Squeak-dev March 2018

squeak-dev@lists.squeakfoundation.org

53 participants
146 discussions

The Trunk: System-cmm.1009.mcz
by commits＠source.squeak.org 03 Apr '18

03 Apr '18

Chris Muller uploaded a new version of System to project The Trunk: http://source.squeak.org/trunk/System-cmm.1009.mcz ==================== Summary ==================== Name: System-cmm.1009 Author: cmm Time: 30 March 2018, 5:47:41.207216 pm UUID: fce32aff-3701-47be-9d68-125ede7bed60 Ancestors: System-mt.1008 Modernize the fonts of my Community Dark theme to better accomodate larger displays. =============== Diff against System-mt.1008 =============== Item was changed: ----- Method: CommunityTheme class>>addDarkFonts: (in category 'instance creation') ----- addDarkFonts: aUserInterfaceTheme "Set-up fonts." aUserInterfaceTheme + set: #balloonHelpFont to: (StrikeFont familyName: 'Darkmap DejaVu Sans' pointSize: 9 emphasized: TextEmphasis italic emphasisCode); + set: #standardButtonFont to: (StrikeFont familyName: 'Darkmap DejaVu Sans' pointSize: 9); + set: #standardCodeFont to: (StrikeFont familyName: 'Bitmap DejaVu Sans' pointSize: 12); + set: #standardFlapFont to: (StrikeFont familyName: 'Bitmap DejaVu Sans' pointSize: 9 emphasized: TextEmphasis bold emphasisCode); - set: #balloonHelpFont to: (StrikeFont familyName: 'Darkmap DejaVu Sans' pointSize: 7 emphasized: TextEmphasis italic emphasisCode); - set: #standardButtonFont to: (StrikeFont familyName: 'Darkmap DejaVu Sans' pointSize: 7); - set: #standardCodeFont to: (StrikeFont familyName: 'Bitmap DejaVu Sans' pointSize: 9); - set: #standardFlapFont to: (StrikeFont familyName: 'Bitmap DejaVu Sans' pointSize: 7 emphasized: TextEmphasis bold emphasisCode); set: #haloLabelFont to: (StrikeFont familyName: 'Darkmap DejaVu Sans' pointSize: 9); + set: #standardListFont to: (StrikeFont familyName: 'Bitmap DejaVu Sans' pointSize: 12); + set: #standardMenuFont to: (StrikeFont familyName: 'Bitmap DejaVu Sans' pointSize: 12); + set: #standardSystemFont to: (StrikeFont familyName: 'Bitmap DejaVu Sans' pointSize: 12); + set: #windowTitleFont to: (StrikeFont familyName: 'Bitmap DejaVu Sans' pointSize: 12 emphasized: TextEmphasis bold emphasisCode)! - set: #standardListFont to: (StrikeFont familyName: 'Bitmap DejaVu Sans' pointSize: 9); - set: #standardMenuFont to: (StrikeFont familyName: 'Bitmap DejaVu Sans' pointSize: 9); - set: #standardSystemFont to: (StrikeFont familyName: 'Bitmap DejaVu Sans' pointSize: 9); - set: #windowTitleFont to: (StrikeFont familyName: 'Bitmap DejaVu Sans' pointSize: 9 emphasized: TextEmphasis bold emphasisCode)!

4 3

Re: [squeak-dev] [Newbies] AMB
by David T. Lewis 01 Apr '18

01 Apr '18

I am CCing squeak-dev because Stéphane Rollandin is active on that list. Stef did the Squeak implementation and can probably answer your questions. Dave On Sat, Mar 31, 2018 at 03:51:22PM -0700, Sean P. DeNigris wrote: > Is anyone familiar with the AMB project, installable from SqMap catalog? > > I got excited by the description's assertion that "it can do incredible > things" (it even provided an example. After reading the description, tests, > code, and references, I have an inkling what it does, but don't quite > understand what is "incredible" about it and don't want to miss the magic!! > > The catalog description is: > > This is the "ambigous", "nondeterminist", "angelic" operator aka amb > > ... more seriously described as a declarative control flow statement able to > enumerate over finite domains all values that make a program terminate. It > uses continuations to walk the tree of possible tuples, backtracking when > the program fails. > > The amazing thing is the simplicity of the code: four short or very short > methods on the instance side, the class side methods only providing > syntactic sugar. > > And it can do incredible things such as: > > AMB assert: [:string :n :char | (string occurrencesOf: char) = n] > over: #('maman' 'barbapapa' 'guiliguili' 'arthur' 'chtulu') > and: #(2 4 1) > and: #($m $b $a $i $u) > > See AMBTest for more magic. > > I would have been unable to implement this by myself ! This is a port of the > code by Dorai Sitaram which is the reference documentation for amb: > > http://www.ccs.neu.edu/home/dorai/t-y-scheme/t-y-scheme-Z-H-16.html#node_ch… > > > > ----- > Cheers, > Sean > -- > Sent from: http://forum.world.st/Squeak-Beginners-f107673.html > _______________________________________________ > Beginners mailing list > Beginners(a)lists.squeakfoundation.org > http://lists.squeakfoundation.org/mailman/listinfo/beginners

3 3

Regarding Raven: DOCap System
by henry 31 Mar '18

31 Mar '18

@m_m#3045 for concurrency without shared state, running on a single green thread (smalltalk’s Process) with support for remote async sends to other images, there is my Raven [1], using Cryptography [2], ParrotTalk [3] and STON. This is an implementation of eright’s [4] ELib. I am working on implementing STON for Java, inside the ASN1 project [5] as I already have ParrotTalk in Java [6], binary interoperate with Squeak/Pharo, and my Raven work in progress in Java [7]. This model allows message send redirection to a promise or farRef (or nearRef) as an asynchronous send returning a promise, which supports whenResolved/whenBroken handlers to be registered per PromiseERef. I’ll be happy to answer any questions you may be having about Raven. Raven is an implementation of an Actor model called Communicating Event Loops, in some literature [8] [9]. I hope this helps! [1] http://www.squeaksource.com/Cryptography/Raven-HenryHouse.21.mcz [2] http://www.squeaksource.com/Cryptography/Cryptography-rww.115.mcz [3] http://www.squeaksource.com/Cryptography/ParrotTalk-rww.19.mcz [4] http://erights.org/elib/index.html [5] https://github.com/CallistoHouseLtd/ASN1 [6] https://github.com/CallistoHouseLtd/ParrotTalk [7] https://github.com/CallistoHouseLtd/Raven [8] https://eighty-twenty.org/2016/10/18/actors-hopl [9] http://soft.vub.ac.be/Publications/2016/vub-soft-tr-16-11.pdf Sent from ProtonMail Mobile

1 1

A fix for ProportionalSplitterMorph, take 2
by Stéphane Rollandin 31 Mar '18

31 Mar '18

Repeat: Attached is a fix for ProportionalSplitterMorph. Currently, moving a window splitter changes all relevant frame layout offsets; fractions are not touched. But offsets are absolute values, in pixels, and a frame with a large offset will be badly redimensionned when the system window extent is reduced. To see this, open a browser, and move the main horizontal splitter way up. Then shrink the browser: the code pane will keep a largeish height (because of its large vertical offsets), and eventually be moved above the browser top. To fix this, I propose in the attached changeset to keep track of the overall splitter delta, then on #mouseUp to recompute the fractions and reset the offsets to the typical small values they have. I don't know if I am being clear, but it works... try it. (the #balanceOffsets method may possibly be simplified I guess) Stef

2 1

The Inbox: Morphic-dtl.1408.mcz
by commits＠source.squeak.org 31 Mar '18

31 Mar '18

David T. Lewis uploaded a new version of Morphic to project The Inbox: http://source.squeak.org/inbox/Morphic-dtl.1408.mcz ==================== Summary ==================== Name: Morphic-dtl.1408 Author: dtl Time: 31 March 2018, 1:28:09.631739 pm UUID: b81c45a6-9270-4233-81ee-a070042bd449 Ancestors: Morphic-cmm.1407 To: "liste [Squeak-dev]" <squeak-dev(a)lists.squeakfoundation.org> From: St??phane Rollandin <lecteur(a)zogotounga.net> Date: Sat, 31 Mar 2018 10:47:47 +0200 Subject: [squeak-dev] A fix for ProportionalSplitterMorph, take 2 Attached is a fix for ProportionalSplitterMorph. Currently, moving a window splitter changes all relevant frame layout offsets; fractions are not touched. But offsets are absolute values, in pixels, and a frame with a large offset will be badly redimensionned when the system window extent is reduced. To see this, open a browser, and move the main horizontal splitter way up. Then shrink the browser: the code pane will keep a largeish height (because of its large vertical offsets), and eventually be moved above the browser top. To fix this, I propose in the attached changeset to keep track of the overall splitter delta, then on #mouseUp to recompute the fractions and reset the offsets to the typical small values they have. I don't know if I am being clear, but it works... try it. (the #balanceOffsets method may possibly be simplified I guess) Stef [-- Attachment #2: ProportionalSplitterMorph.spfa.2.cs --] =============== Diff against Morphic-cmm.1407 =============== Item was added: + ----- Method: ProportionalSplitterMorph>>balanceOffsets (in category 'layout') ----- + balanceOffsets + + self layoutFrame hasFixedHeight ifTrue: [ + | otop obot ctop cbot topf | + + otop := (owner submorphs detect: [:m | + m layoutFrame topFraction isZero] ifNone: [^ self]) in: [:tm | + tm top - (tm layoutFrame topOffset ifNil: [0])]. + + obot := (owner submorphs detect: [:m | + m layoutFrame bottomFraction = 1] ifNone: [^ self]) in: [:tm | + tm bottom - (tm layoutFrame bottomOffset ifNil: [0])]. + + ctop := (self layoutFrame topFraction * (obot - otop)) rounded + + otop + (self layoutFrame topOffset ifNil: [0]). + cbot := (self layoutFrame bottomFraction * (obot - otop)) rounded + + otop + (self layoutFrame bottomOffset ifNil: [0]). + + topf := self layoutFrame topFraction. + self layoutFrame topFraction: ((ctop + cbot) * 0.5 - otop) / (obot - otop) asFloat. + self layoutFrame bottomFraction: self layoutFrame topFraction. + self layoutFrame topOffset: ctop - + (self layoutFrame topFraction * (obot - otop) + otop) rounded. + self layoutFrame bottomOffset: cbot - + (self layoutFrame bottomFraction * (obot - otop) + otop) rounded.. + + (leftOrTop copy union: rightOrBottom) do: [:m | + (m layoutFrame topFraction closeTo: topf) ifTrue: [ + m layoutFrame topFraction: self layoutFrame topFraction. + m layoutFrame topOffset: + m layoutFrame topOffset - (self valueOfProperty: #fullDelta) y]. + (m layoutFrame bottomFraction closeTo: topf) ifTrue: [ + m layoutFrame bottomFraction: self layoutFrame topFraction. + m layoutFrame bottomOffset: + m layoutFrame bottomOffset - (self valueOfProperty: #fullDelta) y.]]] . + + self layoutFrame hasFixedWidth ifTrue: [ + | oleft oright cleft cright leftf | + + oleft := (owner submorphs detect: [:m | + m layoutFrame leftFraction isZero] ifNone: [^ self]) in: [:tm | + tm left - (tm layoutFrame leftOffset ifNil: [0])]. + + oright := (owner submorphs detect: [:m | + m layoutFrame rightFraction = 1] ifNone: [^ self]) in: [:tm | + tm right - (tm layoutFrame rightOffset ifNil: [0])]. + + cleft := (self layoutFrame leftFraction * (oright - oleft)) rounded + + oleft + (self layoutFrame leftOffset ifNil: [0]). + cright := (self layoutFrame rightFraction * (oright - oleft)) rounded + + oleft + (self layoutFrame rightOffset ifNil: [0]). + + leftf := self layoutFrame leftFraction. + self layoutFrame leftFraction: ((cleft + cright) * 0.5 - oleft) / (oright - oleft) asFloat. + self layoutFrame rightFraction: self layoutFrame leftFraction. + + self layoutFrame leftOffset: cleft - + (self layoutFrame leftFraction * (oright - oleft) + oleft) rounded. + self layoutFrame rightOffset: cright - + (self layoutFrame rightFraction * (oright - oleft) + oleft) rounded. + + (leftOrTop copy union: rightOrBottom) do: [:m | + (m layoutFrame leftFraction closeTo: leftf) ifTrue: [ + m layoutFrame leftFraction: self layoutFrame leftFraction. + m layoutFrame leftOffset: + m layoutFrame leftOffset - (self valueOfProperty: #fullDelta) x]. + (m layoutFrame rightFraction closeTo: leftf) ifTrue: [ + m layoutFrame rightFraction: self layoutFrame leftFraction. + m layoutFrame rightOffset: + m layoutFrame rightOffset - (self valueOfProperty: #fullDelta) x.]]] . + + self removeProperty: #fullDelta. + owner layoutChanged + ! Item was changed: ----- Method: ProportionalSplitterMorph>>mouseDown: (in category 'events') ----- mouseDown: anEvent "If the user manually drags me, don't override him with auto positioning." + self setProperty: #fullDelta toValue: 0@0. + anEvent redButtonChanged ifTrue: [ self withSiblingSplittersDo: [ : each | each stopStepping ] ] ifFalse: [ anEvent shiftPressed ifTrue: [ self startStepping ] ifFalse: [ self startStepping. self withSiblingSplittersDo: [ : each | each startStepping ] ] ]. (self class showSplitterHandles not and: [ self bounds containsPoint: anEvent cursorPoint ]) ifTrue: [ oldColor := self color. self color: Color black ]. ^ super mouseDown: anEvent! Item was changed: ----- Method: ProportionalSplitterMorph>>mouseUp: (in category 'events') ----- mouseUp: anEvent (self bounds containsPoint: anEvent cursorPoint) ifFalse: [anEvent hand showTemporaryCursor: nil]. self class fastSplitterResize ifTrue: [self updateFromEvent: anEvent]. traceMorph ifNotNil: [traceMorph delete. traceMorph := nil]. + self color: self getOldColor. + + self balanceOffsets! - self color: self getOldColor! Item was changed: ----- Method: ProportionalSplitterMorph>>repositionBy: (in category 'layout') ----- repositionBy: delta | selfTop selfBottom selfLeft selfRight | + + self setProperty: #fullDelta toValue: ((self valueOfProperty: #fullDelta) ifNil: [0@0]) + delta. + leftOrTop do: [ : each | | firstRight firstBottom firstLeft firstTop | firstRight := each layoutFrame rightOffset ifNil: [ 0 ]. firstBottom := each layoutFrame bottomOffset ifNil: [ 0 ]. each layoutFrame rightOffset: firstRight + delta x. each layoutFrame bottomOffset: firstBottom + delta y. each layoutFrame hasFixedHeight ifTrue: [ firstTop := each layoutFrame topOffset ifNil: [ 0 ]. each layoutFrame topOffset: firstTop + delta y ]. each layoutFrame hasFixedWidth ifTrue: [ firstLeft := each layoutFrame leftOffset ifNil: [ 0 ]. each layoutFrame leftOffset: firstLeft + delta x. ] ]. rightOrBottom do: [ : each | | secondLeft secondTop secondRight secondBottom | secondLeft := each layoutFrame leftOffset ifNil: [ 0 ]. secondTop := each layoutFrame topOffset ifNil: [ 0 ]. each layoutFrame leftOffset: secondLeft + delta x. each layoutFrame topOffset: secondTop + delta y. each layoutFrame hasFixedHeight ifTrue: [ secondBottom := each layoutFrame bottomOffset ifNil: [ 0 ]. each layoutFrame bottomOffset: secondBottom + delta y. ]. each layoutFrame hasFixedWidth ifTrue: [ secondRight := each layoutFrame rightOffset ifNil: [ 0 ]. each layoutFrame rightOffset: secondRight + delta x. ] ]. selfTop := self layoutFrame topOffset ifNil: [ 0 ]. selfBottom := self layoutFrame bottomOffset ifNil: [ 0 ]. selfLeft := self layoutFrame leftOffset ifNil: [ 0 ]. selfRight := self layoutFrame rightOffset ifNil: [ 0 ]. self layoutFrame topOffset: selfTop + delta y ; bottomOffset: selfBottom + delta y ; leftOffset: selfLeft + delta x ; rightOffset: selfRight + delta x. self owner layoutChanged. self movements removeFirst; add: (splitsTopAndBottom ifTrue: [ delta y sign ] ifFalse: [ delta x sign ])!

1 0

The Trunk: Morphic-cmm.1407.mcz
by commits＠source.squeak.org 31 Mar '18

31 Mar '18

Chris Muller uploaded a new version of Morphic to project The Trunk: http://source.squeak.org/trunk/Morphic-cmm.1407.mcz ==================== Summary ==================== Name: Morphic-cmm.1407 Author: cmm Time: 30 March 2018, 5:59:05.775759 pm UUID: d60bae6d-7ac8-4ac6-90d4-ef2adbbe7d68 Ancestors: Morphic-mt.1406 - Allow particular subwidgets of dialogs to be positioned directly under hand instead of only under the center of the entire dialog (under the hand). - Fix for PluggableMultiColumnListMorph when Smart Splitters is enabled. =============== Diff against Morphic-mt.1406 =============== Item was changed: ----- Method: DialogWindow>>getUserResponse (in category 'running') ----- getUserResponse | hand world | (ProvideAnswerNotification signal: self message asString) ifNotNil: [:answer| ^ answer = #default ifTrue: [result] ifFalse: [answer]]. self message ifEmpty: [messageMorph delete]. "Do not waste space." self paneMorph submorphs ifEmpty: [self paneMorph delete]. "Do not waste space." hand := self currentHand. world := self currentWorld. self fullBounds. + self moveToPreferredPosition. - self center: preferredPosition. - self bounds: (self bounds translatedToBeWithin: world bounds). self openInWorld: world. hand showTemporaryCursor: nil. "Since we are out of context, reset the cursor." hand keyboardFocus in: [:priorKeyboardFocus | self exclusive ifTrue: [hand newMouseFocus: self]. hand newKeyboardFocus: self. [self isInWorld] whileTrue:[world doOneSubCycle]. hand newKeyboardFocus: priorKeyboardFocus. hand releaseMouseFocus]. ^ result! Item was changed: ----- Method: DialogWindow>>getUserResponseAtHand (in category 'running') ----- getUserResponseAtHand + "Remove unnecessary widgets." + self message ifEmpty: [ messageMorph delete ]. + self paneMorph submorphs ifEmpty: [ self paneMorph delete ]. + "Modal dialogs must be as convenient as possible, put default button directly under hand." + self + ensureSelectedButton ; + preferredPosition: self selectedButton. + ^ self getUserResponse! - - ^ self getUserResponseAtHand: ActiveHand! Item was removed: - ----- Method: DialogWindow>>getUserResponseAtHand: (in category 'running') ----- - getUserResponseAtHand: aHand - - self message ifEmpty: [messageMorph delete]. "Do not waste space." - self paneMorph submorphs ifEmpty: [self paneMorph delete]. "Do not waste space." - - self moveSelectedButtonToHand: aHand. - ^ self getUserResponse! Item was removed: - ----- Method: DialogWindow>>moveSelectedButtonToHand: (in category 'position') ----- - moveSelectedButtonToHand: aHand - "Just let the user confirm the selected button without having to reposition the mouse." - - self ensureSelectedButton. - self moveTo: self fullBounds center + (aHand position - self selectedButton center).! Item was removed: - ----- Method: DialogWindow>>moveTo: (in category 'position') ----- - moveTo: position - - preferredPosition := position.! Item was changed: ----- Method: DialogWindow>>moveToHand: (in category 'position') ----- + moveToHand: aHand + self preferredPosition: aHand position! - moveToHand: aHand - - self moveTo: aHand position.! Item was added: + ----- Method: DialogWindow>>moveToPreferredPosition (in category 'initialization') ----- + moveToPreferredPosition + self center: + (preferredPosition isPoint + ifTrue: [ preferredPosition ] + ifFalse: [ self center + ActiveHand position - preferredPosition center ]). + self bounds: (self bounds translatedToBeWithin: self currentWorld bounds)! Item was added: + ----- Method: DialogWindow>>positionOverWidgetNamed: (in category 'initialization') ----- + positionOverWidgetNamed: nameSymbol + self allMorphsDo: + [ : each | each knownName = nameSymbol ifTrue: [ self preferredPosition: each ] ]! Item was added: + ----- Method: DialogWindow>>preferredPosition: (in category 'position') ----- + preferredPosition: aPointOrMorph + "Specify that the dialog should be positioned at an absolute Point on the screen, or the particular Morph widget which should be positioned under the hand." + preferredPosition := aPointOrMorph! Item was changed: ----- Method: FillInTheBlankMorph class>>request:initialAnswer:centerAt:inWorld:onCancelReturn:acceptOnCR:answerExtent: (in category 'instance creation') ----- request: queryString initialAnswer: defaultAnswer centerAt: aPoint inWorld: aWorld onCancelReturn: returnOnCancel acceptOnCR: acceptBoolean answerExtent: answerExtent "Create an instance of me whose question is queryString with the given initial answer. Invoke it centered at the given point, and answer the string the user accepts. If the user cancels, answer returnOnCancel." "FillInTheBlankMorph request: 'Type something, then type CR.' initialAnswer: 'yo ho ho!!' centerAt: Display center" | aFillInTheBlankMorph | aFillInTheBlankMorph := self new setQuery: queryString initialAnswer: defaultAnswer answerExtent: answerExtent acceptOnCR: acceptBoolean. aFillInTheBlankMorph createAcceptButton action: [aFillInTheBlankMorph textPane accept]. aFillInTheBlankMorph createCancelButton action: [aFillInTheBlankMorph closeDialog: returnOnCancel]. + aFillInTheBlankMorph preferredPosition: aPoint. - aFillInTheBlankMorph moveTo: aPoint. ^ aFillInTheBlankMorph getUserResponse ! Item was changed: ----- Method: FillInTheBlankMorph class>>requestPassword:initialAnswer:centerAt:inWorld:onCancelReturn:acceptOnCR: (in category 'instance creation') ----- requestPassword: queryString initialAnswer: defaultAnswer centerAt: aPoint inWorld: aWorld onCancelReturn: returnOnCancel acceptOnCR: acceptBoolean "Create an instance of me whose question is queryString with the given initial answer. Invoke it centered at the given point, and answer the string the user accepts. If the user cancels, answer returnOnCancel." "FillInTheBlankMorph request: 'Type something, then type CR.' initialAnswer: 'yo ho ho!!' centerAt: Display center" | aFillInTheBlankMorph | aFillInTheBlankMorph := self new setPasswordQuery: queryString initialAnswer: defaultAnswer answerHeight: 50 acceptOnCR: acceptBoolean. aFillInTheBlankMorph createAcceptButton action: [aFillInTheBlankMorph textPane accept]. aFillInTheBlankMorph createCancelButton action: [aFillInTheBlankMorph closeDialog: returnOnCancel]. + aFillInTheBlankMorph preferredPosition: aPoint. - aFillInTheBlankMorph moveTo: aPoint. ^ aFillInTheBlankMorph getUserResponse! Item was added: + ----- Method: PluggableMultiColumnListMorph>>charactersOccluded (in category 'private') ----- + charactersOccluded + "Not meaningful in multi-column lists, since they should truncate their column widths according to how much space is needed vs. available to show a bit of each." + ^ 0! Item was changed: ----- Method: UserDialogBoxMorph class>>confirm:orCancel:title:at: (in category 'utilities') ----- confirm: aString orCancel: cancelBlock title: titleString at: aPointOrNil (self new title: titleString; message: aString; createButton: 'Yes' translated value: true; createButton: 'No' translated value: false; createButton: 'Cancel' translated value: nil; selectedButtonIndex: 1; "YES" registerKeyboardShortcuts; yourself) in: [:dialog | ^ (aPointOrNil ifNil: [dialog getUserResponseAtHand] ifNotNil: [ + dialog preferredPosition: aPointOrNil. - dialog moveTo: aPointOrNil. dialog getUserResponse]) ifNil: [ cancelBlock value ]]! Item was changed: ----- Method: UserDialogBoxMorph class>>confirm:title:trueChoice:falseChoice:at: (in category 'utilities') ----- confirm: aString title: titleString trueChoice: trueChoice falseChoice: falseChoice at: aPointOrNil "UserDialogBoxMorph confirm: 'Make your choice carefully' withCRs title: 'Do you like chocolate?' trueChoice: 'Oh yessir!!' falseChoice: 'Not so much...'" ^self new title: titleString; message: aString; createButton: trueChoice translated value: true; createCancelButton: falseChoice translated value: false; selectedButtonIndex: 1; registerKeyboardShortcuts; + preferredPosition: (aPointOrNil ifNil: [ActiveWorld center]); - moveTo: (aPointOrNil ifNil: [ActiveWorld center]); getUserResponse! Item was changed: ----- Method: UserDialogBoxMorph class>>confirm:title:trueChoice:falseChoice:default:triggerAfter:at: (in category 'utilities') ----- confirm: aString title: titleString trueChoice: trueChoice falseChoice: falseChoice default: default triggerAfter: seconds at: aPointOrNil "UserDialogBoxMorph confirm: 'I like hot java' title: 'What do you say?' trueChoice: 'You bet!!' falseChoice: 'Nope' default: false triggerAfter: 12 at: 121@212" ^self new title: titleString; message: aString; createButton: trueChoice translated value: true; createCancelButton: falseChoice translated value: false; selectedButtonIndex: (default ifTrue: [1] ifFalse: [2]); registerKeyboardShortcuts; + preferredPosition: (aPointOrNil ifNil: [ActiveWorld center]); - moveTo: (aPointOrNil ifNil: [ActiveWorld center]); getUserResponseAfter: seconds!

2 1

The Trunk: Compiler-eem.380.mcz
by commits＠source.squeak.org 31 Mar '18

31 Mar '18

Eliot Miranda uploaded a new version of Compiler to project The Trunk: http://source.squeak.org/trunk/Compiler-eem.380.mcz ==================== Summary ==================== Name: Compiler-eem.380 Author: eem Time: 20 March 2018, 3:30:10.256928 pm UUID: 3133d60f-54b2-410e-92ae-ef5cc782ab9c Ancestors: Compiler-eem.379 Rename Encoder's addedSelectorAndMethodClassLiterals, a name which is now misleading given the new full block support, to addedExtraLiterals. Requires Compiler-eem.379. =============== Diff against Compiler-eem.379 =============== Item was changed: ----- Method: BytecodeEncoder>>allLiteralsForBlockMethod (in category 'results') ----- allLiteralsForBlockMethod + addedExtraLiterals ifFalse: + [addedExtraLiterals := true. - "N.B. addedSelectorAndMethodClassLiterals is a misnomer. It should be something like - addedImplicitLiterals or addedExtraLiterals." - addedSelectorAndMethodClassLiterals ifFalse: - [addedSelectorAndMethodClassLiterals := true. "Put the optimized selectors in literals so as to browse senders more easily" optimizedSelectors := optimizedSelectors reject: [:e| literalStream originalContents hasLiteral: e]. optimizedSelectors isEmpty ifFalse: [ "Use one entry per literal if enough room, else make anArray" literalStream position + optimizedSelectors size + 2 >= self maxNumLiterals ifTrue: [self litIndex: optimizedSelectors asArray] ifFalse: [optimizedSelectors do: [:e | self litIndex: e]]]. "Add a slot for outerCode" self litIndex: nil]. ^literalStream contents! Item was changed: ----- Method: BytecodeEncoder>>resetForFullBlockGeneration (in category 'code generation') ----- resetForFullBlockGeneration literalStream := WriteStream on: (Array new: 8). + addedExtraLiterals := false. - addedSelectorAndMethodClassLiterals := addedExtraLiterals := false. optimizedSelectors := Set new! Item was changed: ----- Method: BytecodeEncoder>>resetLiteralStreamForFullBlock (in category 'code generation') ----- resetLiteralStreamForFullBlock literalStream := WriteStream on: (Array new: 32). + addedExtraLiterals := false. - addedSelectorAndMethodClassLiterals := addedExtraLiterals := false. optimizedSelectors := Set new! Item was changed: ParseNode subclass: #Encoder + instanceVariableNames: 'scopeTable nTemps supered requestor class selector literalStream selectorSet litIndSet litSet sourceRanges globalSourceRanges addedExtraLiterals optimizedSelectors cue' - instanceVariableNames: 'scopeTable nTemps supered requestor class selector literalStream selectorSet litIndSet litSet sourceRanges globalSourceRanges addedSelectorAndMethodClassLiterals optimizedSelectors cue' classVariableNames: '' poolDictionaries: '' category: 'Compiler-Kernel'! !Encoder commentStamp: 'cwp 12/26/2012 23:29' prior: 0! I encode names and literals into tree nodes with byte codes for the compiler. Byte codes for literals are not assigned until the tree-sizing pass of the compiler, because only then is it known which literals are actually needed. I also keep track of sourceCode ranges during parsing and code generation so I can provide an inverse map for the debugger.! Item was changed: ----- Method: Encoder>>allLiterals (in category 'results') ----- allLiterals + addedExtraLiterals ifFalse: + [addedExtraLiterals := true. - addedSelectorAndMethodClassLiterals ifFalse: - [addedSelectorAndMethodClassLiterals := true. "Put the optimized selectors in literals so as to browse senders more easily" optimizedSelectors := optimizedSelectors reject: [:e| literalStream originalContents hasLiteral: e]. optimizedSelectors isEmpty ifFalse: [ "Use one entry per literal if enough room, else make anArray" literalStream position + optimizedSelectors size + 2 >= self maxNumLiterals ifTrue: [self litIndex: optimizedSelectors asArray] ifFalse: [optimizedSelectors do: [:e | self litIndex: e]]]. "Add a slot for selector or MethodProperties" self litIndex: nil. self litIndex: self associationForClass]. ^literalStream contents! Item was changed: ----- Method: Encoder>>initScopeAndLiteralTables (in category 'initialize-release') ----- initScopeAndLiteralTables scopeTable := StdVariables copy. litSet := StdLiterals copy. "comments can be left hanging on nodes from previous compilations. probably better than this hack fix is to create the nodes afresh on each compilation." scopeTable do: [:varNode| varNode comment: nil]. litSet do: [:varNode| varNode comment: nil]. selectorSet := StdSelectors copy. litIndSet := Dictionary new: 16. literalStream := WriteStream on: (Array new: 32). + addedExtraLiterals := false. - addedSelectorAndMethodClassLiterals := addedExtraLiterals := false. optimizedSelectors := Set new!

4 9

A fix for ProportionalSplitterMorph
by Stéphane Rollandin 31 Mar '18

31 Mar '18

Attached is a fix for ProportionalSplitterMorph. Currently, moving a window splitter changes all relevant frame layout offsets; fractions are not touched. But offsets are absolute values, in pixels, and a frame with a large offset will be badly redimensionned when the system window extent is reduced. To see this, open a browser, and move the main horizontal splitter way up. Then shrink the browser: the code pane will keep a largeish height (because of its large vertical offsets), and eventually be moved above the browser top. To fix this, I propose in the attached changeset to keep track of the overall splitter delta, then on #mouseUp to recompute the fractions and reset the offsets to the typical small values they have. I don't know if I am being clear, but it works... try it. (the #balanceOffsets method may possibly be simplified I guess) Stef

2 2

Fwd: OSProcess Feeback
by Sean P. DeNigris 29 Mar '18

29 Mar '18

I assume this is still the best place? This was posted by Luke Gorrie on the Pharo #general Discord channel, which I guess David will never see… > OSProcess is not working out that well for me. I know exactly what system > calls I want to make (read, write, select/poll, etc) and when I want to > make them. The challenge is working out how to make the object-oriented > interface do what I want. For example I was totally surprised yesterday > that process upToEnd only returns output that was previously read from the > process somehow and does not actually check for new output. > It's also unhelpful that many of the convenience methods, like > outputAndError, can deadlock because they are just not written correctly > with respect to how Unix works e.g. that you have to read from stdout and > stderr in parallel and can't do one to completion and then the other. > I don't like to complain because obviously a lot of work goes into the > framework but I don't think it is really the right tool for the job for my > own application. Could be more suitable if I were running many processes > and piping them together in Smalltalk, in the spirit of a Unix shell, etc. > But I am just running one subprocess at a time and processing its output > stream. Also the question was raised whether OSP is still actively maintained for Pharo. Thanks! ----- Cheers, Sean -- Sent from: http://forum.world.st/Squeak-Dev-f45488.html

2 1

kudos to Eliot
by gettimothy 29 Mar '18

29 Mar '18

Eliot, >From the cheap seats, your work and dedication is wonderful. You are a great role-model for what a programmer should be and you have set standards that I only aspire to at the moment. Thanks for all you have done and are doing. Smalltalk is a great language and your VM kicks butt. cheers. tty

11 10

Jump to page: