Squeak-dev April 2016

squeak-dev@lists.squeakfoundation.org

39 participants
244 discussions

New Cog VMs available
by Eliot Miranda 23 Apr '16

23 Apr '16

... at http://www.mirandabanda.org/files/Cog/VM/VM.r3684 CogVM binaries as per VMMaker.oscog-eem.1834/r3684 General: Correct undue sign extension wile promoting 32 to 64 bits int in fetchLong64:ofObject: This was causing trouble in non spur object memory (V3). Fix slowdown of update on 64-bit x64 by using int variables for the tides in sqExternalSemaphores.c and inline assembler for the sqCompareAndSwap in sqAtomicOps.h. Fix slips in offset time primitives. They need to pop argumentCount + 1, not argumentCount. Define an optional primitive as the fast primitive fail code so that unselected optional primitives are in the prmitiveTable as fast primitive fails. Spur: Have WideString>>at: fail to answer an out-of-range character in the interpreter primitive (as well as in the JIT). Fix bug in following state on primitive failure. The old code would always follow to depth 1, even if the accessor depth was 0. Hard-code primitiveSize's depth to 0 (accessing length in at:[put:] and size causes the stack depth computation to answer 1 instead of 0 for these primitives. Fix assert and dequeueMourner for case where mournQueue is nil. - fixed a bug in receiver accessing in immutability primitive for mirror primitive. - change primitiveSetOrHasIdentityHash to patch the class table if the new hash is set to a behavior (the primitive knows it's a behavior if the second (optional) argument is true) For example: FullBlockClosure tryPrimitive: 161 withArgs: {38.true} No longer follow the method and context fields in a closure in activateNewClosureMethod:numArgs:mayContextSwitch:; the caller will have failed if these are forwarders, so no need to check again. Cogit: Add a primitive that answers pc map data for methods which can be used to better decorate methods in the VM Profiler. Refactor the pc map enumeration facilities so that the Sista pic data primitive can share the same enumerator. Do this by collapsing the isBackwardBranch and annotation parameters into a single parameter. Follow selectors in the openPICList post Spur become. Fix the ARM's caller-saved register mask now that we can name all the actual registers. Reworked machine code generation of immutability so for common stores it uses a single trampoline for both store checks and immutability checks. - improved support for register allocation: branch merge successfully compiled with register moved instead of spilling. Sista Cogit: Don't bother to add counters to conditional jumps implementing and: and or:. Added the remote inst var access bytecode in sista V1 bytecode set without interfering with existing code. Plugins: Upgrade LargeIntegersPlugin to v2.0 LargeInteger primitives now deal with 32-bits digits. No change to image code. Memory is 8 bytes aligned on Spur. When storing 32/64 bits large integers values, allways fill the eight bytes whatever the effectivily used size, rather than bother with dissertion of size. Generate integer type checking as C macros rather than direct/indirect interpreterProxy function call in plugins. This, and 32-bit accessing mean singificantly faster large integer arithmetic. Fix primAlienReplace to use positiveMachineIntegerValueOf: instead of positive32BitValueOf:. BitBltPlugin operates on 32bit word units, therefore it's better to declare its operands as 'unsigned int' rather than sqInt. On 32bit VM, this doesn't change anything, but on 64bits spur, it makes this snippet work: | wideString source pos blt expectedWideString | source := #[1 64 255 14 1 64 48 251]. expectedWideString := WideString fromByteArray: source. wideString := WideString new: source size // 4. pos := 0. blt := (BitBlt toForm: (Form new hackBits: wideString)) sourceForm: (Form new hackBits: source). blt combinationRule: Form over; sourceX: 0; sourceY: pos // 4; height: wideString byteSize // 4; width: 4; destX: 0; destY: 0; copyBits. wideString restoreEndianness. self assert: wideString = expectedWideString Hence it fixes loading/diffing MCZ with wide character. Mac OS Builds: Don't link with -fvisibility=hidden; it breaks external plugins. Use -fvisibility=default instead. This fixes e.g. UnixOSProcessPlugin on Mac OS X. Windows Builds: The Windows VMs are no longer "dpiAware". If you want one-to-one pixel mapping check the README for ins ructions. It's a simple edit. _,,,^..^,,,_ best, Eliot

1 0

Who breaks it, fixes it -- A Question about our Development Model
by marcel.taeumel 21 Apr '16

21 Apr '16

Hey Squeak Community, according to our development model here: https://squeakboard.wordpress.com/2009/07/02/a-new-community-development-mo… If I break something, I am responsible for fixing it. So, if I meant to fix a bug such as here: http://forum.world.st/The-Trunk-Morphic-mt-1119-mcz-td4890816.html And somebody raises an issue about new bugs added or some use case missed, is there a general need to revert that before further disscussions? More importantly, wouldn't I be in charge of reverting it? And not somebody else... Note that we all assume that any trunk committer thoroughly investigates the changes made and fixes yet discusses obvious issues --- before committing such changes. This fundamental level of trust comes with being allowed to commit to trunk in the first place. Nevertheless, remaining issues can easily discussed on the mailing list and the responsible committer can fix/revert them right away -- or delegate it to someone else if time is rare. At least, this is what I expect from every trunk committer. :-) It seems* to work pretty well: http://forum.world.st/The-Trunk-ToolBuilder-Morphic-kfr-159-mcz-td4879754.h… http://forum.world.st/The-Trunk-Collections-mt-688-mcz-td4889675.html http://forum.world.st/The-Trunk-Monticello-mt-629-mcz-td4887731.html http://forum.world.st/The-Trunk-Morphic-cmm-615-mcz-td4524250.html http://forum.world.st/The-Trunk-Graphics-tfel-327-mcz-td4880035.html Sometimes, it seems* to not work: http://forum.world.st/The-Trunk-Morphic-cmm-1052-mcz-td4862408.htm http://forum.world.st/The-Trunk-SMLoader-cmm-85-mcz-td4850418.html http://forum.world.st/The-Trunk-Morphic-cmm-1120-mcz-td4890853.html http://forum.world.st/The-Trunk-Tools-cmm-693-mcz-td4890497.html What is your opinion? Best, Marcel * No limit or warranty. Off-list discussions not considered. Just exemplified. -- View this message in context: http://forum.world.st/Who-breaks-it-fixes-it-A-Question-about-our-Developme… Sent from the Squeak - Dev mailing list archive at Nabble.com.

5 4

The Trunk: Tools-bf.694.mcz
by commits＠source.squeak.org 20 Apr '16

20 Apr '16

Bert Freudenberg uploaded a new version of Tools to project The Trunk: http://source.squeak.org/trunk/Tools-bf.694.mcz ==================== Summary ==================== Name: Tools-bf.694 Author: bf Time: 20 April 2016, 6:24:57.82409 pm UUID: b13d8848-e0bb-479f-b9b5-a243e111cea2 Ancestors: Tools-cmm.693 Always accept dropped variable references in a workspace. =============== Diff against Tools-cmm.693 =============== Item was changed: ----- Method: Workspace>>acceptDroppingMorph:event:inMorph: (in category 'drag and drop') ----- acceptDroppingMorph: dropee event: evt inMorph: targetMorph "Return the dropee to its old position, and add a reference to it at the cursor point." | bindingName externalName reference | (dropee isKindOf: TransferMorph) ifTrue: [reference := dropee passenger. externalName := dropee passenger className] ifFalse: [reference := dropee. dropee externalName]. externalName := externalName isOctetString ifTrue: [externalName] ifFalse: ['a' , externalName]. bindingName := externalName withFirstCharacterDownshifted , reference identityHash printString. targetMorph correctSelectionWithString: bindingName , ' '. (self bindingOf: bindingName) value: reference. + (dropee isKindOf: TransferMorph) + ifFalse: [dropee rejectDropMorphEvent: evt]. - dropee rejectDropMorphEvent: evt. ^ true"success"! Item was changed: ----- Method: Workspace>>wantsDroppedMorph:event:inMorph: (in category 'drag and drop') ----- wantsDroppedMorph: dropee event: evt inMorph: target + ^ acceptDroppedMorphs or: [dropee isKindOf: TransferMorph] - ^ acceptDroppedMorphs !

1 0

The Trunk: Morphic-bf.1122.mcz
by commits＠source.squeak.org 20 Apr '16

20 Apr '16

Bert Freudenberg uploaded a new version of Morphic to project The Trunk: http://source.squeak.org/trunk/Morphic-bf.1122.mcz ==================== Summary ==================== Name: Morphic-bf.1122 Author: bf Time: 20 April 2016, 6:09:34.906493 pm UUID: a3dac7ab-cefe-4c77-9a8f-2d45f7f0dd9e Ancestors: Morphic-tfel.1121 Prevent DNU when keystroke event causes text editor to get closed (owner is nil) =============== Diff against Morphic-tfel.1121 =============== Item was changed: ----- Method: StringMorphEditor>>keyStroke: (in category 'event handling') ----- keyStroke: evt "This is hugely inefficient, but it seems to work, and it's unlikely it will ever need to be any more efficient -- it's only intended to edit single-line strings." | char priorEditor newSel | (((char := evt keyCharacter) = Character enter) or: [(char = Character cr) or: [char = $s and: [evt commandKeyPressed]]]) ifTrue: [owner doneWithEdits; acceptContents. self flag: #arNote. "Probably unnecessary" evt hand releaseKeyboardFocus. ^ self delete]. (char = $l and: [evt commandKeyPressed]) ifTrue: "cancel" [owner cancelEdits. evt hand releaseKeyboardFocus. ^ self delete]. super keyStroke: evt. + + owner ifNil: [^self]. owner interimContents: self contents asString. newSel := self editor selectionInterval. priorEditor := self editor. "Save editor state" self releaseParagraph. "Release paragraph so it will grow with selection." self paragraph. "Re-instantiate to set new bounds" self installEditorToReplace: priorEditor. "restore editor state" self editor selectFrom: newSel first to: newSel last. !

1 0

The Trunk: MorphicExtras-tfel.172.mcz
by commits＠source.squeak.org 20 Apr '16

20 Apr '16

Tim Felgentreff uploaded a new version of MorphicExtras to project The Trunk: http://source.squeak.org/trunk/MorphicExtras-tfel.172.mcz ==================== Summary ==================== Name: MorphicExtras-tfel.172 Author: tfel Time: 20 April 2016, 2:25:18.788021 pm UUID: cd9fb2c0-84a9-ee4c-a244-98522b2ac822 Ancestors: MorphicExtras-mt.171 make postscript printing work again with current default fonts =============== Diff against MorphicExtras-mt.171 =============== Item was added: + ----- Method: PostscriptCanvas class>>fontsForDejaVuSans (in category 'font mapping') ----- + fontsForDejaVuSans + + | d | + + "Bold = 1, Ital = 2, Under = 4, Narrow = 8, Struckout = 16" + d := Dictionary new. + d + at: 0 put: #('Helvetica-Bold' 1.0); + at: 1 put: #('Helvetica-Bold' 1.0); + at: 2 put: #('Helvetica-Oblique' 1.0); + at: 3 put: #('Helvetica-BoldOblique' 1.0). + ^d! Item was changed: ----- Method: PostscriptCanvas class>>initializeFontMap (in category 'font mapping') ----- initializeFontMap "Initialize the dictionary mapping font names to substitutions for Postscript code generation." "PostscriptCanvas initializeFontMap" | f | FontMap := Dictionary new. FontMap at: 'NewYork' put: (f := self fontsForNewYork); at: 'Accuny' put: f; at: 'Helvetica' put: (f := self fontsForHelvetica); at: 'Accujen' put: f; at: 'Palatino' put: self fontsForPalatino; at: 'ComicBold' put: (f := self fontsForComicBold); at: 'Accuat' put: self fontsForAccuAt; + at: 'Bitmap DejaVu Sans' put: self fontsForDejaVuSans; + at: 'ComicPlain' put: self fontsForComicPlain! Item was changed: ----- Method: PostscriptCanvas class>>postscriptFontInfoForFont: (in category 'font mapping') ----- postscriptFontInfoForFont: font | decoded decodedName keys match fontName | fontName := font textStyleName asString. decoded := TextStyle decodeStyleName: fontName. decodedName := decoded second. keys := self fontMap keys asArray sort: [ :a :b | a size > b size ]. - match := keys select: [ :k | decoded first = k or: [ fontName = k ] ]. match do: [ :key | | subD desired mask | subD := self fontMap at: key. desired := font emphasis. mask := 31. [ desired := desired bitAnd: mask. subD at: desired ifPresent: [ :answer | ^answer]. mask := mask bitShift: -1. desired > 0 ] whileTrue. ]. "No explicit lookup found; try to convert the style name into the canonical Postscript name. This name will probably still be wrong." fontName := String streamContents: [ :s | s nextPutAll: decodedName. decoded third do: [ :nm | s nextPut: $-; nextPutAll: nm ]. (font emphasis = 0 and: [ (decoded last includes: 0) not ]) ifTrue: [ s nextPutAll: '-Regular' ]. (font emphasis = 1 and: [ (decoded first anyMask: 1) not ]) ifTrue: [ s nextPutAll: '-Bold' ]. (font emphasis = 2 and: [ (decoded first anyMask: 2) not ]) ifTrue: [ s nextPutAll: '-Italic' ]. (font emphasis = 3 and: [ (decoded first anyMask: 3) not ]) ifTrue: [ s nextPutAll: '-BoldItalic' ]. ]. + ^ {'(', fontName, ') cvn'. 1.0} - ^ {fontName. 1.0} ! Item was changed: ----- Method: PostscriptCanvas>>image:at:sourceRect:rule: (in category 'private') ----- + image: form at: aPoint sourceRect: sourceRect rule: rule + | aForm | - image: aForm at: aPoint sourceRect: sourceRect rule: rule self preserveStateDuring: [:inner | inner translate: aPoint + self origin. + aForm := form depth <= 8 ifTrue: [form asFormOfDepth: 32] ifFalse: [form]. + target write: ((aForm colorsUsed includes: Color transparent) + ifTrue: [| top f2 c2 offset | + "tfel: This was taken from SketchMorph, but is actually needed for all + forms that use transparency" + offset := currentTransformation ifNil: [0@0] ifNotNilDo: [:t | t offset]. + top := self topLevelMorph. + f2 := Form extent: aForm extent depth: self depth. + c2 := f2 getCanvas. + c2 fillColor: Color white. + c2 + translateBy: offset - self origin - aPoint + clippingTo: f2 boundingBox + during: [:c | top fullDrawOn: c]. + f2] + ifFalse: [aForm])]. - target write: aForm] ! Item was removed: - ----- Method: SketchMorph>>drawPostscriptOn: (in category '*MorphicExtras-*morphic-Postscript Canvases') ----- - drawPostscriptOn: aCanvas - - | top f2 c2 tfx clrs | - - tfx := self transformFrom: self world. - tfx angle = 0.0 ifFalse: [^super drawPostscriptOn: aCanvas]. "can't do rotated yet" - clrs := self rotatedForm colorsUsed. - (clrs includes: Color transparent) - ifFalse: [^super drawPostscriptOn: aCanvas]. "no need for this, then" - - "Smalltalk at: #Q put: OrderedCollection new" - "Q add: {self. tfx. clrs}." - "(self hasProperty: #BOB) ifTrue: [self halt]." - - top := aCanvas topLevelMorph. - f2 := Form extent: self extent depth: self rotatedForm depth. - c2 := f2 getCanvas. - c2 fillColor: Color white. - c2 translateBy: bounds origin negated clippingTo: f2 boundingBox during: [ :c | - top fullDrawOn: c - ]. - aCanvas paintImage: f2 at: bounds origin - - !

1 0

ftp down?
by Max Leske 20 Apr '16

20 Apr '16

Hi I’m trying to get the Fuel build set up for 5.0 but I can’t seem to connect to ftp.squeak.org. It’s available via http but not via ftp. Alternatively, how do you access that server for the Squeak builds? Should I use a different protocol? Cheers, Max

4 4

The Trunk: Morphic-tfel.1121.mcz
by commits＠source.squeak.org 20 Apr '16

20 Apr '16

Tim Felgentreff uploaded a new version of Morphic to project The Trunk: http://source.squeak.org/trunk/Morphic-tfel.1121.mcz ==================== Summary ==================== Name: Morphic-tfel.1121 Author: tfel Time: 20 April 2016, 1:45:24.733021 pm UUID: 796c4fe3-21f2-fd4e-a4f9-ccc9905b3503 Ancestors: Morphic-cmm.1120 fix incorrect use of Canvas>>isVisible:. Needs to get a rectangle (but accidentally worked for morphs too *only* on FormCanvas) =============== Diff against Morphic-cmm.1120 =============== Item was changed: ----- Method: SimpleHierarchicalListMorph>>drawLinesOn: (in category 'drawing') ----- drawLinesOn: aCanvas | lColor | lColor := self lineColor. aCanvas transformBy: scroller transform clippingTo: scroller innerBounds during:[:clippedCanvas | scroller submorphs select: [:submorph | submorph visible] thenDo: [ :submorph | ((submorph isExpanded or: [clippedCanvas isVisible: submorph fullBounds] ) + or: [ submorph nextSibling notNil and: [clippedCanvas isVisible: submorph nextSibling bounds]]) - or: [ submorph nextSibling notNil and: [clippedCanvas isVisible: submorph nextSibling]]) ifTrue: [submorph drawLinesOn: clippedCanvas lineColor: lColor indentThreshold: 0] ] ] smoothing: scroller smoothing. !

1 0

Squeak IDE videos
by Chris Muller 20 Apr '16

20 Apr '16

In order to communicate and document some of Squeak's IDE capabilities, I made some videos today which describe some of my favorite features of its IDE. The pinch-zoom-and-resize feature: https://youtu.be/_aMmeNZ39Io The Tracing Messages Browser: https://youtu.be/y96IvSwfXxg The list-filtering feature: https://youtu.be/11CLNKl5334 The Smart-Splitters feature: https://youtu.be/JPQrDMIpjis Focus Follows Mouse feature (old): https://youtu.be/q0WqBjtIbK0

4 3

List filtering stability issues and how to solve them
by marcel.taeumel 20 Apr '16

20 Apr '16

Hi, there. After my somewhat unlucky attempt to improve the list filtering feature in Squeak: http://forum.world.st/The-Trunk-Morphic-mt-1119-mcz-tp4890816.html http://forum.world.st/The-Trunk-Morphic-cmm-1120-mcz-tp4890853.html And the great explanation of the status quo by Chris: http://forum.world.st/Squeak-IDE-videos-td4890897.html https://youtu.be/11CLNKl5334 I want to explain a little bit more, what the issues with the list filtering is at the moment. ----------- Here are the *good *things: - You can just start typing and the filter results appear. - Matches that have the search term at the beginning come first, other matches follow alphabetically. - The parts that matched in an item are highlighted. - Hitting backspace or losing focus (controlled by preference) clears the filter. - Re-typing after some timeout describes a new filter. - If there are 0 matches, the new filter is not applied at all to not have an empty list. Here are the *challenging *things considering stability in interactive, graphical tools: - If there is no selection in the list, the first filter result will be selected automatically. - If the current selection is not included in the filter result, the first filter result will be selected automatically. - If the user makes a selection in the filtered list, removing the filter will revert the selection to the one that used to be there before starting to filter. (Which can be several minutes, days, hours, days, ago) Here is a *change set*, which preserves the good things and fixes the issues: list-filtering-stability-fix.cs <http://forum.world.st/file/n4890920/list-filtering-stability-fix.cs> Here is a feature which -- in my opinion -- was not intended by the original idea of filtering lists and which -- unfortunately -- cannot be retained: - Repeatedly hitting <Backspace> toggles between the last selection in a filtered list and the last selection in an unfiltered list. However, I think that such a workflow is difficult to discover and should have a lower priority than the challenges described above. It might be added with a preference, though. For example, "Use two-level selection in lists with filters". How about it? :-) Best, Marcel -- View this message in context: http://forum.world.st/List-filtering-stability-issues-and-how-to-solve-them… Sent from the Squeak - Dev mailing list archive at Nabble.com.

1 1

Daily Commit Log
by commits＠source.squeak.org 19 Apr '16

19 Apr '16

Changes to Trunk (http://source.squeak.org/trunk.html) in the last 24 hours: http://lists.squeakfoundation.org/pipermail/packages/2016-April/016736.html Name: Morphic-mt.1116 Ancestors: Morphic-mt.1115 Adds preference to enclose a selection with brackets etc. to bypass the respective (shift)cmd-key shortcuts, which actually do not work on German keyboard layouts. Updates wording and preference categorization for other text editing preferences. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2016-April/016737.html ============================================= http://lists.squeakfoundation.org/pipermail/packages/2016-April/016738.html ============================================= http://lists.squeakfoundation.org/pipermail/packages/2016-April/016739.html Name: Morphic-mt.1117 Ancestors: Morphic-mt.1116 Fixes return values of enclose feature. Only return true if you perfomed an edit operation. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2016-April/016740.html Name: Morphic-mt.1118 Ancestors: Morphic-mt.1117 Improves documentation and usage of #refreshWorld to avoid direct calls to #displayWorldSafely, which requires a check whether the morph in in a world or not. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2016-April/016741.html Name: EToys-mt.134 Ancestors: EToys-mt.133 Improves usage of #refreshWorld to avoid direct calls to #displayWorldSafely, which requires a check whether the morph in in a world or not. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2016-April/016742.html Name: MorphicExtras-mt.170 Ancestors: MorphicExtras-cmm.169 Improves usage of #refreshWorld to avoid direct calls to #displayWorldSafely, which requires a check whether the morph in in a world or not. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2016-April/016743.html Name: MorphicExtras-mt.171 Ancestors: MorphicExtras-mt.170 Improves usage of #refreshWorld to avoid direct calls to #displayWorldSafely, which requires a check whether the morph in in a world or not. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2016-April/016744.html Name: Morphic-mt.1119 Ancestors: Morphic-mt.1118 Improves stability of filter in lists: 1) Restore prior selection on filter clear 2) Restore prior selection if there are matches (again) 3) Update prior selection of you make a new selection in a filtered list Note that this will still not be propagated between connected lists in, for example, the code browser. While you now will enjoy more stability when filtering the list of methods, filtering the list of system categories might still update/clear all other lists. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2016-April/016745.html Name: Morphic-cmm.1120 Ancestors: Morphic-mt.1119 Undo Morphic-mt.1119 because it totally broke the filtering function. - It broke navigation by filtering. - It broke toggling between two selections with the backspace key. Filtering is its own isolated feature that works on texts of lists, it was never meant to interact with the model or user selections made outside the filtering feature. =============================================

1 0

← Newer
1
2
3
4
5
6
7
8
9
...
25
Older →

Jump to page: