Squeak-dev May 2011

squeak-dev@lists.squeakfoundation.org

84 participants
303 discussions

Re: [squeak-dev] The Trunk: Tools-cmm.348.mcz
by Frank Shearar 03 May '11

03 May '11

On 2011/05/03 01:24, commits(a)source.squeak.org wrote: > Chris Muller uploaded a new version of Tools to project The Trunk: > http://source.squeak.org/trunk/Tools-cmm.348.mcz > > ==================== Summary ==================== > > Name: Tools-cmm.348 > Author: cmm > Time: 2 May 2011, 8:24:10.46 pm > UUID: b02d968c-e94f-452b-aef7-d78fa343ff09 > Ancestors: Tools-cmm.347 > > Fixed performance bug when opening hierarchy browser on Object. > > =============== Diff against Tools-cmm.347 =============== > > Item was changed: > ----- Method: HierarchyBrowser>>initHierarchyForClass: (in category 'initialization') ----- > + initHierarchyForClass: aClassOrMetaClass > + | nonMetaClass superclasses | > - initHierarchyForClass: aClassOrMetaClass > - | nonMetaClass | > centralClass := aClassOrMetaClass. > nonMetaClass := aClassOrMetaClass theNonMetaClass. > self systemOrganizer: SystemOrganization. > metaClassIndicated := aClassOrMetaClass isMeta. > classDisplayList := OrderedCollection new. > + (superclasses := nonMetaClass allSuperclasses reversed) withIndexDo: > + [ : each : indent | classDisplayList add: > + (String streamContents: > + [ : stream | indent - 1 timesRepeat: [ stream nextPutAll: ' ' ]. > + stream nextPutAll: each name ]) ]. > + nonMetaClass > + allSubclassesWithLevelDo: > + [ : eachClass : lvl | classDisplayList add: > + (String streamContents: > + [ : stream | lvl+1 timesRepeat: [ stream nextPutAll: ' ' ]. > + stream nextPutAll: eachClass name ]) ] > + startingLevel: superclasses size. > + self selectClass: nonMetaClass! > - self > - flattenHierarchyTree: (self createHierarchyTreeOf: (nonMetaClass allSuperclasses, { nonMetaClass }, nonMetaClass allSubclasses)) > - on: classDisplayList > - indent: '' > - by: ' '. > - > - self selectClass: nonMetaClass.! That also looks rather more understandable that trying to wrap one's head around how createHierarchyTreeOf: works. And this avoids the ObjectTracer hierarchy issue too, doesn't it? frank

2 1

Fixing support for selectors with underscores
by Levente Uzonyi 03 May '11

03 May '11

Hi, since Squeak 4.1 it's possible [1] to use underscores in class, method and variable names. System wide support for them is still incomplete though. There are two reasons for this: - our developement process doesn't need this feature - noone wanted to fix it The per-class scoping causes so much trouble, that I'd like to remove it completely, leaving only global preferences. Why? - It doesn't work for - class names, - class variable names, - class side methods. - It doesn't work with code loading tools (e.g. Monticello), because there's no guarantee, that #allowUnderscoreSelectors will be loaded before the methods using underscores that way. - It doesn't work for tools viewing external code. Any objections? Cheers, Levente [1] http://forum.world.st/Selectors-with-underscores-Have-your-cake-and-eat-it-… [2] http://bugs.squeak.org/view.php?id=7547

7 15

Re: [squeak-dev] The Trunk: Kernel-cmm.578.mcz
by Enrico Spinielli 03 May '11

03 May '11

Hi, see below On Tue, May 3, 2011 at 01:46, <commits(a)source.squeak.org> wrote: > Chris Muller uploaded a new version of Kernel to project The Trunk: > http://source.squeak.org/trunk/Kernel-cmm.578.mcz > > ==================== Summary ==================== > > Name: Kernel-cmm.578 > Author: cmm > Time: 2 May 2011, 6:46:17.696 pm > UUID: 913d762f-e361-4f5e-814f-0b0692ffa4d9 > Ancestors: Kernel-ul.577 > > - Support for streamlined variable reference / assignment browsing. > - Fix for Month class>>#readFrom:. > > =============== Diff against Kernel-ul.577 =============== > [...] > Item was changed: > ----- Method: Month class>>readFrom: (in category 'squeak protocol') ----- > readFrom: aStream > > | m y c | > m := (ReadWriteStream with: '') reset. > [(c := aStream next) isSeparator] whileFalse: [m nextPut: c]. > [(c := aStream next) isSeparator] whileTrue. >* y := (ReadWriteStream with: '') reset. >* y nextPut: c. >* [aStream atEnd] whileFalse: [y nextPut: aStream next]. > > ^ self > month: m contents > + year: y contents asInteger > - year: y contents > > "Month readFrom: 'July 1998' readStream"! Last part (lines marked with '>*') can probably be simplified (as in Date>>readFrom:) as follows: y := Integer readFrom: aStream. Hope it helps Bye Enrico -- Enrico Spinielli "Do Androids dream of electric sheep?"— Philip K. Dick "Hear and forget; see and remember;do and understand."—Mitchel Resnick

1 0

Re: Hello list
by Rodney Polkinghorne 03 May '11

03 May '11

Nicolas Cellier, replying to me on Wed, 20 Apr: > One question that soon arise is complex extension of some intrinsic > functions, say arcSin for example. > we extend the Number>>#arcSin with a complex extension Scheme has done that for 30 years without drama. > Drawback: we loose the ability to raise a DomainError I think this drawback is minor. Spurious errors shouldn't be raised. Actual errors will be apparent at a higher level, and should be handled there. Usually, the mistake was using real numbers at all. > Anyway I'm interested in cleaning and extending trunk if necessary. Could I help? Rodney

1 0

Re: Hello list
by Rodney Polkinghorne 03 May '11

03 May '11

Bert Freudenberg, replying to me on Wed, 20 Apr: >> Someone must have been there before; Number currently >> accepts an #argument message, and answers 0 or pi according to sign. >> I think everything that knows its modulus and argument should know its >> real and imaginary parts. > That would be trivial to add to Number. See changeset below. > What else do you think is missing? That's all I've found so far. In my image, I'll add anything whose absence requires me to write "+0i". >> I found that, tried it, and got sick of saying "0 asComplex" in about >> two minutes. > So why don't you write 0i instead of "0 asComplex"? It was actually "z asComplex" in the method "foo:", so that "foo: 0" would work. I think "foo: 0i" looks tricky. Rodney 'From Squeak4.2 of 4 February 2011 [latest update: #10966] on 3 May 2011 at 1:03:49 pm'! !Number methodsFor: 'arithmetic' stamp: 'RESP 4/19/2011 11:23'! imaginary "Answer the imaginary part of the receiver (see Complex | imaginary)." ^ self * 0! ! !Number methodsFor: 'arithmetic' stamp: 'RESP 4/19/2011 11:22'! real "Answer the real part of the receiver (see Complex | real)." ^ self! ! !TestCase methodsFor: 'accessing' stamp: 'RESP 4/19/2011 11:49'! assert: aValue fixedPointOf: aMethod aValue = (aValue perform: aMethod) ifFalse: [self signalFailure: 'Assertion failed']. ! ! !ComplexTest methodsFor: 'tests' stamp: 'RESP 5/3/2011 13:02'! testCreation "self run: #testCreation" | c | c := 5 i. self assert: (c real = 0). self assert: (c imaginary = 5). c := 6 + 7 i. self assert: (c real = 6). self assert: ( c imaginary = 7). c := 5.6 - 8 i. self assert: (c real = 5.6). self assert: (c imaginary = -8). c := Complex real: 10 imaginary: 5. self assert: (c real = 10). self assert: (c imaginary = 5). c := Complex abs: 5 arg: (Float pi/2). self assert: (c real rounded = 0). self assert: (c imaginary = 5). "The real and imaginary methods work, for negative, zero, and positive Numbers of all kinds." "Float" self assert: 0.0 fixedPointOf: #real. self assert: 0.0 imaginary = 0. self assert: Float pi fixedPointOf: #real. self assert: Float pi imaginary = 0. self assert: Float pi negated fixedPointOf: #real. self assert: Float pi negated imaginary = 0. "Fraction" self assert: (2/5) fixedPointOf: #real. self assert: (2/5) imaginary = 0. self assert: (-2/5) fixedPointOf: #real. self assert: (-2/5) imaginary = 0. "SmallInteger" self assert: 0 fixedPointOf: #real. self assert: 0 imaginary = 0. self assert: 42 fixedPointOf: #real. self assert: 42 imaginary = 0. self assert: -42 fixedPointOf: #real. self assert: -42 imaginary = 0. "LargePositiveInteger" self assert: ((2 raisedTo: 100) isMemberOf: LargePositiveInteger). self assert: (2 raisedTo: 100) fixedPointOf: #real. self assert: (2 raisedTo: 100) imaginary = 0. "LargeNegativeInteger" self assert: ((2 raisedTo: 100) negated isMemberOf: LargeNegativeInteger). self assert: (2 raisedTo: 100) negated fixedPointOf: #real. self assert: (2 raisedTo: 100) negated imaginary = 0. ! !

1 0

Re: [squeak-dev] The Trunk: Protocols-cmm.37.mcz
by Levente Uzonyi 03 May '11

03 May '11

On Tue, 3 May 2011, commits(a)source.squeak.org wrote: > Chris Muller uploaded a new version of Protocols to project The Trunk: > http://source.squeak.org/trunk/Protocols-cmm.37.mcz > > ==================== Summary ==================== > > Name: Protocols-cmm.37 > Author: cmm > Time: 2 May 2011, 7:55:08.248 pm > UUID: d96b0e7e-5131-4408-a928-c4d68bb8aac9 > Ancestors: Protocols-cmm.36 > > - Reverted half-renamed sender. > > =============== Diff against Protocols-cmm.36 =============== > > Item was changed: > ----- Method: Lexicon>>selectorsDefiningInstVar (in category 'within-tool queries') ----- > selectorsDefiningInstVar > "Return a list of methods that define a given inst var that are in the protocol of this object" > > | aList | > aList := OrderedCollection new. > targetClass withAllSuperclassesDo: > [:aClass | > + (aClass whichMethodsStoreInto: currentQueryParameter asString) do: > - (aClass whichMethodsAssign: currentQueryParameter asString) do: > [ : cm | cm selector isDoIt ifFalse: [aList add: cm selector] AFAIK DoIts are not stored in MethodDictionaries anymore, so that check is superfluous. Levente > ] > ]. > ^ aList! > > >

2 1

The Trunk: Tools-cmm.348.mcz
by commits＠source.squeak.org 03 May '11

03 May '11

Chris Muller uploaded a new version of Tools to project The Trunk: http://source.squeak.org/trunk/Tools-cmm.348.mcz ==================== Summary ==================== Name: Tools-cmm.348 Author: cmm Time: 2 May 2011, 8:24:10.46 pm UUID: b02d968c-e94f-452b-aef7-d78fa343ff09 Ancestors: Tools-cmm.347 Fixed performance bug when opening hierarchy browser on Object. =============== Diff against Tools-cmm.347 =============== Item was changed: ----- Method: HierarchyBrowser>>initHierarchyForClass: (in category 'initialization') ----- + initHierarchyForClass: aClassOrMetaClass + | nonMetaClass superclasses | - initHierarchyForClass: aClassOrMetaClass - | nonMetaClass | centralClass := aClassOrMetaClass. nonMetaClass := aClassOrMetaClass theNonMetaClass. self systemOrganizer: SystemOrganization. metaClassIndicated := aClassOrMetaClass isMeta. classDisplayList := OrderedCollection new. + (superclasses := nonMetaClass allSuperclasses reversed) withIndexDo: + [ : each : indent | classDisplayList add: + (String streamContents: + [ : stream | indent - 1 timesRepeat: [ stream nextPutAll: ' ' ]. + stream nextPutAll: each name ]) ]. + nonMetaClass + allSubclassesWithLevelDo: + [ : eachClass : lvl | classDisplayList add: + (String streamContents: + [ : stream | lvl+1 timesRepeat: [ stream nextPutAll: ' ' ]. + stream nextPutAll: eachClass name ]) ] + startingLevel: superclasses size. + self selectClass: nonMetaClass! - self - flattenHierarchyTree: (self createHierarchyTreeOf: (nonMetaClass allSuperclasses, { nonMetaClass }, nonMetaClass allSubclasses)) - on: classDisplayList - indent: '' - by: ' '. - - self selectClass: nonMetaClass.!

1 0

The Trunk: Protocols-cmm.37.mcz
by commits＠source.squeak.org 03 May '11

03 May '11

Chris Muller uploaded a new version of Protocols to project The Trunk: http://source.squeak.org/trunk/Protocols-cmm.37.mcz ==================== Summary ==================== Name: Protocols-cmm.37 Author: cmm Time: 2 May 2011, 7:55:08.248 pm UUID: d96b0e7e-5131-4408-a928-c4d68bb8aac9 Ancestors: Protocols-cmm.36 - Reverted half-renamed sender. =============== Diff against Protocols-cmm.36 =============== Item was changed: ----- Method: Lexicon>>selectorsDefiningInstVar (in category 'within-tool queries') ----- selectorsDefiningInstVar "Return a list of methods that define a given inst var that are in the protocol of this object" | aList | aList := OrderedCollection new. targetClass withAllSuperclassesDo: [:aClass | + (aClass whichMethodsStoreInto: currentQueryParameter asString) do: - (aClass whichMethodsAssign: currentQueryParameter asString) do: [ : cm | cm selector isDoIt ifFalse: [aList add: cm selector] ] ]. ^ aList!

1 0

The Trunk: Kernel-cmm.579.mcz
by commits＠source.squeak.org 03 May '11

03 May '11

Chris Muller uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-cmm.579.mcz ==================== Summary ==================== Name: Kernel-cmm.579 Author: cmm Time: 2 May 2011, 7:53:17.364 pm UUID: bc0c0f5a-e690-458d-a2c1-5ca67389166c Ancestors: Kernel-cmm.578 - Removed unused methods. - Fixed sender. =============== Diff against Kernel-cmm.578 =============== Item was removed: - ----- Method: Behavior>>whichMethodsStoreInto: (in category 'testing method dictionary') ----- - whichMethodsStoreInto: instVarName - "Answer a collection of selectors whose methods access the argument, - instVarName, as a named instance variable." - "Point whichMethodsStoreInto: 'x'." - | instVarIndex | - instVarIndex := self allInstVarNames - indexOf: instVarName - ifAbsent: [ ^ IdentitySet new ]. - ^ Array streamContents: - [ : stream | self methodDict keysAndValuesDo: - [ : eachSelector : eachMethod | (eachMethod writesField: instVarIndex) ifTrue: [ stream nextPut: eachMethod ] ] ]! Item was removed: - ----- Method: Behavior>>whichSelectorsStoreInto: (in category 'testing method dictionary') ----- - whichSelectorsStoreInto: instVarName - "Answer a collection of selectors whose methods access the argument, - instVarName, as a named instance variable." - | instVarIndex | - instVarIndex := self allInstVarNames indexOf: instVarName ifAbsent: [^IdentitySet new]. - ^ self methodDict keys select: - [:sel | (self methodDict at: sel) writesField: instVarIndex] - - "Point whichSelectorsStoreInto: 'x'."! Item was changed: ----- Method: MethodFinder>>initialize (in category 'initialize') ----- (excessive size, no diff calculated)

1 0

Daily Commit Log
by commits＠source.squeak.org 03 May '11

03 May '11

Changes to Trunk (http://source.squeak.org/trunk.html) in the last 24 hours: http://lists.squeakfoundation.org/pipermail/packages/2011-May/004641.html Name: ToolsTests-fbs.41 Ancestors: ToolsTests-fbs.40 Maintain the currently selected message category and method when resorting the message categories (http://bugs.squeak.org/view.php?id=7626). ============================================= http://lists.squeakfoundation.org/pipermail/packages/2011-May/004642.html Name: Tools-fbs.345 Ancestors: Tools-fbs.344 Levente Uzonyi found that deleting the last message in a MessageSet raised an error. This fixes the error by checking the list of messages: if there are no messages in the Browser (or subclass) instance, then explicitly select "no method selected". ============================================= http://lists.squeakfoundation.org/pipermail/packages/2011-May/004643.html Name: Tools-fbs.346 Ancestors: Tools-cmm.345 Maintain the currently selected message category and method when resorting the message categories (http://bugs.squeak.org/view.php?id=7626). ============================================= http://lists.squeakfoundation.org/pipermail/packages/2011-May/004644.html Name: Tools-cmm.347 Ancestors: Tools-fbs.346, Tools-fbs.345 - Update menu "inst var defs" to "assignments..." and "inst var refs" to "references...". Each option now includes a selection of the available class-variables, enabling "class var refs" to be removed. Each option also has an appropriate hot-key assignment. - Removed "class vars" because one could simply explore the class' classPool. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2011-May/004645.html Name: System-cmm.433 Ancestors: System-mtf.432 - Updates to SystemNavigation to support streamlined variable-reference / assignment browsing. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2011-May/004646.html Name: Tests-cmm.126 Ancestors: Tests-ul.125 Updated test to use new MCDirectoryRepository constructor. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2011-May/004647.html Name: Protocols-cmm.36 Ancestors: Protocols-nice.35 - Use new streamlined variable-reference / assignment browsing. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2011-May/004648.html Name: KernelTests-cmm.188 Ancestors: KernelTests-nice.187 Demonstrate bug with Month class>>readFrom: when used with February. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2011-May/004649.html Name: Kernel-cmm.578 Ancestors: Kernel-ul.577 - Support for streamlined variable reference / assignment browsing. - Fix for Month class>>#readFrom:. =============================================

1 0

Jump to page: