Squeak-dev June 2013

squeak-dev@lists.squeakfoundation.org

45 participants
215 discussions

OpenGL procedural textures generator (by David Faught) for Squeak 4.4
by Nikolay Suslov 09 Jun '14

09 Jun '14

Hello, I have successfully proceeded in running TweakCore on the recent Squeak 4.4 trunk image (from Jenkins). And one of the famous existed applications developed in Tweak is the OpenGL procedural textures generator by David Faught. I make it also loadable to the current Squeak. You could download the ready to run image from here: http://krestianstvo.org/TweakCoreOpenGL-squeak44.zip or execute in the workspace in own image: "1. Load FFI" (Installer repository: 'http://source.squeak.org/FFI' ) install: 'FFI-Pools'; install: 'FFI-Kernel'; install: 'FFI-Tests'.. "2. Load CroquetGL " (Installer repository: 'http://www.squeaksource.com/CroquetGL') install: '3DTransform'; install: 'OpenGL-Pools'; install: 'OpenGL-Core'. "3. Load TweakCore and Procedural textures application for Tweak" (Installer repository: 'http://sdk.krestianstvo.org/sdk/inbox') install: 'tweakcore'; install: 'Tweak-OpenGL-sn.3'. "4. Set the default settings in Tweak" CDefaultWidgetLibrary setDefaultSettings. "5. Run one of two examples" CProjectMorph open: Wrinkle1. "or" CProjectMorph open: Wrinkle2. The attached screenshot shows the running application. Regards, Nikolay

7 9

Mantis usage rules du jour
by tim Rowledge 19 Feb '14

19 Feb '14

I'm submitting a bug fix and thought I'd ask what the current favoured procedure is. Simply submitting is as irritating as it ever was in Mantis, but once I've done that, uploaded my fix and made a note about it… what do we do next? To whom might I assign it for action in getting into the image for the next release? Or is that done some other way now? Is there documentation on the current process that I failed to spot? tim -- tim Rowledge; tim(a)rowledge.org; http://www.rowledge.org/tim Strange OpCodes: XO: Execute Operator

15 34

Re: [squeak-dev] The Inbox: Compiler.quasiquote-eem.248.mcz
by Eliot Miranda 05 Feb '14

05 Feb '14

Hi All, I had fun implementing a quasi-quote for Squeak today. This is a convenient way of embedding substrings in format strings (a little like printf), and, because it uses a different quote character, a convenient way of embedding code form other languages in a string literal. An example of the former usage is `hello [#cruel] world` which evaluates to 'hello cruel world' And `Float pi is [Float pi]` evaluates to 'Float pi is 3.141592653589793' An example of the latter use is that one can write `printf("%s: %c\\n", "a string", 'C');` instead of 'printf("%s: %c\n", "a string", ''C'');' This last example shows a limitation; The use of \ to escape characters ($\ $[ and $`) in quasi-quote might not be such a good choice. Anyway I thought I'd put this in the in-box for people to play with and savage. Please let me know what you think, both about the semantics and the implementation. This is a quick hack and I'm sure that there's plenty of scope for clean-up. cheers Eliot On Tue, Feb 5, 2013 at 9:54 PM, <commits(a)source.squeak.org> wrote: > A new version of Compiler was added to project The Inbox: > http://source.squeak.org/inbox/Compiler.quasiquote-eem.248.mcz > > ==================== Summary ==================== > > Name: Compiler.quasiquote-eem.248 > Author: eem > Time: 5 February 2013, 9:54:20.317 pm > UUID: ef044906-3339-48cc-856b-9b5172e3e81b > Ancestors: Compiler-cwp.247 > > Add a quasi-quote form that allows convenient embedding > of substrings within a format string, and provides a > convenient way of embedding literal strings within an > alternative literal string whose string delimiter is different. > > e.g. > `hello [#cruel] world!` > evaluates to > 'hello cruel world'. > > `S1[B1]...SN[BN]SN+1` > > is equivalent to > { 'S1'. [B1] value. ... 'SN'. [BN] value. 'SN+1' } > concatenateQuasiQuote > where concatenateQuasiQuote sends asString to each > element and answers the concatenation of those elements. > > however, single-statement blocks are inlined, so e.g. the > above `hello [#cruel] world!` is compiled as > { 'hello '. #cruel. ' world!' } concatenateQuasiQuote > > See Tests.quasiquote-eem.188 for tests and examples. > > =============== Diff against Compiler-cwp.247 =============== > > Item was added: > + ----- Method: Array>>concatenateQuasiQuote (in category > '*Compiler-support') ----- > + concatenateQuasiQuote > + "This method is used in compilation of quasi-quote constructs. > + It MUST NOT be deleted or altered." > + > + | s sz | > + sz := self size. > + s := WriteStream on: (String new: sz * 16). > + 1 to: sz do: > + [:i| s nextPutAll: (self at: i) asString]. > + ^s contents! > > Item was removed: > - ----- Method: Decompiler>>checkForBlock:selector:arguments: (in category > 'control') ----- > - checkForBlock: receiver selector: selector arguments: arguments > - selector == #blockCopy: ifTrue: > - [^self checkForBlockCopy: receiver]. > - self assert: selector == #closureCopy:copiedValues:. > - ^self checkForClosureCopy: receiver arguments: arguments! > > Item was added: > + ----- Method: Decompiler>>checkForMacroMessage:selector:arguments: (in > category 'control') ----- > + checkForMacroMessage: rcvr selector: selector arguments: args > + ^ (selector == #concatenateQuasiQuote > + and: [self checkForQuasiQuote: rcvr selector: selector > arguments: args]) > + or: [(#closureCopy:copiedValues: == selector > + and: [self checkForClosureCopy: rcvr arguments: args]) > + or: [#blockCopy: == selector > + and: [self checkForBlockCopy: rcvr]]]! > > Item was added: > + ----- Method: Decompiler>>checkForQuasiQuote:selector:arguments: (in > category 'control') ----- > + checkForQuasiQuote: rcvr "<BraceNode>" selector: selector "<Symbol>" > arguments: args "<Array>" > + stack addLast: > + ((MessageNode new > + receiver: rcvr > + selector: (SelectorNode new key: > #concatenateQuasiQuote code: nil) > + arguments: args > + precedence: 1) > + notePrintingSelector: #printQuasiQuoteOn:indent:; > + yourself). > + ^true! > > Item was changed: > ----- Method: Decompiler>>send:super:numArgs: (in category 'instruction > decoding') ----- > send: selector super: superFlag numArgs: numArgs > > | args rcvr selNode msgNode messages | > args := Array new: numArgs. > (numArgs to: 1 by: -1) do: > [:i | args at: i put: stack removeLast]. > rcvr := stack removeLast. > superFlag ifTrue: [rcvr := constructor codeSuper]. > + (self checkForMacroMessage: rcvr selector: selector arguments: > args) ifFalse: > - ((#(blockCopy: closureCopy:copiedValues:) includes: selector) > - and: [self checkForBlock: rcvr selector: selector arguments: > args]) ifFalse: > [selNode := constructor codeAnySelector: selector. > rcvr == CascadeFlag > ifTrue: > ["May actually be a cascade or an ifNil: > for value." > self willJumpIfFalse > ifTrue: "= generated by a case > macro" > [selector == #= ifTrue: > [" = signals a > case statement..." > statements > addLast: args first. > stack addLast: > rcvr. "restore CascadeFlag" > ^ self]. > selector == #== ifTrue: > [" == signals an > ifNil: for value..." > stack removeLast; > removeLast. > rcvr := stack > removeLast. > stack addLast: > IfNilFlag; > addLast: > (constructor > > codeMessage: rcvr > > selector: selNode > > arguments: args). > ^ self]] > ifFalse: > [(self willJumpIfTrue and: > [selector == #==]) ifTrue: > [" == signals an > ifNotNil: for value..." > stack removeLast; > removeLast. > rcvr := stack > removeLast. > stack addLast: > IfNilFlag; > addLast: > (constructor > > codeMessage: rcvr > > selector: selNode > > arguments: args). > ^ self]]. > msgNode := constructor > > codeCascadedMessage: selNode > arguments: > args. > stack last == CascadeFlag ifFalse: > ["Last message of a cascade" > statements addLast: msgNode. > messages := self popTo: stack > removeLast. "Depth saved by first dup" > msgNode := constructor > > codeCascade: stack removeLast > > messages: messages]] > ifFalse: > [msgNode := constructor > codeMessage: rcvr > selector: selNode > arguments: args]. > stack addLast: msgNode]! > > Item was changed: > ----- Method: MessageNode class>>initialize (in category 'class > initialization') ----- > initialize > "MessageNode initialize" > MacroSelectors := > #( ifTrue: ifFalse: ifTrue:ifFalse: ifFalse:ifTrue: > and: or: > whileFalse: whileTrue: whileFalse whileTrue > to:do: to:by:do: > caseOf: caseOf:otherwise: > ifNil: ifNotNil: ifNil:ifNotNil: ifNotNil:ifNil: > + repeat > + nil "space for concatenateQuasiQuote" ). > - repeat ). > MacroTransformers := > #( transformIfTrue: transformIfFalse: > transformIfTrueIfFalse: transformIfFalseIfTrue: > transformAnd: transformOr: > transformWhile: transformWhile: transformWhile: > transformWhile: > transformToDo: transformToDo: > transformCase: transformCase: > transformIfNil: transformIfNil: > transformIfNilIfNotNil: transformIfNotNilIfNil: > + transformRepeat: > + nil "space for concatenateQuasiQuote" ). > - transformRepeat: ). > MacroEmitters := > #( emitCodeForIf:encoder:value: > emitCodeForIf:encoder:value: > emitCodeForIf:encoder:value: > emitCodeForIf:encoder:value: > emitCodeForIf:encoder:value: > emitCodeForIf:encoder:value: > emitCodeForWhile:encoder:value: > emitCodeForWhile:encoder:value: > emitCodeForWhile:encoder:value: > emitCodeForWhile:encoder:value: > emitCodeForToDo:encoder:value: > emitCodeForToDo:encoder:value: > emitCodeForCase:encoder:value: > emitCodeForCase:encoder:value: > emitCodeForIfNil:encoder:value: > emitCodeForIfNil:encoder:value: > emitCodeForIf:encoder:value: > emitCodeForIf:encoder:value: > + emitCodeForRepeat:encoder:value: > + nil "space for concatenateQuasiQuote"). > - emitCodeForRepeat:encoder:value:). > MacroSizers := > #( sizeCodeForIf:value: sizeCodeForIf:value: > sizeCodeForIf:value: sizeCodeForIf:value: > sizeCodeForIf:value: sizeCodeForIf:value: > sizeCodeForWhile:value: sizeCodeForWhile:value: > sizeCodeForWhile:value: sizeCodeForWhile:value: > sizeCodeForToDo:value: sizeCodeForToDo:value: > sizeCodeForCase:value: sizeCodeForCase:value: > sizeCodeForIfNil:value: sizeCodeForIfNil:value: > sizeCodeForIf:value: sizeCodeForIf:value: > + sizeCodeForRepeat:value: > + nil "space for concatenateQuasiQuote"). > - sizeCodeForRepeat:value:). > MacroPrinters := > #( printIfOn:indent: printIfOn:indent: > printIfOn:indent: printIfOn:indent: > printIfOn:indent: printIfOn:indent: > printWhileOn:indent: printWhileOn:indent: > printWhileOn:indent: printWhileOn:indent: > printToDoOn:indent: printToDoOn:indent: > printCaseOn:indent: printCaseOn:indent: > printIfNil:indent: printIfNil:indent: > printIfNilNotNil:indent: printIfNilNotNil:indent: > + printRepeatOn:indent: > + printQuasiQuoteOn:indent:)! > - printRepeatOn:indent:)! > > Item was added: > + ----- Method: MessageNode>>notePrintingSelector: (in category 'macro > transformations') ----- > + notePrintingSelector: printingSelectorSymbol > + "decompile" > + > + special := MacroPrinters indexOf: printingSelectorSymbol! > > Item was added: > + ----- Method: MessageNode>>printQuasiQuoteOn:indent: (in category > 'printing') ----- > + printQuasiQuoteOn: aStream indent: level > + aStream nextPut: $`. > + receiver elements do: > + [:parseNode| > + (parseNode isLiteralNode > + and: [parseNode key class == 'literal' class]) > + ifTrue: > + [parseNode key do: > + [:char| > + ('`[\' includes: char) ifTrue: > + [aStream nextPut: $\]. > + aStream nextPut: char]] > + ifFalse: > + [(parseNode isMessageNode > + and: [parseNode selector key == #value > + and: [parseNode receiver isBlockNode]]) > + ifTrue: > + [parseNode receiver > printOn: aStream indent: 0] > + ifFalse: > + [aStream nextPut: $[. > + parseNode printOn: > aStream indent: 0. > + aStream nextPut: $]]]]. > + aStream nextPut: $`! > > Item was changed: > ----- Method: Parser>>advance (in category 'scanning') ----- > advance > | this | > prevMark := hereMark. > prevEnd := hereEnd. > this := here. > here := token. > hereType := tokenType. > hereMark := mark. > hereEnd := source position - (aheadChar == DoItCharacter > ifTrue: [hereChar == DoItCharacter > ifTrue: [0] > ifFalse: [1]] > ifFalse: [2]). > + hereType ~~ #backQuote ifTrue: > + [self scanToken]. > - self scanToken. > "Transcript show: 'here: ', here printString, ' mark: ', hereMark > printString, ' end: ', hereEnd printString; cr." > ^this! > > Item was changed: > ----- Method: Parser>>expression (in category 'expression types') ----- > expression > > + (hereType == #word and: [tokenType == #leftArrow]) ifTrue: > + [^self assignment: self variable]. > + hereType == #backQuote > + ifTrue: [self quasiQuoteExpression] > + ifFalse: > + [hereType == #leftBrace > + ifTrue: [self braceExpression] > + ifFalse: > + [self primaryExpression ifFalse: > + [^false]]]. > + (self messagePart: 3 repeat: true) ifTrue: > + [hereType == #semicolon ifTrue: > + [self cascade]]. > + ^true! > - (hereType == #word and: [tokenType == #leftArrow]) > - ifTrue: [^ self assignment: self variable]. > - hereType == #leftBrace > - ifTrue: [self braceExpression] > - ifFalse: [self primaryExpression ifFalse: [^ false]]. > - (self messagePart: 3 repeat: true) > - ifTrue: [hereType == #semicolon ifTrue: [self cascade]]. > - ^ true! > > Item was added: > + ----- Method: Parser>>nonQuasiQuoteExpression (in category 'expression > types') ----- > + nonQuasiQuoteExpression > + > + (hereType == #word and: [tokenType == #leftArrow]) > + ifTrue: [^ self assignment: self variable]. > + hereType == #leftBrace > + ifTrue: [self braceExpression] > + ifFalse: [self primaryExpression ifFalse: [^ false]]. > + (self messagePart: 3 repeat: true) > + ifTrue: [hereType == #semicolon ifTrue: [self cascade]]. > + ^ true! > > Item was added: > + ----- Method: Parser>>quasiQuoteExpression (in category 'expression > types') ----- > + quasiQuoteExpression > + "`quasi-quote` > + => { elements } concatenateQuasiQuote > + => MessageNode receiver: BraceNode selector: > #concatenateQuasiQuote. > + > + The syntax of quasi-quote is > + quasi-quote := $` (characters | blockExpression) * $` > + characters := (unescapedCharacter | $\ escapedCharacter) * > + > + The semantics of quasi-quote are that each blockExpression is > evaluated > + left-to-right in the scope of the enclosing method or block. The > sequence > + of interspersed character sequences and expressions are > concatenated > + left-to-right, sending asString to each element immediately prior > to concatenation. > + The concatenation is then the result of the expression. It is > always a new string. > + > + The implementation inlines single-statement blocks into the brace > expression that > + comprises the receiver of concatenateQuasiQuote" > + > + | elements locations stringStream loc | > + elements := OrderedCollection new. > + locations := OrderedCollection new. > + stringStream := WriteStream on: (String new: 16). > + [loc := hereMark + requestorOffset. > + hereType == #doit ifTrue: > + [^self expected: 'back quote']. > + hereType == #leftBracket > + ifTrue: > + [self scanToken; advance. > + parseNode := nil. > + self blockExpression. > + parseNode statements size = 1 > + ifTrue: > + [elements addLast: parseNode > statements first] > + ifFalse: > + [elements addLast: (MessageNode new > + > receiver: parseNode > + > selector: #value > + > arguments: #() > + > precedence: 1 > + > from: encoder)]. > + source position: hereMark - 1. > + [source peek ~~ $]] whileTrue: > + [source position: source position - 1]. > + source next. > + self step; step. > + self setHereTypeForQuasiQuote. > + locations addLast: loc] > + ifFalse: > + [(self scanQuasiQuoteCharactersUsing: > stringStream) ifNotNil: > + [:lit| > + elements addLast: lit. > + locations addLast: loc]]. > + hereType ~~ #backQuote] whileTrue. > + parseNode := MessageNode new > + receiver: (BraceNode new elements: > elements sourceLocations: locations) > + selector: #concatenateQuasiQuote > + arguments: #() > + precedence: 1 > + from: encoder. > + self scanToken; advance. > + ^true! > > Item was changed: > + ----- Method: Parser>>queriedUnusedTemporaries (in category 'temps') > ----- > - ----- Method: Parser>>queriedUnusedTemporaries (in category 'accessing') > ----- > queriedUnusedTemporaries > > queriedUnusedTemporaries ifNil: > [queriedUnusedTemporaries := Dictionary new]. > ^queriedUnusedTemporaries! > > Item was added: > + ----- Method: Parser>>scanQuasiQuoteCharactersUsing: (in category > 'scanning') ----- > + scanQuasiQuoteCharactersUsing: stringStream > + "Answer the next non-empty sequence of characters in a quasi-quote > string, or nil, if none." > + stringStream reset. > + [hereChar ~~ $` and: [hereChar ~~ $[ and: [hereChar ~~ > DoItCharacter]]] whileTrue: > + [hereChar == $\ > + ifTrue: > + [stringStream nextPut: aheadChar. self > step] > + ifFalse: > + [stringStream nextPut: hereChar]. > + self step]. > + self setHereTypeForQuasiQuote. > + ^stringStream position > 0 ifTrue: > + [encoder encodeLiteral: stringStream contents]! > > Item was added: > + ----- Method: Parser>>setHereTypeForQuasiQuote (in category 'scanning') > ----- > + setHereTypeForQuasiQuote > + "Set hereType appropriately based on hereChar. Used only for > quasi-quote parsing." > + hereChar == $` > + ifTrue: > + [hereType := #backQuote. > + self step] > + ifFalse: > + [hereChar == $[ > + ifTrue: > + [hereType := #leftBracket. > + self step] > + ifFalse: > + [hereChar == DoItCharacter ifTrue: > + [hereType := #doit]]]! > > Item was changed: > + ----- Method: Parser>>tempsMark (in category 'temps') ----- > - ----- Method: Parser>>tempsMark (in category 'accessing') ----- > tempsMark > ^ tempsMark! > > Item was changed: > + ----- Method: Parser>>tempsMark: (in category 'temps') ----- > - ----- Method: Parser>>tempsMark: (in category 'accessing') ----- > tempsMark: aNumber > tempsMark := aNumber! > > Item was changed: > ----- Method: Scanner class>>initializeTypeTable (in category > 'initialization') ----- > initializeTypeTable > "self initializeTypeTable" > > | newTable | > newTable := Array new: 256 withAll: #xBinary. "default" > newTable atAll: #(9 10 12 13 32 ) put: #xDelimiter. "tab lf ff cr > space" > newTable atAll: ($0 asciiValue to: $9 asciiValue) put: #xDigit. > > 1 to: 255 > do: [:index | > (Character value: index) isLetter > ifTrue: [newTable at: index put: > #xLetter]]. > > newTable at: $" asciiValue put: #xDoubleQuote. > newTable at: $# asciiValue put: #xLitQuote. > newTable at: $$ asciiValue put: #xDollar. > newTable at: $' asciiValue put: #xSingleQuote. > + newTable at: $` asciiValue put: #backQuote. > newTable at: $: asciiValue put: #xColon. > newTable at: $( asciiValue put: #leftParenthesis. > newTable at: $) asciiValue put: #rightParenthesis. > newTable at: $. asciiValue put: #period. > newTable at: $; asciiValue put: #semicolon. > newTable at: $[ asciiValue put: #leftBracket. > newTable at: $] asciiValue put: #rightBracket. > newTable at: ${ asciiValue put: #leftBrace. > newTable at: $} asciiValue put: #rightBrace. > newTable at: $^ asciiValue put: #upArrow. > newTable at: $_ asciiValue put: #xUnderscore. > newTable at: $| asciiValue put: #verticalBar. > TypeTable := newTable "bon voyage!!"! > > > -- best, Eliot

3 5

Loading FFI is broken
by Frank Shearar 20 Nov '13

20 Nov '13

(Installer monticello mc: (MCHttpRepository new location: 'http://source.squeak.org/FFI')) install: 'FFI-Kernel-eem.24.mcz'. "There's a -tbn.25, but that's not important to this post" fails with an MNU: ExternalFunction class >> callingConventionFor:. As far as I can see what's happening is this: * during the loading of the mcz ExternalFunction is defined, * a method is parsed (#XOpenDisplay, which has a pragma <cdecl: X11Display* ''XOpenDisplay'' (char*) module:''X11''>) * Parser >> externalFunctionDeclaration checks whether ExternalFunction is defined. * It is, so tries to evaluate `descriptorClass callingConvention: here` and boom, because ExternalFunction class >> callingConventionFor: _has not been loaded yet_. I've seen this kind of issue with Helvetia code: sometimes you simply have to load class-side methods first. Thoughts? Lukas worked around the issue with his Helvetia code by directly patching Monticello, and ripping out its "try to do atomic loading" mechanism. frank

12 39

Re: [squeak-dev] The Inbox: System-ul.528.mcz
by Frank Shearar 14 Sep '13

14 Sep '13

On 12 May 2013 21:35, <commits(a)source.squeak.org> wrote: > A new version of System was added to project The Inbox: > http://source.squeak.org/inbox/System-ul.528.mcz > > ==================== Summary ==================== > > Name: System-ul.528 > Author: ul > Time: 12 May 2013, 10:33:50.489 pm > UUID: d59743da-b5a5-4413-81f4-eb6dd7f7704d > Ancestors: System-fbs.527 > > Added support for defining the exit code when quitting. > > =============== Diff against System-fbs.527 =============== ORSM! So I just need to say SmalltalkImage current snapshot: false andQuitWithExitCode: 1 to signal a failure. Thanks! frank

2 5

BitBlt alpha blend rules broken?
by tim Rowledge 09 Sep '13

09 Sep '13

The guy that is hacking BitBlt for ARM for me is having all sorts of fun and making good progress. I even have a VM working with an early version and am currently in the process of testing the results. He's started on some of the more….curious… corners and right now is of the opinion that alphaBlend is getting its maths wrong. I'll just quote him rather than paraphrasing; > "The next combinationRule I've tackled is alphaBlend. This is clearly only > applicable to 32bpp source and destination images. Despite what the > comments in the original source say, this *does* calculate the alpha > component of the output, rather than zero it. In fact, it's using the > formula: > > da' = sa + (1-sa).da > dr' = sa.sr + (1-sa).dr > dg' = sa.sg + (1-sa).dg > db' = sa.sb + (1-sa).db > > where sa = source alpha, etc. > > This is actually incorrect maths for an alpha blend operation with non- > premultiplied colours, as appears to be the intent. For the RGB colour > components, the contribution from the first term is da' times too large, > and the contribution from the second term is da'/da times too large. In > other words, you only get the correct results if the destination image > was fully opaque (which implies that the result is likewise). > > Nevertheless, I suspect we'd run into problems if we attempted to fix > this, so I'll endeavour to replicate the behaviour, bugs and all. > > Next issue was the fact that for some crazy reason, they'd decided to > renormalise the colour components after the multiplication by doing a > divide by 255, rounding to +infinity. This is neither the "best" approach > (arguably round to nearest with rounding half-values to odd or even would > be the best), nor is it particularly efficient to implement - divisions > rarely are. The current code also has a completely pointless bitwise AND > with 0xFF following the division, since the inputs to the division cannot > exceed 0xFF*0xFF to begin with." We're talking about BitBltSimulation>alphaBlend:with: in this case. It's very old code. I'm not into colour blending maths etc and I can't see why we're using 255 instead of 256 (which would allow use of some ARM pixel engine instructions) nor quite what the intent is here. Thoughts, anyone? tim -- tim Rowledge; tim(a)rowledge.org; http://www.rowledge.org/tim Strange OpCodes: IAM: Increase Amperage Above Maximum

3 4

Phexample on Squeak
by Frank Shearar 21 Jul '13

21 Jul '13

Hi Niko, I tried Phexample in Squeak and things work great. I had to load in Regex-Core, but that's no biggie. I'll try my hand at editing the ConfigurationOf and send you a patch. I have only one blip: ForExampleStack >> #shouldFailWhenPopEmpty fails because instead of 'this collection is empty' the example fails with 'this stack is empty'. I'm not sure how to fix that, given that it's not your problem. One easy way to get a green bar would be to change the error message in Squeak (making it less accurate). Any thoughts? frank

3 9

Re: [squeak-dev] The Inbox: Files-cmm.125.mcz
by Levente Uzonyi 04 Jul '13

04 Jul '13

On Fri, 28 Jun 2013, commits(a)source.squeak.org wrote: > A new version of Files was added to project The Inbox: > http://source.squeak.org/inbox/Files-cmm.125.mcz > > ==================== Summary ==================== > > Name: Files-cmm.125 > Author: cmm > Time: 27 June 2013, 7:24:33.37 pm > UUID: 0b00b2c4-3430-442b-a924-72040ff43d73 > Ancestors: Files-fbs.124 > > FileDirectory>>statsForDirectoryTree: is a prime example of FileDirectory living up to its poor reputation. It does one useless thing, and does it inefficiently. It has no senders, deservedly. > Tree-walking operations like gathering stats can be written in apps more generally using directoryTreeDo:. I wouldn't say it's inefficient. It uses BFS (instead of DFS, which is used by #directoryTreeDo:entries:), which is somewhat better than DFS, because of the way FilePlugin works: you have to iterate over the contents of the whole directory if you want to it to be efficient. If you use DFS, then the contents of all parent directories up to the root will have to be stored in memory during the iteration, while if you use BFS, then only the sibling directories (without their contents) or the siblings's parents will be in memory. I made two "new" implementations of #statsForDirectoryTree:. This one optimizes memory usage for BFS: statsForDirectoryTree2: rootedPathName | dirs files bytes todo | dirs := files := bytes := 0. todo := OrderedCollection with: rootedPathName. [ todo isEmpty ] whileFalse: [ | p | p := todo removeFirst. self directoryContentsFor: p do: [ :entry | entry isDirectory ifTrue: [ todo addLast: p , self pathNameDelimiter asString , entry name. dirs := dirs + 1] ifFalse: [ files := files + 1. bytes := bytes + entry fileSize ] ] ]. ^{ dirs. files. bytes } This one uses #directoryTreeDo:: statsForDirectoryTree3: rootedPathName | dirs files bytes | dirs := files := bytes := 0. (FileDirectory on: rootedPathName) directoryTreeDo: [ :path | | entry | entry := path last. entry isDirectory ifTrue: [ dirs := dirs + 1 ] ifFalse: [ files := files + 1. bytes := bytes + entry fileSize ] ]. ^{ dirs. files. bytes } Let's see the numbers: "Make sure the disk won't be accessed" FileDirectory default statsForDirectoryTree: '/usr/'. FileDirectory default statsForDirectoryTree: '/usr/'. "Run the benchmark" #(statsForDirectoryTree: statsForDirectoryTree2: statsForDirectoryTree3:) collect: [ :each | each -> ((1 to: 5) collect: [ :run | Smalltalk garbageCollect. [ FileDirectory default perform: each with: '/usr/' ] timeToRun ]) ]. { #statsForDirectoryTree:->#(6340 6300 6430 6316 6228) . #statsForDirectoryTree2:->#(5918 6072 6122 6296 6306) . #statsForDirectoryTree3:->#(8576 8374 8470 8506 8228) } It's a bit noisy (i was too lazy to exit other programs), but the one using #directoryTreeDo: seems to be clearly slower than the existing algorithm using BFS. Levente > > =============== Diff against Files-fbs.124 =============== > > Item was changed: > + ----- Method: FileDirectory>>directoryTreeDo:entries: (in category 'private') ----- > - ----- Method: FileDirectory>>directoryTreeDo:entries: (in category 'enumeration') ----- > directoryTreeDo: oneArgBlock entries: entriesCollection > "Value oneArgBlock with the path (an OrderedCollection of FileDirectory's) to each DirectoryEntry and the DirectoryEntry itself." > self entries do: > [ : each | > entriesCollection add: each. > oneArgBlock value: entriesCollection. > each isDirectory ifTrue: > [ | subdir | > subdir := each asFileDirectory. > subdir > directoryTreeDo: oneArgBlock > entries: entriesCollection ]. > entriesCollection removeLast ]! > > Item was removed: > - ----- Method: FileDirectory>>statsForDirectoryTree: (in category 'enumeration') ----- > - statsForDirectoryTree: rootedPathName > - "Return the size statistics for the entire directory tree starting at the given root. The result is a three element array of the form: (<number of folders><number of files><total bytes in all files>). This method also serves as an example of how recursively enumerate a directory tree." > - "FileDirectory default statsForDirectoryTree: '\smalltalk'" > - > - | dirs files bytes todo entries p | > - dirs := files := bytes := 0. > - todo := OrderedCollection with: rootedPathName. > - [todo isEmpty] whileFalse: [ > - p := todo removeFirst. > - entries := self directoryContentsFor: p. > - entries do: [:entry | > - entry isDirectory > - ifTrue: [ > - todo addLast: p , self pathNameDelimiter asString , entry name. > - dirs := dirs + 1] > - ifFalse: [ > - files := files + 1. > - bytes := bytes + entry fileSize]]]. > - ^ Array with: dirs with: files with: bytes > - ! > > >

7 15

Someone update Installer in 4.3 and 4.4 please?
by Frank Shearar 01 Jul '13

01 Jul '13

I'm away from my computer for the next few hours or I'd do this myself: with the Gemstone -> GemTalk Systems rename, we need to change the URL in Installer class >> #squeakSource3 (or whatever #ss3 uses) to point to the new URL. This will allow correctly written Installer scripts (i.e., ones that use #ss3 and not #url:) to carry on working. Dale Henrichs has found that Squeak MC HTTP repositories don't like 3xx responses, and so a number of his builds fail in Squeak. Pharo uses ZnHTTP, which handles HTTP redirection correctly. Thanks! frank

3 3

Daily Commit Log
by commits＠source.squeak.org 01 Jul '13

01 Jul '13

Changes to Trunk (http://source.squeak.org/trunk.html) in the last 24 hours: http://lists.squeakfoundation.org/pipermail/packages/2013-June/006095.html Name: System-fbs.555 Ancestors: System-fbs.554 More consolidation of the "System-Object Events" pieces. Note the removal of indirection through Object class >> #flushEvents. At some point we might want to deprecate this method because it doesn't do anything useful. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2013-June/006096.html Name: Kernel-fbs.778 Ancestors: Kernel-fbs.777 More consolidation of the "System-Object Events" pieces. Note the removal of indirection through Object class >> #flushEvents. At some point we might want to deprecate this method because it doesn't do anything useful. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2013-June/006097.html Name: Multilingual-fbs.162 Ancestors: Multilingual-nice.161 TrueType shouldn't depend on Morphic (because Morphic should be unloadable), and since it's display related let's put it in Multilingual-Display (for the moment). ============================================= http://lists.squeakfoundation.org/pipermail/packages/2013-June/006098.html Name: TrueType-fbs.20 Ancestors: TrueType-ul.19 TrueType shouldn't depend on Morphic (because Morphic should be unloadable), and since it's display related let's put it in Multilingual-Display (for the moment). ============================================= http://lists.squeakfoundation.org/pipermail/packages/2013-June/006099.html Name: Tests-fbs.228 Ancestors: Tests-fbs.227 TrueType -/-> Morphic. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2013-June/006100.html Name: TrueType-fbs.21 Ancestors: TrueType-fbs.20 System-Tools is a better place for TTFileDescription>>profileAll because System-Tools provides reasonably generic 'tools' things that don't otherwise have anything to do with TrueType. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2013-June/006101.html Name: System-fbs.556 Ancestors: System-fbs.555 System-Tools is a better place for TTFileDescription>>profileAll because System-Tools provides reasonably generic 'tools' things that don't otherwise have anything to do with TrueType. =============================================

1 0

Jump to page: