looked at the wiki page > http://code.google.com/p/pharo/wiki/CurrentlyWorkingOn > but it does not say much. > > I would like to work through some tests, but I cannot find an > overview. > > The tests CharacterTest>>testHex and IntegerTest>>testHex are broken > but seem to be very new. Stef, are you working on these? You seem to > have written the tests ... > > - on > > > _______________________________________________ > 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

2 1