Squeak-dev August 2022

squeak-dev@lists.squeakfoundation.org

30 participants
156 discussions

The Trunk: Graphics-kfr.530.mcz
by commits＠source.squeak.org 10 Aug '22

10 Aug '22

Marcel Taeumel uploaded a new version of Graphics to project The Trunk: http://source.squeak.org/trunk/Graphics-kfr.530.mcz ==================== Summary ==================== Name: Graphics-kfr.530 Author: kfr Time: 9 August 2022, 7:12:20.983085 am UUID: ee4a3ab4-fc38-f349-b0bc-f6f109ba9d01 Ancestors: Graphics-kfr.529 Avoid call to allButFirstDo: in Rectangle class>>encompassing: and Rectangle class>>merging:. But keep the call to Rectangle class>>origin: corner: so subclasses can use these methods =============== Diff against Graphics-kfr.527 =============== Item was added: + ----- Method: Form>>at:bias: (in category 'processing') ----- + at: aPoint bias: aQuantError + | x y color point oldColor | + x := aPoint x. + y := aPoint y. + + point := (x + 1 min: self width)@ y. + oldColor := self colorAt: point. + color := oldColor + (aQuantError * 0.4375). "7/16" + self colorAt: point put: color. + + point := (x - 1 max: 0)@ (y + 1 min: self height). + oldColor := self colorAt: point. + color := oldColor + (aQuantError * 0.1875)."3/16" + self colorAt: point put: color. + + point := x @ (y + 1 min: self height). + oldColor := self colorAt: point. + color := oldColor + (aQuantError * 0.3125 ) ."5/16" + self colorAt: point put: color. + + point := (x + 1 min: self width) @ (y + 1 min: self height). + oldColor := self colorAt: point. + color := oldColor + (aQuantError * 0.0625 ) ."1/16" + self colorAt: point put: color. + ! Item was added: + ----- Method: Form>>dither: (in category 'processing') ----- + dither: aDepth + | newColor oldColor quantError result | + "Floyd Steinberg dithering for depth 1 to 8 bits" + result := self copy. + (0 to: height) do:[:y| + (0 to: width) do:[:x| + oldColor := (result colorAt: x@y). + newColor := Color colorFromPixelValue: (oldColor pixelValueForDepth: aDepth) depth: aDepth. + result colorAt: x@y put: newColor. + quantError := (oldColor - newColor). + result at: x@y bias: quantError]]. + ^result.! Item was changed: ----- Method: Form>>processUsingKernel:factor:bias: (in category 'processing') ----- processUsingKernel: filter factor: factor bias: bias + | image result color | - | image result | image := self. result := Form extent: image extent depth: image depth. 0 to: image height - 1 do: [:y | 0 to: image width - 1 do: [:x | | r g b | r := g := b := 0.0. 0 to: filter rowCount - 1 do: [:filterY | 0 to: filter columnCount - 1 do: [:filterX | | imageX imageY | imageX := (x - (filter columnCount // 2) + filterX + image width) \\ image width. imageY := (y - (filter rowCount // 2) + filterY + image height) \\ image height. + color := image colorAt: imageX@imageY. + r := r + (color red * - r := r + ((image colorAt: imageX@imageY) red * (filter at: filterY + 1 at: filterX + 1)). + g := g + (color green * - g := g + ((image colorAt: imageX@imageY) green * (filter at: filterY + 1 at: filterX + 1)). + b := b + (color blue * - b := b + ((image colorAt: imageX@imageY) blue * (filter at: filterY + 1 at: filterX + 1))]]. result colorAt: x@y put: (Color r: ((factor * r + bias) min: 1.0 max: 0.0) g: ((factor * g + bias) min: 1.0 max: 0.0) b: ((factor * b + bias) min: 1.0 max: 0.0))]]. ^ result ! Item was changed: ----- Method: Rectangle class>>encompassing: (in category 'instance creation') ----- encompassing: listOfPoints "A number of callers of encompass: should use this method." | topLeft bottomRight | + topLeft := bottomRight := nil. + listOfPoints do: + [:p | topLeft == nil + ifTrue: [topLeft := bottomRight := p] + ifFalse: [topLeft := topLeft min: p. + bottomRight := bottomRight max: p]]. + ^self origin: topLeft corner: bottomRight! - topLeft := bottomRight := listOfPoints first. - listOfPoints allButFirstDo: - [:p |topLeft := topLeft min: p. - bottomRight := bottomRight max: p]. - ^self origin: topLeft corner: bottomRight - ! Item was changed: ----- Method: Rectangle class>>merging: (in category 'instance creation') ----- merging: listOfRects "A number of callers of merge: should use this method." + | minX minY maxX maxY | - | bottomRight topLeft | - topLeft := listOfRects first topLeft. - bottomRight := listOfRects first bottomRight. listOfRects + do: [:r | minX + ifNil: [minX := r topLeft x. minY := r topLeft y. + maxX := r bottomRight x. maxY := r bottomRight y] + ifNotNil: [minX := minX min: r topLeft x. minY := minY min: r topLeft y. + maxX := maxX max: r bottomRight x. maxY := maxY max: r bottomRight y]]. + ^self origin: minX@minY corner: maxX@maxY! - allButFirstDo: [:r | topLeft := topLeft min: r topLeft. - bottomRight := bottomRight max: r bottomRight]. - ^self origin: topLeft corner: bottomRight. - !

1 0

The Trunk: Graphics-mt.531.mcz
by commits＠source.squeak.org 10 Aug '22

10 Aug '22

Marcel Taeumel uploaded a new version of Graphics to project The Trunk: http://source.squeak.org/trunk/Graphics-mt.531.mcz ==================== Summary ==================== Name: Graphics-mt.531 Author: mt Time: 10 August 2022, 10:15:13.741587 am UUID: d7b78221-c876-6a43-8455-dba637d950dc Ancestors: Graphics-eem.529, Graphics-kfr.530 Merge kfr.528, kfr.529, kfr.530 from inbox. Graphics-kfr.530: Avoid call to allButFirstDo: in Rectangle class>>encompassing: and Rectangle class>>merging:. But keep the call to Rectangle class>>origin: corner: so subclasses can use these methods Graphics-kfr.529: Floyd Steinberg dither for forms of 1 to 8 bit depth. Graphics-kfr.528: Color>>at: is slow. Speed up by using variable =============== Diff against Graphics-eem.529 =============== Item was added: + ----- Method: Form>>at:bias: (in category 'processing') ----- + at: aPoint bias: aQuantError + | x y color point oldColor | + x := aPoint x. + y := aPoint y. + + point := (x + 1 min: self width)@ y. + oldColor := self colorAt: point. + color := oldColor + (aQuantError * 0.4375). "7/16" + self colorAt: point put: color. + + point := (x - 1 max: 0)@ (y + 1 min: self height). + oldColor := self colorAt: point. + color := oldColor + (aQuantError * 0.1875)."3/16" + self colorAt: point put: color. + + point := x @ (y + 1 min: self height). + oldColor := self colorAt: point. + color := oldColor + (aQuantError * 0.3125 ) ."5/16" + self colorAt: point put: color. + + point := (x + 1 min: self width) @ (y + 1 min: self height). + oldColor := self colorAt: point. + color := oldColor + (aQuantError * 0.0625 ) ."1/16" + self colorAt: point put: color. + ! Item was added: + ----- Method: Form>>dither: (in category 'processing') ----- + dither: aDepth + | newColor oldColor quantError result | + "Floyd Steinberg dithering for depth 1 to 8 bits" + result := self copy. + (0 to: height) do:[:y| + (0 to: width) do:[:x| + oldColor := (result colorAt: x@y). + newColor := Color colorFromPixelValue: (oldColor pixelValueForDepth: aDepth) depth: aDepth. + result colorAt: x@y put: newColor. + quantError := (oldColor - newColor). + result at: x@y bias: quantError]]. + ^result.! Item was changed: ----- Method: Form>>processUsingKernel:factor:bias: (in category 'processing') ----- processUsingKernel: filter factor: factor bias: bias + | image result color | - | image result | image := self. result := Form extent: image extent depth: image depth. 0 to: image height - 1 do: [:y | 0 to: image width - 1 do: [:x | | r g b | r := g := b := 0.0. 0 to: filter rowCount - 1 do: [:filterY | 0 to: filter columnCount - 1 do: [:filterX | | imageX imageY | imageX := (x - (filter columnCount // 2) + filterX + image width) \\ image width. imageY := (y - (filter rowCount // 2) + filterY + image height) \\ image height. + color := image colorAt: imageX@imageY. + r := r + (color red * - r := r + ((image colorAt: imageX@imageY) red * (filter at: filterY + 1 at: filterX + 1)). + g := g + (color green * - g := g + ((image colorAt: imageX@imageY) green * (filter at: filterY + 1 at: filterX + 1)). + b := b + (color blue * - b := b + ((image colorAt: imageX@imageY) blue * (filter at: filterY + 1 at: filterX + 1))]]. result colorAt: x@y put: (Color r: ((factor * r + bias) min: 1.0 max: 0.0) g: ((factor * g + bias) min: 1.0 max: 0.0) b: ((factor * b + bias) min: 1.0 max: 0.0))]]. ^ result ! Item was changed: ----- Method: Rectangle class>>encompassing: (in category 'instance creation') ----- encompassing: listOfPoints "A number of callers of encompass: should use this method." | topLeft bottomRight | + topLeft := bottomRight := nil. + listOfPoints do: + [:p | topLeft == nil + ifTrue: [topLeft := bottomRight := p] + ifFalse: [topLeft := topLeft min: p. + bottomRight := bottomRight max: p]]. + ^self origin: topLeft corner: bottomRight! - topLeft := bottomRight := listOfPoints first. - listOfPoints allButFirstDo: - [:p |topLeft := topLeft min: p. - bottomRight := bottomRight max: p]. - ^self origin: topLeft corner: bottomRight - ! Item was changed: ----- Method: Rectangle class>>merging: (in category 'instance creation') ----- merging: listOfRects "A number of callers of merge: should use this method." + | minX minY maxX maxY | - | bottomRight topLeft | - topLeft := listOfRects first topLeft. - bottomRight := listOfRects first bottomRight. listOfRects + do: [:r | minX + ifNil: [minX := r topLeft x. minY := r topLeft y. + maxX := r bottomRight x. maxY := r bottomRight y] + ifNotNil: [minX := minX min: r topLeft x. minY := minY min: r topLeft y. + maxX := maxX max: r bottomRight x. maxY := maxY max: r bottomRight y]]. + ^self origin: minX@minY corner: maxX@maxY! - allButFirstDo: [:r | topLeft := topLeft min: r topLeft. - bottomRight := bottomRight max: r bottomRight]. - ^self origin: topLeft corner: bottomRight. - !

1 0

Rectangle #encompassing: & #merging:
by Stéphane Rollandin 10 Aug '22

10 Aug '22

A minor hurdle in my effort to port my code to 6.0: The new implementations of #encompassing: and #merging: in Rectangle class use #allButFirstDo: on their argument, and that makes the methods unable to accept a Set as argument - which was fine before. Stef

4 3

The Trunk: Sound-eem.92.mcz
by commits＠source.squeak.org 10 Aug '22

10 Aug '22

Eliot Miranda uploaded a new version of Sound to project The Trunk: http://source.squeak.org/trunk/Sound-eem.92.mcz ==================== Summary ==================== Name: Sound-eem.92 Author: eem Time: 9 August 2022, 7:08:53.016079 pm UUID: 6e13b207-ddca-4f1c-854d-8b6124f946f0 Ancestors: Sound-eem.91 fix a typo =============== Diff against Sound-eem.91 =============== Item was changed: ----- Method: SoundPlayer class>>soundQuickStart (in category 'preferences') ----- soundQuickStart <preference: 'Quickstart Sounds' category: 'media' + description: 'If true, attempt to start playing sounds immediately rather than waiting for the next chance to fill the sound buffers"' - description: 'If true, attempt to start playing sounds immediately rather than waiting for the next chance to fill the sound bufffers"' type: #Boolean> ^SoundsShouldStartQuick ifNil: [self defaultQuickStartForPlatform]!

1 0

decomposing a MixedSound into left and right channels
by Eliot Miranda 10 Aug '22

10 Aug '22

Hi Stéphane, Hi All, I'm working on AAC encode/decode and mp4 packaging. I'm trying to test using the stereoBachFugue, but I can't even get the following to generate any sound in stereo. What am I doing wrong? testSqueakStereoSound | aMixedSound left right samples | (aMixedSound := MixedSound new) add: (AbstractSound bachFugueVoice1On: FMSound oboe1) pan: 0.4; add: (AbstractSound bachFugueVoice2On: FMSound organ1) pan: 0.1; add: (AbstractSound bachFugueVoice3On: PluckedSound default) pan: 0.9; add: (AbstractSound bachFugueVoice4On: FMSound brass1) pan: 0.6. left := WriteStream on: SoundBuffer new. right := WriteStream on: SoundBuffer new. samples := aMixedSound computeSamplesForSeconds: aMixedSound duration. [1 to: samples size by: 2 do: [:i| left nextPut: (samples at: i). right nextPut: (samples at: i + 1)]] value. "to step over quickly in the debugger" ^MixedSound new add: (SampledSound samples: left contents samplingRate: aMixedSound samplingRate) pan: 0; add: (SampledSound samples: right contents samplingRate: aMixedSound samplingRate) pan: 1; play "self testSqueakStereoSound" The idea is to pan the harpsichord and organ almost hard right and left, and pan the oboe and tuba a little around the centre, generate the stereo samples via computeSamplesForSeconds:, decompose the two channels into two separate monophonic channels, and recombine them. But I on;y ever get silence. I can see that the decomposition looks correct, but why do neither the resulting MixedSound, nor either the left or right channel produce any sound? _,,,^..^,,,_ best, Eliot

1 0

Unable to change font style or emphasis from halo in scrolled text
by Stéphane Rollandin 08 Aug '22

08 Aug '22

Hello, In 6.0, if you type a long text in a workspace, then select a word at the bottom and attempt to set it in bold emphasis, you will see that the halo to do so is not accessible anymore. To be precise: once the cursor is on the selected text - the first alt-click brings the Workspace halo - the second one yield the PluggableTextMorphPlus halo - the third one yield the TextMorphForEditView halo It is the later halo that is scrolled with the text display, making it out of sight if we scroll too much. Stef

2 5

A puzzle
by Stéphane Rollandin 08 Aug '22

08 Aug '22

Does anyone know why, in 6.0, evaluating Player subclass: #PlayerSubclass instanceVariableNames: '' classVariableNames: '' poolDictionaries: '' category: 'SomeCategory' results in PlayerSubclass actually having a nil category? (already happening in 5.3 BTW) Stef

2 1

The Trunk: Morphic-mt.2022.mcz
by commits＠source.squeak.org 08 Aug '22

08 Aug '22

Marcel Taeumel uploaded a new version of Morphic to project The Trunk: http://source.squeak.org/trunk/Morphic-mt.2022.mcz ==================== Summary ==================== Name: Morphic-mt.2022 Author: mt Time: 8 August 2022, 12:09:48.81904 pm UUID: 00f8498c-9d4e-794a-a5bb-56f9322b1b4d Ancestors: Morphic-eem.2021 Fixes halo box for transformed targets. Always show the box of the rendered morph, not the renderer. I will backport this to Squeak 6.0. Thanks to Stef (spfa) for the tip! =============== Diff against Morphic-eem.2021 =============== Item was changed: ----- Method: HaloMorph>>addHandlesForWorldHalos (in category 'private') ----- addHandlesForWorldHalos "Add handles for world halos, like the man said" | box w | w := self world ifNil:[target world]. self removeAllMorphs. "remove old handles, if any" self bounds: target bounds. box := w bounds insetBy: self handleSize // 2. target addWorldHandlesTo: self box: box. Preferences uniqueNamesInHalos ifTrue: [innerTarget assureExternalName]. self + addNameBeneath: box - addNameBeneath: ((box insetBy: (0@0 corner: 0@10)) - scaleBy: RealEstateAgent scaleFactor) string: innerTarget externalName. growingOrRotating := false. self layoutChanged. self changed. ! Item was changed: ----- Method: HaloMorph>>basicBox (in category 'private') ----- basicBox | minSide outset anExtent aBox w | minSide := 4 * self handleSize. + outset := (8 * RealEstateAgent scaleFactor) truncated. - outset := 8 * RealEstateAgent scaleFactor. anExtent := (self extent + self handleSize + outset) max: minSide asPoint. aBox := Rectangle center: self center extent: anExtent. w := self world ifNil:[target outermostWorldMorph]. ^ w ifNil: [aBox] ifNotNil: [aBox intersect: (w viewBox insetBy: self handleSize // 2)] ! Item was changed: ----- Method: SimpleHaloMorph>>updateBounds (in category 'updating') ----- updateBounds + self bounds: (self haloBoundsFor: self target renderedMorph).! - self bounds: (self haloBoundsFor: self target).!

1 0

The Inbox: Monticello-jr.783.mcz
by commits＠source.squeak.org 08 Aug '22

08 Aug '22

A new version of Monticello was added to project The Inbox: http://source.squeak.org/inbox/Monticello-jr.783.mcz ==================== Summary ==================== Name: Monticello-jr.783 Author: jr Time: 8 August 2022, 12:06:12.526085 am UUID: 01da1063-034e-5444-be24-853128075cad Ancestors: Monticello-mt.782 Fix: when writing only modified definitions to a file tree repository, writing the MCTraitDefinition of a trait that uses another trait without also writing the MCClassTraitDefinition would trim away the word "classTrait" from the value of the "classtraitcomposition" property in the properties.json file. Compare with MCFileTreeStCypressWriter>>#writeTraitDefinition: Usually, when introducing trait T to a class C (or another trait), the properties C become { "classtraitcomposition" : "T classTrait", "traitcomposition" : "T", ... }. Through the bug, this can be inadvertently changed to "classtraitcomposition" : "T". When the package is subsequently loaded from the files, instance methods of T will additionally appear on the class side of C. Probably the T classTrait methods will also be missing from C class, but I did not verify this. When using plain file tree repositories, the bug does not appear because all files are removed and written again using all the definitions, including the MCClassTraitDefinition. Squot tries to optimize commit performance by passing only the changes through to the file tree writer. The situation before this patch requires to also collect and pass along the MCClassTraitDefinitions even if these were not changed. MCTraitDefinition should not override classTraitCompositionString and it should not conflate it with the traitCompositionString. Just use the method inherited from MCClassDefinition. I checked in Pharo 10: the override of classTraitCompositionString does not exist there either. =============== Diff against Monticello-mt.782 =============== Item was removed: - ----- Method: MCTraitDefinition>>classTraitCompositionString (in category 'accessing') ----- - classTraitCompositionString - ^self traitComposition ifNil: ['{}'].!

1 0

Unable to install a specific TrueType font in 6.0
by Stéphane Rollandin 07 Aug '22

07 Aug '22

Hello, Attached is a TrueType font I use in my game. In a 5.3 image, I can load it fine. For example from a file list, using 'install ttf style' works. In a 6.0 image, 'install TrueType font' seems to work, but I have no glyphs for most characters and the font is unusable. I kind of need this specific font if I want to port my game to 6.0 in its current state, but am at a loss about what to do. Any help appreciated... Stef

3 4

Jump to page: