- Pharo-dev - lists.pharo.org

Re: [Pharo-project] [ANN] new update

by Adrian Lienhard

First, I think you have to be logged in. Then, click on an issue and then click the link named "Add a comment andmake changes below" at the left hand side. A form is opened with a text field named Status. Adrian On Jun 26, 2008, at 08:45 , StÃ©phane Ducasse wrote: > how please tell me. > > On Jun 26, 2008, at 12:42 AM, Norbert Hartl wrote: > >> On Wed, 2008-06-25 at 23:50 +0200, Adrian Lienhard wrote: >>> I think closed is better. Everyone knows what it means. >> >> Ok, I changed the issues to "closed" >> >> Norbert >>> Adrian >>> >>> On Jun 25, 2008, at 23:44 , Norbert Hartl wrote: >>> >>>> On Wed, 2008-06-25 at 23:34 +0200, Adrian Lienhard wrote: >>>>> Thanks, Stef! >>>>> Looks good. >>>>> >>>>> In the bug tracker, the status of 3 harvested changes is >>>>> "statusVerified". This does not seem to be a valid state. >>>>> Shouldn't >>>>> it >>>>> be "Closed" instead?! >>>>> >>>> Yes, I created the "Closed" status exactly for this. I don't think >>>> there is much to do after it is harvested therefor closed. Would >>>> it be better to call it "harvested" ? >>>> >>>> Norbert >>>>> Adrian >>>>> >>>>> On Jun 25, 2008, at 22:12 , StÃ©phane Ducasse wrote: >>>>> >>>>>> here is a couple of fixed. >>>>>> BTW yesterday I issued another image. >>>>>> >>>>>> Stef >>>>>> >>>>>> >>>>>> >>>>>> Couple of changes to make more tests work plus fix of >>>>>> finalization >>>>>> logic: >>>>>> - changed WeakRegistry to use WeakIdentityKeyDictionary (http://bugs.squeak.org/view.php?id=6347 >>>>>> ) >>>>>> - FinalizerTest reverted to real test case >>>>>> - added missing class variables to DigitalSigntureAlgorithm >>>>>> - Fix issue #13: TraitsTests pollutes ProtoObject subclasses >>>>>> (Mantis >>>>>> 7090) >>>>>> - make BitmapStreamTest clean up the file >>>>>> bitmapStreamTest.ref.extSeg >>>>>> ------------------------------------------------------ >>>>>> update10043 >>>>>> MVC cleaning step 3 >>>>>> ------------------------------------------------------ >>>>>> update10044 >>>>>> cleaning unimplemented calls >>>>>> ------------------------------------------------------ >>>>>> update10045 >>>>>> Some more cleaning made by Norbert >>>>>> ------------------------------------------------------ >>>>>> update10046 >>>>>> KernelTests-Numbers -- Nan and hex tests broken >>>>>> >>>>>> Collection methods #select: and #collect: have no test >>>>>> >>>>>> MC fixes >>>>>> >>>>>> Remove testAllMethodsWithSourceString >>>>>> >>>>>> >>>>>> >>>>>> _______________________________________________ >>>>>> 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 >>> >>> >>> _______________________________________________ >>> 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

June 26, 2008

Re: [Pharo-project] milestone 1

by StÃ©phane Ducasse

I mean not having squeak in the flap name :) Now this is cool that you come up with this list because I like it a lot since I did not want to push it alone. But now this is different. This is a good goal. It would be good if we can really push that. BTW I will be at RMLL then on "holiday" working on the seaside book. So I may be not reactuve as now (I hope my parent will have a good connection). So next week I will be connected day long, the week after nothing clear, the week after I hope that the box will arrived and that the network will be working. STef On Jun 25, 2008, at 11:59 PM, Adrian Lienhard wrote: > I agree. For Milestone 1 I suggest: > > - carefully check list of 3.10 (and maybe Croquet) fixes and > integrate the ones that make sense (notice, on our wiki there exist > pages listing 3.10.* and Croquet fixes). > > - Check for important fixes posted on Mantis that were not > integrated in 3.10 or were posted more recently. > > - Make all remaining tests pass or mark them as expected failures > (only 2 left!) > > Anything else that would be valuable and does not take too much > effort? > > This version would then be a good candidate for a first release. It > would be good if we have a first release as this allows people to > start moving to Pharo. And its also good that we do not have > fundamental changes that would require code migration. > > Adrian > > PS: Stef, I do not understand what you mean by "putting Pharo in the > flap" > > On Jun 25, 2008, at 09:11 , StÃ©phane Ducasse wrote: > >> for milestone 1 >> - it would be good to have a Pharo in the flap (if any flap) and >> any other better looking lesssqueakish look :L) >> if one of you has time :) > > >> >> >> - when do we want to finish. It would be good to have milestone 1 >> out for esug or before. >> - I would really like to see the items of 3.10 examined. >> >> for milestone 2 >> >> - better building process >> >> - something that I would like us to do is to have a task force >> writing some good class comments. >> If you can think about that it would be good. >> >> Purpose of the class >> Main methods >> Main collaborators >> Example >> Internal representation >> >> 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 >

June 26, 2008

Re: [Pharo-project] [ANN] new update

by StÃ©phane Ducasse

how please tell me. On Jun 26, 2008, at 12:42 AM, Norbert Hartl wrote: > On Wed, 2008-06-25 at 23:50 +0200, Adrian Lienhard wrote: >> I think closed is better. Everyone knows what it means. > > Ok, I changed the issues to "closed" > > Norbert >> Adrian >> >> On Jun 25, 2008, at 23:44 , Norbert Hartl wrote: >> >>> On Wed, 2008-06-25 at 23:34 +0200, Adrian Lienhard wrote: >>>> Thanks, Stef! >>>> Looks good. >>>> >>>> In the bug tracker, the status of 3 harvested changes is >>>> "statusVerified". This does not seem to be a valid state. Shouldn't >>>> it >>>> be "Closed" instead?! >>>> >>> Yes, I created the "Closed" status exactly for this. I don't think >>> there is much to do after it is harvested therefor closed. Would >>> it be better to call it "harvested" ? >>> >>> Norbert >>>> Adrian >>>> >>>> On Jun 25, 2008, at 22:12 , StÃ©phane Ducasse wrote: >>>> >>>>> here is a couple of fixed. >>>>> BTW yesterday I issued another image. >>>>> >>>>> Stef >>>>> >>>>> >>>>> >>>>> Couple of changes to make more tests work plus fix of finalization >>>>> logic: >>>>> - changed WeakRegistry to use WeakIdentityKeyDictionary (http://bugs.squeak.org/view.php?id=6347 >>>>> ) >>>>> - FinalizerTest reverted to real test case >>>>> - added missing class variables to DigitalSigntureAlgorithm >>>>> - Fix issue #13: TraitsTests pollutes ProtoObject subclasses >>>>> (Mantis >>>>> 7090) >>>>> - make BitmapStreamTest clean up the file >>>>> bitmapStreamTest.ref.extSeg >>>>> ------------------------------------------------------ >>>>> update10043 >>>>> MVC cleaning step 3 >>>>> ------------------------------------------------------ >>>>> update10044 >>>>> cleaning unimplemented calls >>>>> ------------------------------------------------------ >>>>> update10045 >>>>> Some more cleaning made by Norbert >>>>> ------------------------------------------------------ >>>>> update10046 >>>>> KernelTests-Numbers -- Nan and hex tests broken >>>>> >>>>> Collection methods #select: and #collect: have no test >>>>> >>>>> MC fixes >>>>> >>>>> Remove testAllMethodsWithSourceString >>>>> >>>>> >>>>> >>>>> _______________________________________________ >>>>> 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 >> >> >> _______________________________________________ >> 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 >

June 26, 2008

Re: [Pharo-project] [ANN] new update

by StÃ©phane Ducasse

Yes but I have no idea no idea how to interact and change these status. Again a geek interface. Stef On Jun 25, 2008, at 11:34 PM, Adrian Lienhard wrote: > Thanks, Stef! > Looks good. > > In the bug tracker, the status of 3 harvested changes is > "statusVerified". This does not seem to be a valid state. Shouldn't > it be "Closed" instead?! > > Adrian > > On Jun 25, 2008, at 22:12 , StÃ©phane Ducasse wrote: > >> here is a couple of fixed. >> BTW yesterday I issued another image. >> >> Stef >> >> >> >> Couple of changes to make more tests work plus fix of finalization >> logic: >> - changed WeakRegistry to use WeakIdentityKeyDictionary (http://bugs.squeak.org/view.php?id=6347 >> ) >> - FinalizerTest reverted to real test case >> - added missing class variables to DigitalSigntureAlgorithm >> - Fix issue #13: TraitsTests pollutes ProtoObject subclasses >> (Mantis 7090) >> - make BitmapStreamTest clean up the file bitmapStreamTest.ref.extSeg >> ------------------------------------------------------ >> update10043 >> MVC cleaning step 3 >> ------------------------------------------------------ >> update10044 >> cleaning unimplemented calls >> ------------------------------------------------------ >> update10045 >> Some more cleaning made by Norbert >> ------------------------------------------------------ >> update10046 >> KernelTests-Numbers -- Nan and hex tests broken >> >> Collection methods #select: and #collect: have no test >> >> MC fixes >> >> Remove testAllMethodsWithSourceString >> >> >> >> _______________________________________________ >> 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 >

June 26, 2008

Re: [Pharo-project] ReleaseTest>>testUnimplementedCalls

by StÃ©phane Ducasse

I would keep it Now may be we should improve SUnit to add the possibility to have long tests.... I think that this is really important that we do not get stuck with what we have. Stef On Jun 26, 2008, at 12:38 AM, Norbert Hartl wrote: > I thought once more about this test. I'm reluctant to have > this test removed. On the other hand is Adrien right. He > states that a lot of test classes are doing calls to methods > they created while the test runs. > > One solution would be to change the test to use perform: to > invoke those methods. But that would making tests a bit > tedious. So this is not my preferred solution. > > I tweaked SystemNavigation>>allUnimplementedNonPrimitiveCalls > to > > allUnimplementedNonPrimitiveCallsX > "Answer an Array of each message that is sent by an expression in a > method but is not implemented by any object in the system." > | all meth dict | > dict := Dictionary new. > all := self systemNavigation allImplementedMessages. > self systemNavigation allBehaviorsDo: [:cl | > cl selectorsDo: [:sel | > meth := cl compiledMethodAt: sel. > meth primitive = 0 ifTrue: [ > meth messages do: [:m | > (all includes: m) ifFalse: [ > ((dict at: cl ifAbsentPut: [ Dictionary new ]) > at: sel ifAbsentPut: [ OrderedCollection new]) > add: m > ] > ] > ] > ] > ]. > ^ dict > > So the test would be like > > (SystemNavigation default allUnimplementedNonPrimitiveCallsX > associationsSelect: [:a| > (a key allSuperclasses includes: TestCase) not > ]) isEmpty > > Would this or something similar be useful or is this just a lot of > noise without much of a benefit? > > Norbert > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >

June 26, 2008

Re: [Pharo-project] milestone 1

by Serge Stinckwich

2008/6/25 Adrian Lienhard <adi(a)netstyle.ch>: > I agree. For Milestone 1 I suggest: > > - carefully check list of 3.10 (and maybe Croquet) fixes and integrate the > ones that make sense (notice, on our wiki there exist pages listing 3.10.* > and Croquet fixes). > > - Check for important fixes posted on Mantis that were not integrated in > 3.10 or were posted more recently. > > - Make all remaining tests pass or mark them as expected failures (only 2 > left!) > > Anything else that would be valuable and does not take too much effort? > > This version would then be a good candidate for a first release. It would be > good if we have a first release as this allows people to start moving to > Pharo. And its also good that we do not have fundamental changes that would > require code migration. > > Adrian > > PS: Stef, I do not understand what you mean by "putting Pharo in the flap" I guess he would like to have the Pharo Logo into a flap. -- Serge Stinckwich Smalltalkers do: [:it | All with: Class, (And love: it)] http://blog.doesnotunderstand.org/

June 26, 2008

Re: [Pharo-project] [ANN] new update

by Norbert Hartl

On Wed, 2008-06-25 at 23:50 +0200, Adrian Lienhard wrote: > I think closed is better. Everyone knows what it means. Ok, I changed the issues to "closed" Norbert > Adrian > > On Jun 25, 2008, at 23:44 , Norbert Hartl wrote: > > > On Wed, 2008-06-25 at 23:34 +0200, Adrian Lienhard wrote: > >> Thanks, Stef! > >> Looks good. > >> > >> In the bug tracker, the status of 3 harvested changes is > >> "statusVerified". This does not seem to be a valid state. Shouldn't > >> it > >> be "Closed" instead?! > >> > > Yes, I created the "Closed" status exactly for this. I don't think > > there is much to do after it is harvested therefor closed. Would > > it be better to call it "harvested" ? > > > > Norbert > >> Adrian > >> > >> On Jun 25, 2008, at 22:12 , StÃ©phane Ducasse wrote: > >> > >>> here is a couple of fixed. > >>> BTW yesterday I issued another image. > >>> > >>> Stef > >>> > >>> > >>> > >>> Couple of changes to make more tests work plus fix of finalization > >>> logic: > >>> - changed WeakRegistry to use WeakIdentityKeyDictionary (http://bugs.squeak.org/view.php?id=6347 > >>> ) > >>> - FinalizerTest reverted to real test case > >>> - added missing class variables to DigitalSigntureAlgorithm > >>> - Fix issue #13: TraitsTests pollutes ProtoObject subclasses (Mantis > >>> 7090) > >>> - make BitmapStreamTest clean up the file > >>> bitmapStreamTest.ref.extSeg > >>> ------------------------------------------------------ > >>> update10043 > >>> MVC cleaning step 3 > >>> ------------------------------------------------------ > >>> update10044 > >>> cleaning unimplemented calls > >>> ------------------------------------------------------ > >>> update10045 > >>> Some more cleaning made by Norbert > >>> ------------------------------------------------------ > >>> update10046 > >>> KernelTests-Numbers -- Nan and hex tests broken > >>> > >>> Collection methods #select: and #collect: have no test > >>> > >>> MC fixes > >>> > >>> Remove testAllMethodsWithSourceString > >>> > >>> > >>> > >>> _______________________________________________ > >>> 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 > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

June 25, 2008

[Pharo-project] ReleaseTest>>testUnimplementedCalls

by Norbert Hartl

I thought once more about this test. I'm reluctant to have this test removed. On the other hand is Adrien right. He states that a lot of test classes are doing calls to methods they created while the test runs. One solution would be to change the test to use perform: to invoke those methods. But that would making tests a bit tedious. So this is not my preferred solution. I tweaked SystemNavigation>>allUnimplementedNonPrimitiveCalls to allUnimplementedNonPrimitiveCallsX "Answer an Array of each message that is sent by an expression in a method but is not implemented by any object in the system." | all meth dict | dict := Dictionary new. all := self systemNavigation allImplementedMessages. self systemNavigation allBehaviorsDo: [:cl | cl selectorsDo: [:sel | meth := cl compiledMethodAt: sel. meth primitive = 0 ifTrue: [ meth messages do: [:m | (all includes: m) ifFalse: [ ((dict at: cl ifAbsentPut: [ Dictionary new ]) at: sel ifAbsentPut: [ OrderedCollection new]) add: m ] ] ] ] ]. ^ dict So the test would be like (SystemNavigation default allUnimplementedNonPrimitiveCallsX associationsSelect: [:a| (a key allSuperclasses includes: TestCase) not ]) isEmpty Would this or something similar be useful or is this just a lot of noise without much of a benefit? Norbert

June 25, 2008

Re: [Pharo-project] milestone 1

by Adrian Lienhard

I agree. For Milestone 1 I suggest: - carefully check list of 3.10 (and maybe Croquet) fixes and integrate the ones that make sense (notice, on our wiki there exist pages listing 3.10.* and Croquet fixes). - Check for important fixes posted on Mantis that were not integrated in 3.10 or were posted more recently. - Make all remaining tests pass or mark them as expected failures (only 2 left!) Anything else that would be valuable and does not take too much effort? This version would then be a good candidate for a first release. It would be good if we have a first release as this allows people to start moving to Pharo. And its also good that we do not have fundamental changes that would require code migration. Adrian PS: Stef, I do not understand what you mean by "putting Pharo in the flap" On Jun 25, 2008, at 09:11 , StÃ©phane Ducasse wrote: > for milestone 1 > - it would be good to have a Pharo in the flap (if any flap) and any > other better looking lesssqueakish look :L) > if one of you has time :) > > > - when do we want to finish. It would be good to have milestone 1 > out for esug or before. > - I would really like to see the items of 3.10 examined. > > for milestone 2 > > - better building process > > - something that I would like us to do is to have a task force > writing some good class comments. > If you can think about that it would be good. > > Purpose of the class > Main methods > Main collaborators > Example > Internal representation > > Stef > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

June 25, 2008

Re: [Pharo-project] [ANN] new update

by Adrian Lienhard

I think closed is better. Everyone knows what it means. Adrian On Jun 25, 2008, at 23:44 , Norbert Hartl wrote: > On Wed, 2008-06-25 at 23:34 +0200, Adrian Lienhard wrote: >> Thanks, Stef! >> Looks good. >> >> In the bug tracker, the status of 3 harvested changes is >> "statusVerified". This does not seem to be a valid state. Shouldn't >> it >> be "Closed" instead?! >> > Yes, I created the "Closed" status exactly for this. I don't think > there is much to do after it is harvested therefor closed. Would > it be better to call it "harvested" ? > > Norbert >> Adrian >> >> On Jun 25, 2008, at 22:12 , StÃ©phane Ducasse wrote: >> >>> here is a couple of fixed. >>> BTW yesterday I issued another image. >>> >>> Stef >>> >>> >>> >>> Couple of changes to make more tests work plus fix of finalization >>> logic: >>> - changed WeakRegistry to use WeakIdentityKeyDictionary (http://bugs.squeak.org/view.php?id=6347 >>> ) >>> - FinalizerTest reverted to real test case >>> - added missing class variables to DigitalSigntureAlgorithm >>> - Fix issue #13: TraitsTests pollutes ProtoObject subclasses (Mantis >>> 7090) >>> - make BitmapStreamTest clean up the file >>> bitmapStreamTest.ref.extSeg >>> ------------------------------------------------------ >>> update10043 >>> MVC cleaning step 3 >>> ------------------------------------------------------ >>> update10044 >>> cleaning unimplemented calls >>> ------------------------------------------------------ >>> update10045 >>> Some more cleaning made by Norbert >>> ------------------------------------------------------ >>> update10046 >>> KernelTests-Numbers -- Nan and hex tests broken >>> >>> Collection methods #select: and #collect: have no test >>> >>> MC fixes >>> >>> Remove testAllMethodsWithSourceString >>> >>> >>> >>> _______________________________________________ >>> 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

June 25, 2008