Squeak-dev May 2014

squeak-dev@lists.squeakfoundation.org

43 participants
219 discussions

Re: [squeak-dev] The Trunk: Environments-cwp.43.mcz
by Nicolas Cellier 27 May '14

27 May '14

2014-01-03 21:58 GMT+01:00 <commits(a)source.squeak.org>: > Colin Putney uploaded a new version of Environments to project The Trunk: > http://source.squeak.org/trunk/Environments-cwp.43.mcz > > ==================== Summary ==================== > > Name: Environments-cwp.43 > Author: cwp > Time: 1 January 2014, 1:27:27.436 pm > UUID: 3ccd33e2-0400-405d-b213-fc170cf0a3e6 > Ancestors: Environments-cwp.42 > > Rewrite import/export to be eager, rather than lazy. (step 1 of 3) > > =============== Diff against Environments-cwp.42 =============== > > Item was removed: > - ----- Method: BindingPolicy class>>namespace: (in category 'create') > ----- > - namespace: aDictionary > - ^ self namespace: aDictionary next: nil! > > Item was removed: > - ----- Method: BindingPolicy class>>namespace:next: (in category > 'create') ----- > - namespace: aDictionary next: anImport > - ^ self > - namespace: aDictionary > - policy: AllNamePolicy new > - next: anImport! > > Item was removed: > - ----- Method: BindingPolicy class>>namespace:policy: (in category > 'create') ----- > - namespace: aDictionary policy: aNamePolicy > - ^ self > - namespace: aDictionary > - policy: aNamePolicy > - next: nil! > > Item was removed: > - ----- Method: BindingPolicy class>>namespace:policy:next: (in category > 'create') ----- > - namespace: aDictionary policy: aNamePolicy next: anImport > - ^ self new > - initializeWithNamespace: aDictionary > - policy: aNamePolicy > - next: anImport! > > Item was removed: > - ----- Method: BindingPolicy class>>null (in category 'create') ----- > - null > - ^ self namespace: IdentityDictionary new! > > Item was removed: > - ----- Method: BindingPolicy>>forgetName: (in category 'private') ----- > - forgetName: aSymbol > - self name: aSymbol do: [:foreign | > - namespace removeKey: foreign ifAbsent: [ > - next ifNotNil: [next forgetName: aSymbol]]].! > > Item was removed: > - ----- Method: BindingPolicy>>initializeWithNamespace:policy:next: (in > category 'initialize-release') ----- > - initializeWithNamespace: aDictionary policy: aNamePolicy next: anImport > - namespace := aDictionary. > - policy := aNamePolicy. > - next := anImport! > > Item was changed: > Object subclass: #Environment > + instanceVariableNames: 'info declarations bindings undeclared > policies observers' > - instanceVariableNames: 'info imports exports declarations > references public undeclared bindings policies observers' > classVariableNames: 'Default Instances' > poolDictionaries: '' > category: 'Environments-Core'! > > !Environment commentStamp: 'cmm 12/20/2013 14:10' prior: 0! > I am a context for compiling methods. I maintain the namespace of > classes and global variables that are visible to the methods compiled > within me. > > I have the following instance variables: > > info <EnvironmentInfo> > Metadata about me and the code I contain. > > imports <Import> > Rules for importing globals from other environments. > > exports <Export> > Rules for exposing globals to other environments. > > declarations <IdentityDictionary> > Bindings for globals that have been declared inside me. > > references <IdentityDictionary> > Bindings for globals that are used by methods compiled inside me. > > public <IdentityDictionary> > Bindings for classes that have been declared inside me, and which > satisfy the export rules contain in 'exports'. > > undeclared <Dictionary> > Bindings for globals that are used by methods compiled inside me, but > which aren't present in 'references' and couldn't be found via the rules in > 'imports'.! > > <rant> Hmm, the change of Environment definition would have deserved an update of Environment comment... Some senders of bindingOf: are currently not working (see MCEnvironmentLoadTest), though they once did. This is because bindingOf: uses bindings inst var which is left empty by #testLoadIntoEnvironment. This inst var is not documented and it's not obvious to guess its purpose. It seems that declarations are filled, so I should probably use declarationOf: instead of bindingOf: But that's a blind guess, that means that I change code without understanding the details, and that ain't good. If I want to understand, I will have to open many browsers on senders chains, tests, etc... instead of reading a simple comment. And if I want to review the senders of bindingOf:ifAbsent: / bindingOf: then good luck... Very few will be writing Environment code. Many will read it. My feeling as a reader is that the writers have lack of empathy with the readers. Maybe I'm dumb, but I like to be dumb. Being smart costs too much energy, and in my experience, the smarter the code, the merrier the missinterpretations (that is the bugs). </rant> Item was removed: > - ----- Method: Environment>>migrate (in category 'initialize-release') > ----- > - migrate > - | newDeclarations source dest index policy | > - bindings := IdentityDictionary new. > - newDeclarations := IdentityDictionary new. > - source := Array new: declarations size. > - dest := Array new: declarations size. > - index := 1. > - declarations associationsDo: > - [:ea || binding | > - binding := ea key => ea value. > - source at: index put: ea. > - dest at: index put: binding. > - newDeclarations add: binding. > - bindings add: binding. > - index := index + 1]. > - declarations := newDeclarations. > - source elementsForwardIdentityTo: dest. > - > - policy := BindingPolicy > - environment: self > - policy: AllNamePolicy new > - addSelector: #showBinding: > - removeSelector: #hideBinding:. > - policies := Array with: policy. > - > - observers := IdentitySet new.. > - ! > > Item was removed: > - ----- Method: Environment>>public (in category 'accessing') ----- > - public > - ^ public! > > Item was removed: > - ----- Method: Environment>>publicizeContents (in category 'private') > ----- > - publicizeContents > - declarations associationsDo: [:binding | exports bind: binding]. > - ! > > Item was removed: > - ----- Method: Environment>>rebindUndeclared (in category 'private') ----- > - rebindUndeclared > - undeclared keys do: > - [:name | > - (imports valueOf: name) ifNotNil: > - [:v | > - references declare: name from: undeclared. > - references at: name put: v]]! > > Item was removed: > - BindingPolicy subclass: #Export > - instanceVariableNames: '' > - classVariableNames: '' > - poolDictionaries: '' > - category: 'Environments-Policies'! > > Item was removed: > - ----- Method: Export>>bind: (in category 'binding') ----- > - bind: aBinding > - > - self name: aBinding key do: > - [:foreign | ^ namespace add: (aBinding asBinding: > foreign)]. > - ^ next ifNotNil: [next bind: aBinding]! > > Item was removed: > - ----- Method: Export>>bind:to: (in category 'binding') ----- > - bind: aSymbol to: anObject > - ^ self bind: aSymbol -> anObject! > > Item was removed: > - BindingPolicy subclass: #Import > - instanceVariableNames: '' > - classVariableNames: '' > - poolDictionaries: '' > - category: 'Environments-Policies'! > > Item was removed: > - ----- Method: Import>>bindingOf: (in category 'binding') ----- > - bindingOf: aSymbol > - self name: aSymbol do: > - [:foreign | > - ^ namespace > - associationAt: foreign > - ifAbsent: [next ifNotNil: [next bindingOf: > aSymbol]]]. > - ^ next ifNotNil: [next bindingOf: aSymbol]! > > Item was removed: > - ----- Method: Import>>valueOf: (in category 'binding') ----- > - valueOf: aSymbol > - ^ (self bindingOf: aSymbol) value! > > Item was added: > + (PackageInfo named: 'Environments') postscript: '"Recompile all methods > to fix errant bindings" > + Compiler recompileAll. > + '! > > >

1 2

The Trunk: Tests-nice.298.mcz
by commits＠source.squeak.org 27 May '14

27 May '14

Nicolas Cellier uploaded a new version of Tests to project The Trunk: http://source.squeak.org/trunk/Tests-nice.298.mcz ==================== Summary ==================== Name: Tests-nice.298 Author: nice Time: 27 May 2014, 12:30:53.426 am UUID: 9a8de954-af83-44b6-8b85-25057d5bff13 Ancestors: Tests-nice.297 Fix MCEnvironmentLoadTest 1) its environment failed to import itself, thus #bindingOf: failed to find #MCMockClassA, making testLoadIntoEnvironment fail (fixed in setUp) 2) its environment failed to import binding of #MCMockAPoolDictionary, and this was triggering a UI action - not good for automated tests (fixed in setUp) 3) the imported classes where not deleted, making further MCPackageTest testUnload fail (fixed in tearDown) =============== Diff against Tests-nice.297 =============== Item was changed: + ----- Method: MCEnvironmentLoadTest>>setUp (in category 'running') ----- - ----- Method: MCEnvironmentLoadTest>>setUp (in category 'as yet unclassified') ----- setUp + environment := Environment named: 'MCEnvironmentLoadTest'. + + "This is required in order to not importSelf more than once" + environment stopObserving: environment. + + "This is required to make own declarations visible in environment" + environment importSelf. + + "Import the required PoolDictionary used by the package" + environment bindingOf: #MCMockAPoolDictionary + ifAbsent: [environment from: Smalltalk globals import: #MCMockAPoolDictionary] + + "Curiously, the superclass MCMock will be used, but does not have to be imported, don't ask why..."! - environment := Environment named: 'MCEnvironmentLoadTest'.! Item was added: + ----- Method: MCEnvironmentLoadTest>>tearDown (in category 'running') ----- + tearDown + (environment allClassesAndTraits sorted: [:a :b | + (a allSuperclasses includes: b) or: [a name < b name]]) + do: [:each | each removeFromSystem]! Item was changed: + ----- Method: MCEnvironmentLoadTest>>testLoadIntoEnvironment (in category 'testing') ----- - ----- Method: MCEnvironmentLoadTest>>testLoadIntoEnvironment (in category 'as yet unclassified') ----- testLoadIntoEnvironment [MCPackageLoader installSnapshot: self mockSnapshot] on: CurrentEnvironment do: [:e | e resume: environment]. self assert: (environment bindingOf: #MCMockClassA) notNil description: 'Environment has no binding for MCMockClassA'.!

1 0

#streamContents: on OrderedCollection broken?
by Marcel Taeumel 26 May '14

26 May '14

Should this work? OrderedCollection streamContents: [:result | result nextPutAll: #(a b c) asOrderedCollection]. Well, this works: OrderedCollection streamContents: [:result | result nextPutAll: #(a b c)]. And this works, too: OrderedCollection streamContents: [:result | result nextPut: #foo; nextPutAll: #(a b c) asOrderedCollection]. Best, Marcel -- View this message in context: http://forum.world.st/streamContents-on-OrderedCollection-broken-tp4760369.… Sent from the Squeak - Dev mailing list archive at Nabble.com.

5 7

issue allocating more memory for the image
by Chris Muller 26 May '14

26 May '14

Using latest Cog image 4.0-2776 in Linux, whenever I've had a memory-intensive operation to run, I would simply start up with 2G allocation: squeak -mmap 2000m trunk.image & But, for some reason, when I do this today, I'm getting a message on stdout: sound: /dev/dsp: No such file or directory sound: /dev/dsp: No such file or directory AND, if I simply do it in a workspace self halt and click Proceed in the debugger, I'm getting a pop-up that warns me: 'WARNING: There is not enough space to start the low space watcher. If you proceed, you will not be warned again, and the system may run out of memory and crash. If you do proceed, you can start the low space notifier when more space becomes available simply by opening and then closing a debugger (e.g., by hitting Cmd-period.) Do you want to proceed?' This happens in a standard trunk image. I've never had this issue before, does anyone know why its happening? Thanks.

5 16

Daily Commit Log
by commits＠source.squeak.org 26 May '14

26 May '14

Changes to Trunk (http://source.squeak.org/trunk.html) in the last 24 hours: http://lists.squeakfoundation.org/pipermail/packages/2014-May/007059.html Name: Tests-nice.297 Ancestors: Tests-eem.296 Restore the original MCMockClassA methods #a #b #c to answer 'a' 'b' 'c' as originally intended. Correct the testMcdSerialization test which did not correctly took a snaphsot of modified package (see mockVersionWithAncestor:). Correct MCSerializationTest which omitted to restoreMocks (see tearDown). Add a 30s timeout to MCSerializationTest (tests may timeout in case of huge package cache). Remove MCSnapshotBrowserTest class>>resources which is same as super. Classify some as yet unclassified. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2014-May/007060.html Name: CollectionsTests-nice.218 Ancestors: CollectionsTests-nice.217 Fix a test that did not work anymore after change of Fraction>>hash =============================================

1 0

trivial code changes (was: [squeak-dev] The Trunk: Compression-nice.42.mcz)
by Chris Muller 26 May '14

26 May '14

> On Sat, May 24, 2014 at 2:27 PM, Nicolas Cellier > <nicolas.cellier.aka.nice(a)gmail.com> wrote: >> >> Sorry to push such insignificant change (it's waiting a more serious >> occasion in one of my images for 3 years), >> but I wanted to trigger a build.squeak.org CI build... > > > nothing to apologize for. (allSatisfy:) not. Really?? :-) Nicolas didn't need to apologize, but I also think he expresses totally the right attitude about making trivial improvements to code in _general_. e.g., being conservative about it rather than aggressive about it. (Eliot, I'm not saying you think otherwise -- nothing about you, in fact, nor any comment about (allSatisfy:) not. This is just a general comment). By trying to wait for a more significant change to be needed, Nicolas shows the right level of consideration and respect for "treading on others' code," and for the idea of having a rich change history, full of _meaningful_ changes, instead of a bloated history dilute with unmeaningful ones). Kudos Nicolas. I see a lot of code that could be "improved", like factoring out return carats for example, but I would never replace a "di 1998" edition with "cmm 2014" only for that. I would not like anyone else to do it only for that either. If I feel strongly about a particular ugly code, I'll still look for at least one additional "excuse" before committing a trivial change, like Nicolas did today.

1 0

The Trunk: CollectionsTests-nice.218.mcz
by commits＠source.squeak.org 25 May '14

25 May '14

Nicolas Cellier uploaded a new version of CollectionsTests to project The Trunk: http://source.squeak.org/trunk/CollectionsTests-nice.218.mcz ==================== Summary ==================== Name: CollectionsTests-nice.218 Author: nice Time: 25 May 2014, 11:46:20.057 pm UUID: 09e9444f-318a-4c31-9ba9-139ee94fb3e0 Ancestors: CollectionsTests-nice.217 Fix a test that did not work anymore after change of Fraction>>hash =============== Diff against CollectionsTests-nice.217 =============== Item was changed: ----- Method: WeakSetTest>>testCollisions (in category 'testing') ----- testCollisions | ws o1 o2 o5 on remember forget | "create a weak set" ws := WeakSet new: 15. "select some fractions wanting same place in ws array" + o1 := (2 to: 200) select: [:i | (ws scanFor: 1 / i) = 1]. + o2 := (2 to: 200) select: [:i | (ws scanFor: 1 / i) = 2]. + o5 := (2 to: 200) select: [:i | (ws scanFor: 1 / i) = 5]. + on := (2 to: 200) select: [:i | (ws scanFor: 1 / i) = (ws array size - 1)]. - o1 := (2 to: 100) select: [:i | (ws scanFor: 1 / i) = 1]. - o2 := (2 to: 100) select: [:i | (ws scanFor: 1 / i) = 2]. - o5 := (2 to: 100) select: [:i | (ws scanFor: 1 / i) = 5]. - on := (2 to: 100) select: [:i | (ws scanFor: 1 / i) = (ws array size - 1)]. "Add some fractions to the weak set, and remember a pointer for a few of them" remember := OrderedCollection new. forget := OrderedCollection new. ws add: (remember add: 1 / o1 first). ws add: (forget add: 1 / on second). ws add: (forget add: 1 / o1 second). ws add: (forget add: 1 / o5 second). ws add: (forget add: 1 / o2 second). ws add: (forget add: 1 / o1 third). ws add: (remember add: 1 / o2 first). ws add: (forget add: 1 / o5 third). ws add: (forget add: 1 / on third). ws add: (remember add: 1 / o2 fourth). ws add: (remember add: 1 / on first). ws add: (remember add: 1 / o5 first). "forget and reclaim all entries but those in remember" forget := nil. Smalltalk garbageCollect. remember do: [:m | self assert: (ws includes: m)]. ws add: 1/on second. remember do: [:m | self assert: (ws includes: m)]. ws add: (remember add: 1 / o1 fourth). ws add: (remember add: 1 / on fourth). remember remove: (ws remove: (1 / o5 first)). remember remove: (ws remove: (1 / on first)). remember remove: (ws remove: (1 / o2 first)). remember remove: (ws remove: (1 / o1 first)). remember do: [:m | self assert: (ws includes: m)]. ws add: 1/on second. ws add: 1/o5 second. remember do: [:m | self assert: (ws includes: m)]. !

1 0

Re: [squeak-dev] The Trunk: Files-eem.134.mcz
by Nicolas Cellier 25 May '14

25 May '14

These changes are making some of the BitmapStreamTests fail. Believe it or not, a WordArray is passed to this method... And this seems a normal op, since #primitiveFileWrite is prepared to handle that... 2014-05-03 2:19 GMT+02:00 <commits(a)source.squeak.org>: > Eliot Miranda uploaded a new version of Files to project The Trunk: > http://source.squeak.org/trunk/Files-eem.134.mcz > > ==================== Summary ==================== > > Name: Files-eem.134 > Author: eem > Time: 2 May 2014, 5:19:13.06 pm > UUID: eaf22de5-75b8-4350-b6a5-2fc8ae1077a7 > Ancestors: Files-eem.133 > > Make the Files-eem.133 fix robust when the collection type > doesn't match buffer1's type. > > =============== Diff against Files-eem.133 =============== > > Item was changed: > ----- Method: StandardFileStream>>nextPutAll: (in category 'read, write, > position') ----- > nextPutAll: aString > "Write all the characters of the given string to this file." > | size | > rwmode ifFalse: [^ self error: 'Cannot write a read-only file']. > collection ifNotNil: [ > position < readLimit ifTrue: [ self flushReadBuffer ] ]. > (size := aString size) > 0 ifTrue: > + [buffer1 > + at: 1 > + put: ((buffer1 at: 1) isCharacter > + ifTrue: [(aString at: size) > asCharacter] > + ifFalse: [(aString at: size) > asInteger]). > - [buffer1 at: 1 put: (aString at: size). > self primWrite: fileID from: aString startingAt: 1 count: > aString basicSize]. > ^ aString > ! > > >

1 0

Re: [squeak-dev] The Trunk: Compression-nice.42.mcz
by Nicolas Cellier 25 May '14

25 May '14

Sorry to push such insignificant change (it's waiting a more serious occasion in one of my images for 3 years), but I wanted to trigger a build.squeak.org CI build... 2014-05-24 23:24 GMT+02:00 <commits(a)source.squeak.org>: > Nicolas Cellier uploaded a new version of Compression to project The Trunk: > http://source.squeak.org/trunk/Compression-nice.42.mcz > > ==================== Summary ==================== > > Name: Compression-nice.42 > Author: nice > Time: 24 May 2014, 11:24:05.218 pm > UUID: c9c66aed-c87d-4c7f-8518-fda5842f46c2 > Ancestors: Compression-eem.41 > > Cosmetic refactoring: > (aCollection anySatisfy: aBlock) not -> (aCollection noneSatisfy: aBloc) > > =============== Diff against Compression-eem.41 =============== > > Item was changed: > ----- Method: Archive>>canWriteToFileNamed: (in category 'archive > operations') ----- > canWriteToFileNamed: aFileName > "Catch attempts to overwrite existing zip file" > + ^members noneSatisfy: [ :ea | ea usesFileNamed: aFileName ]. > - ^(members anySatisfy: [ :ea | ea usesFileNamed: aFileName ]) not. > ! > > >

2 1

The Trunk: Tests-nice.297.mcz
by commits＠source.squeak.org 25 May '14

25 May '14

Nicolas Cellier uploaded a new version of Tests to project The Trunk: http://source.squeak.org/trunk/Tests-nice.297.mcz ==================== Summary ==================== Name: Tests-nice.297 Author: nice Time: 25 May 2014, 10:03:43.314 pm UUID: 2d50690f-9f66-4c65-b5a5-59cacb10ac11 Ancestors: Tests-eem.296 Restore the original MCMockClassA methods #a #b #c to answer 'a' 'b' 'c' as originally intended. Correct the testMcdSerialization test which did not correctly took a snaphsot of modified package (see mockVersionWithAncestor:). Correct MCSerializationTest which omitted to restoreMocks (see tearDown). Add a 30s timeout to MCSerializationTest (tests may timeout in case of huge package cache). Remove MCSnapshotBrowserTest class>>resources which is same as super. Classify some as yet unclassified. =============== Diff against Tests-eem.296 =============== Item was changed: ----- Method: MCMockClassA>>a (in category 'numeric') ----- a + ^ 'a'! - ^ 'a2'! Item was changed: ----- Method: MCMockClassA>>b (in category 'numeric') ----- b + ^ 'b'! - ^ 'b1'! Item was changed: ----- Method: MCMockClassA>>c (in category 'numeric') ----- c + ^ 'c'! - ^ 'c1'! Item was added: + ----- Method: MCSerializationTest>>tearDown (in category 'running') ----- + tearDown + self restoreMocks. + super tearDown.! Item was changed: ----- Method: MCSerializationTest>>testMcdSerialization (in category 'testing') ----- testMcdSerialization | stream expected actual | + <timeout: 30> expected := self mockDiffyVersion. stream := RWBinaryOrTextStream on: String new. MCMcdWriter fileOut: expected on: stream. actual := MCMcdReader versionFromStream: stream reset. self assertVersion: actual matches: expected.! Item was changed: ----- Method: MCSerializationTest>>testMczSerialization (in category 'testing') ----- testMczSerialization + <timeout: 30> self assertVersionsMatchWith: MCMczWriter. self assertExtensionProvidedBy: MCMczWriter. self assertVersionInfosMatchWith: MCMczWriter. self assertDependenciesMatchWith: MCMczWriter.! Item was removed: - ----- Method: MCSnapshotBrowserTest class>>resources (in category 'as yet unclassified') ----- - resources - ^ Array with: MCSnapshotResource! Item was changed: + ----- Method: MCTestCase class>>isAbstract (in category 'testing') ----- - ----- Method: MCTestCase class>>isAbstract (in category 'as yet unclassified') ----- isAbstract ^ self = MCTestCase! Item was changed: + ----- Method: MCTestCase class>>resources (in category 'accessing') ----- - ----- Method: MCTestCase class>>resources (in category 'as yet unclassified') ----- resources ^ Array with: MCSnapshotResource! Item was changed: ----- Method: MCTestCase>>mockVersionWithAncestor: (in category 'mocks') ----- mockVersionWithAncestor: aMCVersion ^ MCVersion package: self mockPackage info: (self mockVersionInfoWithAncestor: aMCVersion info) + snapshot: self mockPackage snapshot! - snapshot: self mockSnapshot!

1 0

← Newer
1
2
3
4
5
6
7
8
9
...
22
Older →

Jump to page: