Squeak-dev June 2022

squeak-dev@lists.squeakfoundation.org

30 participants
309 discussions

[ANN] OpenSmalltalk VM 2022 released (tag 202205110711)
by Marcel Taeumel 11 Oct '22

11 Oct '22

Hi all! We just released the next version of the OpenSmalltalk VM. Please find the binaries here: https://github.com/OpenSmalltalk/opensmalltalk-vm/releases/tag/202205110711 [https://github.com/OpenSmalltalk/opensmalltalk-vm/releases/tag/202205110711] (see VMMaker.oscog-mt.3184 and update.oscog-mt.6.mcm) That version will be used in the upcoming Squeak 6.0 and also updated bundles for Squeak 5.3. And probably in upcoming Cuis releases. :-) Here is an attempt of a change log (since 2020): - Adds ARMv8/Aarch64/ARM64 JIT incl. support for Apple M1 - Adds "fast C primitives" via #FastCPrimitiveFlag - Adds support for catching exceptions in FFI callouts - Adds #primitiveScreenScaleFactor (for DPI-aware images) - Adds primitives 568 and 578 complementing 88 (primitiveSuspend) - Adds #primitiveMultipleBytecodeSetsActive to update image format for SistaV1 - Adds VectorEnginePlugin - Fixes regressions in ARMv6 support - Fixes performance regressions of -metal and -opengl backends on macOS - Fixes -core-graphics backend on macOS - Fixes Retina scaling on macOS, i.e., support "backing scale factor" - Fixes primitive 126 to fail on graphics backends w/o composition buffer - Fixes regressions in vm-display-fbdev on Linux - Fixes time sync (e.g., for DST) on Windows - Fixes UDP binding on Windows I am sure that I forgot something especially in plugin code. Please expand on this. BIG THANKS to everybody who has worked on this release! Personally, I would like to thank Eliot, who is a great software architect who keeps on making the OSVM faster with every commit. Thank you! Best, Marcel (on behalf of the OSVM core dev team)

9 18

PSA | Slow .changes access | Real-time protection in Windows Defender
by Marcel Taeumel 25 Sep '22

25 Sep '22

Hi all -- Provided that you have a .changes file that is bigger than a couple of bytes (here: ~200 MiB), you might experience a serious lag (here: ~5 seconds) when: - Starting the .image - Evaluating some do-its (or print-its) - Saving a method All operations that typically read from or write to the .changes file. The Windows Defender's "real-time protection service" seems to monitor file access. Whatever triggers it, the defender might try to get exclusive read access to a file. Then it scans the last bytes and if it finds something strange, it will continue, eventually scanning THE ENTIRE FILE! So, what is "something strange"? Well, try to evaluate this: '<a href="https://www.microsoft.com"></a>' size. Now, depending on the size of your .changes file, the Windows Defender will begin to scan the entire file for more such patterns. Note that the Defender will never complain or blacklist that file. It is just curious for the moment. Why is the Defender able to get exclusive read access on a do-it? Take a look at SmalltalkImage >> #forceChangesToDisk. There, you can see that we close and re-open the .changes file. That's exactly the time when the Windows Defender kicks in. Avoid closing that file, and you will not experience any lag. Try to save a method with such contents. You will be able to observe the same amount of extra lag. Why is regular read access during code browsing not affected? Because we already have the open file handle to the .changes and .sources files. The Defender seems to need exclusive read access. In my experience, this happened, for example, during VM/image startup. (Note that this has nothing to do with the Defender sending stuff to Microsoft via Internet. I did all the tests offline. No Ethernet connection. No WiFi.) *** What can we do about it? Different approaches work: - Save a method (or do-it sth.) to add about 1000 bytes of uninteresting content to the .changes file; then everything is fast again - Shrink your .changes file via #condenseChanges (and don't forget to save your image afterwards!!) to speed up the Defender's scan - Disable the "real-time protection service" in the Windows settings (not recommended) - Avoid strange content in your source code (or do-its) - Do not close the file so often. #primFlush: is not affected. Best, Marcel

4 5

The Trunk: Tools-mt.1087.mcz
by commits＠source.squeak.org 20 Sep '22

20 Sep '22

Marcel Taeumel uploaded a new version of Tools to project The Trunk: http://source.squeak.org/trunk/Tools-mt.1087.mcz ==================== Summary ==================== Name: Tools-mt.1087 Author: mt Time: 15 December 2021, 4:52:00.953892 pm UUID: 627887b0-5064-ab43-b7ac-60b329a8b606 Ancestors: Tools-mt.1086 Tweak BlockClosureInspector to show all the closured stuff including temps from the outer context and inst-vars from the current receiver. Adds some caching to make it faster. Support cache invalidation only for changing the #receiver field. Not sure whether it makes sense to support changing the outerContext as well... =============== Diff against Tools-mt.1086 =============== Item was changed: Inspector subclass: #BlockClosureInspector + instanceVariableNames: 'debuggerMap tempVarNames instVarNames' - instanceVariableNames: '' classVariableNames: '' poolDictionaries: '' category: 'Tools-Inspector'! Item was added: + ----- Method: BlockClosureInspector>>debuggerMap (in category 'private') ----- + debuggerMap + + ^ debuggerMap ifNil: [debuggerMap := self object method debuggerMap]! Item was changed: ----- Method: BlockClosureInspector>>hasBindingOf: (in category 'user interface - styling') ----- hasBindingOf: aString self object method literalsDo: [:literal | (literal isVariableBinding and: [literal key = aString]) ifTrue: [^ true]]. + + (self tempVarNames includes: aString) ifTrue: [^ true]. + (self instVarNames includes: aString) ifTrue: [^ true]. + + ^ false! - ^ self object outerContext tempNames includes: aString! Item was added: + ----- Method: BlockClosureInspector>>instVarNames (in category 'private') ----- + instVarNames + + ^ instVarNames ifNil: [instVarNames := (self object receiver perform: #class "avoid inlining; support proxies") allInstVarNames]! Item was added: + ----- Method: BlockClosureInspector>>replaceSelectionValue: (in category 'selection') ----- + replaceSelectionValue: anObject + + | success | + (success := super replaceSelectionValue: anObject) ifTrue: [ + self selectedField key = #receiver + ifTrue: [instVarNames := nil]]. + ^ success! Item was changed: ----- Method: BlockClosureInspector>>streamBindingsOn: (in category 'fields - streaming') ----- streamBindingsOn: aStream + | compiledCode instNames tempNames | - | compiledCode tempNames | self flag: #decompile. "mt: Use #to: and #do: instead of #to:do: to avoid inlining to preserve bindings in enumeration block for later decompilation. See InspectorField." + + (1 to: (instNames := self instVarNames) size) do: [:index | + aStream nextPut: ((self newFieldForType: #instVar key: index) + name: ('[[[{1}]]]' format: {instNames at: index}); + valueGetter: [:closure | closure receiver instVarAt: index]; + valueSetter: [:closure :value | closure receiver instVarAt: index put: value]; + yourself)]. + + tempNames := self tempVarNames. + "No need to show closured bindings from outerContext. The important temps are already covered in the indexed fields below." + (1 to: tempNames size - self object basicSize) do: [:index | + aStream nextPut: ((self newFieldForType: #tempVar key: index) + name: ('[[{1}]]' format: {tempNames at: index}); + valueGetter: [:closure | closure outerContext tempAt: index]; + valueSetter: [:closure :value | closure outerContext tempAt: index put: value]; + yourself)]. + (1 to: (compiledCode := self object method) numLiterals) do: [:index | | literal | literal := compiledCode literalAt: index. literal isVariableBinding ifTrue: [ aStream nextPut: ((self newFieldForType: #tempVar key: index) name: ('[{1}]' format: {literal key}); valueGetter: [:closure | (closure method literalAt: index) value]; valueSetter: [:closure :value | (closure method literalAt: index) value: value]; yourself)]]. + - - tempNames := self object outerContext tempNames. - "No need to show closured bindings from outerContext. The important temps are already covered in the indexed fields below." - "(1 to: tempNames size - self object basicSize) do: [:index | - aStream nextPut: ((self newFieldForType: #tempVar key: index) - name: ('[{1}]' format: {tempNames at: index}); - valueGetter: [:closure | closure outerContext tempAt: index]; - valueSetter: [:closure :value | closure outerContext tempAt: index put: value]; - yourself)]." - (1 to: self object basicSize) do: [:index | aStream nextPut: ((self newFieldForType: #tempVar key: index) name: ('[{1}]' format: {tempNames at: index + tempNames size - self object basicSize}); valueGetter: [:closure | closure basicAt: index]; valueSetter: [:closure :value | closure basicAt: index put: value]; yourself)]. "See commentary in #doItReceiver. No need to expose the closured 'self' here because there is already the #receiver instVar. fieldReceiver ^ (self newFieldForType: #tempVar key: #receiver) name: '[self]'; valueGetter: [:obj | obj receiver]; valueSetter: [:obj :value | obj receiver: value]; yourself "! Item was added: + ----- Method: BlockClosureInspector>>tempVarNames (in category 'private') ----- + tempVarNames + + ^ tempVarNames ifNil: [tempVarNames := self debuggerMap tempNamesForContext: self object outerContext]!

2 2

The Trunk: Graphics-mt.520.mcz
by commits＠source.squeak.org 15 Sep '22

15 Sep '22

Marcel Taeumel uploaded a new version of Graphics to project The Trunk: http://source.squeak.org/trunk/Graphics-mt.520.mcz ==================== Summary ==================== Name: Graphics-mt.520 Author: mt Time: 8 June 2022, 2:50:00.935335 pm UUID: 8fb5feb3-71e4-1240-8c76-b10b119a18e2 Ancestors: Graphics-mt.519 Fixes a rounding issue with fractional points in integer rectangles. =============== Diff against Graphics-mt.519 =============== Item was changed: ----- Method: Rectangle>>pointAtFraction: (in category 'rectangle functions') ----- pointAtFraction: relativePoint + "For values between 0.0 and 1.0, answers a point that lies within the receiver. + This method is a more general form of #center (meaning 0.5(a)0.5), #topLeft + (meaning 0.0(a)0.0), #bottomCenter (meaning 0.5(a)1.0), etc. + NOTE THAT for proportions that equal 1.0, the resulting point will NOT lie + within the receiver, i.e., #containsPoint: will answer false." + - | result | result := self origin + (self extent * relativePoint). ^ self isIntegerRectangle + ifTrue: [result floor] - ifTrue: [result rounded] ifFalse: [result]!

2 1

Difference between Tools and Apps menu?
by Jaromir Matas 15 Sep '22

15 Sep '22

Hi All, another quick question: What is the difference between Tools and Apps menu? Version 4 introduced Tools menu, version 5 added Apps menu. I can't figure out what the difference is other than Apps are "More Tools". But then why would some entries be in both? So it's probably not it... Or is it about frequency, importance? And then there's "Preferences" in Tools and the same tool named "Preference Browser" in Apps which is also confusing. Thanks for any advice :) -- Jaromír Matas mail(a)jaromir.net

5 5

Installer inbox cleanup
by Tim Johnson 15 Sep '22

15 Sep '22

Hi, Can someone please delete my inbox contribution, Installer-Core-tcj-440.mcz from August 2020? It is superseded by Installer-Core-lrnp.447.mcz which is (I believe) the same intention but better executed. Thanks, Tim

2 1

The Trunk: System-tonyg.1224.mcz
by commits＠source.squeak.org 14 Sep '22

14 Sep '22

Tony Garnock-Jones uploaded a new version of System to project The Trunk: http://source.squeak.org/trunk/System-tonyg.1224.mcz ==================== Summary ==================== Name: System-tonyg.1224 Author: tonyg Time: 29 March 2021, 10:15:02.279821 am UUID: 9d897d72-46fc-4205-a27b-89c7a50ac104 Ancestors: System-mt.1223 Teach SystemOrganizer about multiple environments. =============== Diff against System-mt.1223 =============== Item was changed: Categorizer subclass: #SystemOrganizer + instanceVariableNames: 'environment' - instanceVariableNames: '' classVariableNames: '' poolDictionaries: '' category: 'System-Support'! !SystemOrganizer commentStamp: '<historical>' prior: 0! My instances provide an organization for the classes in the system, just as a ClassOrganizer organizes the messages within a class. The only difference is the methods for fileIn/Out.! Item was changed: ----- Method: SystemOrganizer>>classesIn: (in category 'query') ----- classesIn: categoryName | classes | classes := OrderedCollection new. self categories withIndexCollect: [:cat :idx | (categoryName match: cat) ifTrue: [classes addAll: (self listAtCategoryNumber: idx)] ifFalse: [nil]]. + ^ classes collect: [:clsName | self environment classNamed: clsName]! - ^ classes collect: [:clsName | Smalltalk classNamed: clsName]! Item was changed: ----- Method: SystemOrganizer>>classify:under: (in category 'accessing') ----- classify: element under: newCategory | oldCategory class | self flag: #environments. "do we want notifications for classes in other environments?" oldCategory := self categoryOfElement: element. super classify: element under: newCategory. + class := self environment at: element ifAbsent: [^ self]. - class := Smalltalk at: element ifAbsent: [^ self]. self == SystemOrganization ifTrue: [ SystemChangeNotifier uniqueInstance class: class recategorizedFrom: oldCategory to: newCategory]! Item was changed: ----- Method: SystemOrganizer>>commentInventory: (in category 'query') ----- commentInventory: categoryName "SystemOrganization commentInventory: 'Morphic*'" | classes commentedClasses | classes := OrderedCollection new. self categories withIndexCollect: [:cat :idx | (categoryName match: cat) ifTrue: [classes addAll: (self listAtCategoryNumber: idx)] ifFalse: [nil]]. + commentedClasses := classes select: [:catCls | (self environment at: catCls) hasComment]. - commentedClasses := classes select: [:catCls | (Smalltalk at: catCls) hasComment]. ^ 'There are ' , classes size asString , ' classes in ' , categoryName , ' of which ' , commentedClasses size asString , ' have comments and ', (classes size - commentedClasses size) asString , ' do not yet have comments.' ! Item was added: + ----- Method: SystemOrganizer>>environment (in category 'accessing') ----- + environment + ^ environment ifNil: [Smalltalk globals]! Item was added: + ----- Method: SystemOrganizer>>environment: (in category 'accessing') ----- + environment: anEnvironment + environment := anEnvironment! Item was changed: ----- Method: SystemOrganizer>>fileOutCategory:on:initializing: (in category 'fileIn/Out') ----- fileOutCategory: category on: aFileStream initializing: aBool "Store on the file associated with aFileStream, all the traits and classes associated with the category and any requested shared pools in the right order." | first poolSet tempClass classes traits | traits := self orderedTraitsIn: category. classes := self superclassOrder: category. poolSet := Set new. classes do: [:class | class sharedPools do: [:eachPool | poolSet add: eachPool]]. poolSet size > 0 ifTrue: [ tempClass := Class new. tempClass shouldFileOutPools ifTrue: [ poolSet := poolSet select: [:aPool | + tempClass shouldFileOutPool: (self environment keyAtIdentityValue: aPool)]. - tempClass shouldFileOutPool: (Smalltalk globals keyAtIdentityValue: aPool)]. poolSet do: [:aPool | tempClass fileOutPool: aPool onFileStream: aFileStream]]]. first := true. traits, classes do: [:each | first ifTrue: [first := false] ifFalse: [aFileStream cr; nextPut: Character newPage; cr]. each fileOutOn: aFileStream moveSource: false toFile: 0 initializing: false]. aBool ifTrue: [classes do: [:cls | cls fileOutInitializerOn: aFileStream]].! Item was changed: ----- Method: SystemOrganizer>>orderedTraitsIn: (in category 'fileIn/Out') ----- orderedTraitsIn: category "Answer an OrderedCollection containing references to the traits in the category whose name is the argument, category (a string). The traits are ordered so they can be filed in." | behaviors traits | behaviors := (self listAtCategoryNamed: category asSymbol) + collect: [:title | self environment at: title]. - collect: [:title | Smalltalk at: title]. traits := behaviors reject: [:each | each isBehavior]. ^traits asArray sort: [:t1 :t2 | (t2 traitComposition allTraits includes: t1) or: [(t1 traitComposition allTraits includes: t2) not]]! Item was changed: ----- Method: SystemOrganizer>>removeMissingClasses (in category 'remove') ----- removeMissingClasses "Remove any class names that are no longer in the Smalltalk dictionary. Used for cleaning up after garbage collecting user-generated classes." "SystemOrganization removeMissingClasses" elementArray copy do: [:el | + (self environment includesKey: el) ifFalse: [self removeElement: el]]. - (Smalltalk includesKey: el) ifFalse: [self removeElement: el]]. ! Item was changed: ----- Method: SystemOrganizer>>superclassOrder: (in category 'fileIn/Out') ----- superclassOrder: category "Answer an OrderedCollection containing references to the classes in the category whose name is the argument, category (a string). The classes are ordered with superclasses first so they can be filed in." | behaviors classes | behaviors := (self listAtCategoryNamed: category asSymbol) + collect: [:title | self environment at: title]. - collect: [:title | Smalltalk at: title]. classes := behaviors select: [:each | each isBehavior]. ^ChangeSet superclassOrder: classes!

5 7

ResumableTestFailure
by Thiede, Christoph 13 Sep '22

13 Sep '22

Hi all! :-) I was just working a bit with Squeak by Example (the book) and found out that resumable assertions as described in the book (see chapter SUnit, Advanced, Continuing after a failure) do not any longer work in a trunk image. If you write a test case with some resumable assertions, turn on logging, and run the test, only the first failure will be logged, not the rest, too. (Methods to reproduce: isLogging ^ true "You will need to open a Transcript" testResumable #(1 2 3 4 5) do: [:i | self assert: i even description: ['{1} is not even' format: {i}] resumable: true] ) A second issue I experienced frequently when I felt the desire to work with resumable assertions was that they are basically only available for two selectors: #assert:description:resumable: and #deny:description:identical:. I often would like to insert a resumable into any test without being that verbose, and nearly redoubling the number of assertion selectors by adding a resumable argument to each of them did not feel right, either. Please find the attached changeset, which fixes both issues. * In TestResult >> #runCase:, resumable assertions are identified and - finally - resumed, after marking the test as failed. Unfortunately, other test runners (yes, I'm thinking of smalltalkCI) will need to reimplement this feature unless they reuse the TestResult implementation. * In TestCase, I refactored some methods that raise assertion failures and added the most basic #signalFailure:resumable: as the single source of signalization. Instead of setting the resumable bit manually for each assertion, you now can also make the whole test execution resumable by: * overriding #isResumable, * setting a resumable flag using #resumable:, * setting a resumable flag temporarily using #beResumable:during:, or * adding a <resumable> or <resumable: true> pragma to the test method. * In addition, I tested the added and fixed behavior in ResumableTestFailureTestCase. Looking forward to your review! :-) Best, Christoph

2 2

The Inbox: Kernel-ct.1294.mcz
by commits＠source.squeak.org 09 Sep '22

09 Sep '22

A new version of Kernel was added to project The Inbox: http://source.squeak.org/inbox/Kernel-ct.1294.mcz ==================== Summary ==================== Name: Kernel-ct.1294 Author: ct Time: 9 January 2020, 3:42:36.672555 pm UUID: 7409dcc3-160e-9d44-9712-f266e0f25973 Ancestors: Kernel-nice.1292 Proposal: Add #haltOnceIf: to Object, which combines the strengths of #haltOnce and #haltIf:. Especially useful in performance-critical situations: It HaltOnce is disabled, the condition will not be evaluated at all. =============== Diff against Kernel-nice.1292 =============== Item was changed: ----- Method: Object>>haltIf: (in category 'debugging') ----- + haltIf: aCondition + "This is the typical message to use for inserting breakpoints during debugging. Param can be a block or expression, halt if true. - haltIf: condition - "This is the typical message to use for inserting breakpoints during - debugging. Param can be a block or expression, halt if true. If the Block has one arg, the receiver is bound to that. + If the condition is a selector, we look up in the callchain. Halt if any method's selector equals selector." - If the condition is a selector, we look up in the callchain. Halt if - any method's selector equals selector." - | cntxt | + ^ (self meetsHaltCondition: aCondition) + ifTrue: [self halt]! - condition isSymbol ifTrue:[ - "only halt if a method with selector symbol is in callchain" - cntxt := thisContext. - [cntxt sender isNil] whileFalse: [ - cntxt := cntxt sender. - (cntxt selector = condition) ifTrue: [Halt signal]. - ]. - ^self. - ]. - (condition isBlock - ifTrue: [condition cull: self] - ifFalse: [condition] - ) ifTrue: [ - Halt signal - ].! Item was added: + ----- Method: Object>>haltOnceIf: (in category 'debugging') ----- + haltOnceIf: aCondition + "Check aCondition and halt, unless we have already done it once. See #meetsHaltCondition:." + + self haltOnceEnabled ifFalse: [^ self]. + ^ (self meetsHaltCondition: aCondition) + ifTrue: [ + self clearHaltOnce; halt].! Item was added: + ----- Method: Object>>meetsHaltCondition: (in category 'debugging') ----- + meetsHaltCondition: aCondition + "If the Block has one arg, the receiver is bound to that. + If the condition is a selector, we look up in the callchain. Halt if any method's selector equals selector." + + aCondition isSymbol ifTrue: [ | ctxt | + "only halt if a method with selector symbol is in callchain" + ctxt := thisContext. + [(ctxt := ctxt sender) isNil] whileFalse: [ + (ctxt selector = aCondition) ifTrue: [^ true] ]. + ^ false ]. + ^ aCondition isBlock + ifTrue: [aCondition cull: self] + ifFalse: [aCondition value]!

2 1

[Sounds] How to turn off reverb
by Thiede, Christoph 20 Jul '22

20 Jul '22

Hi all, I never have dealt with the Sound system in Squeak before so here is a newbie question from my side: How can I turn off the reverb which you hear every time you play an FM sound? Take this example: AbstractSound dial: '#'. Either my speakers are broken or there is a reverb that is hidden somewhere in the implementation of FMSound. :-) I can also reproduce this when creating an FMSound manually without defining any envelopes. I also took a look at the ingenious Speech-TTS package from Squeak 3 and could hear a reverb there, too. Is this a bug? Is this a - maybe unfortunate - default setting? How can I turn off this? Thanks in advance! :-) Best, Christoph

8 13

← Newer
1
2
3
4
5
6
7
...
31
Older →

Jump to page: