Squeak-dev November 2022

squeak-dev@lists.squeakfoundation.org

24 participants
94 discussions

How to play YouTube videos in squeak?
by rabbit 24 Dec '22

24 Dec '22

Noch einmal, zwei für einen! I’d like to watch these in Squeak. This would totally make my day! My two favorite pieces of music, anywhere. All time. Infinitely, always & forever! Mozart : Requiem in D K. 626 (Orchestre national de France / James Gaffi...https://youtu.be/Dp2SJN4UiE4 M̤ ̤I̤ ̤L̤ ̤E̤ ̤S̤ ̤D̤ ̤A̤ ̤V̤ ̤I̤ ̤S̤ ̤-̤ ̤K̤i̤n̤d̤ ̤O̤f̤ ̤B̤l̤ṳe̤...https://youtu.be/8uUCIQ50e-4 — Have a good one; keep it, light. Kindly, langohr . .. … ‘…^,^ 🐇🐇🐇 Sent from Callisto House Mobile - Europa :: decentralized mobile homeless solutions ::

4 11

The Trunk: Chronology-Core-dtl.80.mcz
by commits＠source.squeak.org 10 Dec '22

10 Dec '22

David T. Lewis uploaded a new version of Chronology-Core to project The Trunk: http://source.squeak.org/trunk/Chronology-Core-dtl.80.mcz ==================== Summary ==================== Name: Chronology-Core-dtl.80 Author: dtl Time: 19 May 2022, 12:03:05.720962 pm UUID: 1b1b97c4-bc35-408d-b548-f0d0376a0808 Ancestors: Chronology-Core-dtl.79 Change DateAndTime>>offset: to be compliant with the ANSI draft spec. The offset: method should answer an instance equivalent to the receiver, where equivalent means having the same UTC time. Let #offset: adopt the prior behavior of #utcOffset: and let #utcOffset: be a compatibility synonym for #offset:. Add #asLocalAt: to the squeak protocol and let it implement the prior behavior of #offset: Change existing code and tests to use #asLocalAt: but otherwise retain existing behavior. Issue reported at https://github.com/squeak-smalltalk/squeak-object-memory/issues/23 =============== Diff against Chronology-Core-dtl.79 =============== Item was added: + ----- Method: DateAndTime>>asLocalAt: (in category 'squeak protocol') ----- + asLocalAt: anOffset + "Answer a DateAndTime for a different time zone offset that has the same + year, month, day, hour, minute, and second as this instance, and with + printString that matches except for time zone offset." + + | newOffset newMicros | + newOffset := anOffset asDuration asSeconds. + newMicros := localOffsetSeconds - newOffset * 1000000 + utcMicroseconds. + ^ self class utcMicroseconds: newMicros offset: newOffset + ! Item was changed: ----- Method: DateAndTime>>offset: (in category 'ansi protocol') ----- offset: anOffset + "Answer a <DateAndTime> equivalent to the receiver but with its local time + being offset from UTC by offset. - "Answer a DateAndTime for a different time zone offset that has the same - year, month, day, hour, minute, and second as this instance, and with - printString that matches except for time zone offset." + See asLocalAt: for the earlier behavior of this method." + + ^ self class + utcMicroseconds: utcMicroseconds + offset: anOffset asDuration asSeconds! - | newOffset newMicros | - newOffset := anOffset asDuration asSeconds. - newMicros := localOffsetSeconds - newOffset * 1000000 + utcMicroseconds. - ^ self class utcMicroseconds: newMicros offset: newOffset - ! Item was changed: ----- Method: DateAndTime>>utcOffset: (in category 'squeak protocol') ----- utcOffset: anOffset + "Compatiblity. Answer a DateAndTime equivalent to the receiver but offset from UTC by anOffset." + + ^ self offset: anOffset! - "Answer a DateAndTime equivalent to the receiver but offset from UTC by anOffset" - ^ self class - utcMicroseconds: utcMicroseconds - offset: anOffset asDuration asSeconds! Item was changed: ----- Method: Timespan class>>current (in category 'squeak protocol') ----- current + ^ self starting: (DateAndTime now asLocalAt: self defaultOffset)! - ^ self starting: (DateAndTime now offset: self defaultOffset)!

8 11

Building with a strict(er) C99 compiler
by Florian Weimer 03 Dec '22

03 Dec '22

I'm trying to build the VM with a C compiler which does not accept implicit function declarations or implici int. This is in preparation for a future change in GCC and Clang defaults. More information is here: <https://fedoraproject.org/wiki/Changes/PortingToModernC> <https://fedoraproject.org/wiki/Toolchain/PortingToModernC> (The second link points to a container image with an instrumented GCC that can be used for experiments, or you can just inject -Werror=implicit-function-declaration -Werror=implicit-int into the build, which has the same results if compiler flags injection works reliably throughout the build environment.) According to Fedora's records Squeak-4.10.2.2614-src-no-mp3 was downloaded from <http://squeakvm.org/unix/release/>. I got passed the first obstacle with these little changes: diff -ur Squeak-4.10.2.2614-src-no-mp3.orig/Cross/plugins/CroquetPlugin/CroquetPlugin.h Squeak-4.10.2.2614-src-no-mp3/Cross/plugins/CroquetPlugin/CroquetPlugin.h --- Squeak-4.10.2.2614-src-no-mp3.orig/Cross/plugins/CroquetPlugin/CroquetPlugin.h 2006-09-14 19:52:45.000000000 +0200 +++ Squeak-4.10.2.2614-src-no-mp3/Cross/plugins/CroquetPlugin/CroquetPlugin.h 2022-11-29 12:30:00.834158643 +0100 @@ -20,4 +20,6 @@ */ int ioGatherEntropy(char *bufPtr, int bufSize); +int triBoxOverlap(float *a, float *b, float *c, float *d, float *e); + #endif /* CROQUET_PLUGIN_H */ diff -ur Squeak-4.10.2.2614-src-no-mp3.orig/Cross/plugins/CroquetPlugin/TriBoxStub.c Squeak-4.10.2.2614-src-no-mp3/Cross/plugins/CroquetPlugin/TriBoxStub.c --- Squeak-4.10.2.2614-src-no-mp3.orig/Cross/plugins/CroquetPlugin/TriBoxStub.c 2009-05-26 23:15:18.000000000 +0200 +++ Squeak-4.10.2.2614-src-no-mp3/Cross/plugins/CroquetPlugin/TriBoxStub.c 2022-11-29 12:30:10.184066148 +0100 @@ -1,3 +1,5 @@ +#include "CroquetPlugin.h" + /* a stub for triboxoverlap */ int triBoxOverlap(float *a, float *b, float *c, float *d, float *e) { return 0; diff -ur Squeak-4.10.2.2614-src-no-mp3.orig/Cross/plugins/SerialPlugin/SerialPlugin.h Squeak-4.10.2.2614-src-no-mp3/Cross/plugins/SerialPlugin/SerialPlugin.h --- Squeak-4.10.2.2614-src-no-mp3.orig/Cross/plugins/SerialPlugin/SerialPlugin.h 2009-08-27 03:48:06.000000000 +0200 +++ Squeak-4.10.2.2614-src-no-mp3/Cross/plugins/SerialPlugin/SerialPlugin.h 2022-11-29 13:02:56.418443376 +0100 @@ -6,9 +6,13 @@ #pragma export on int serialPortClose(int portNum); +int serialPortCloseByName(const char *portName); + int serialPortCount(void); int serialPortOpen(int portNum, int baudRate, int stopBitsType, int parityType, int dataBits, int inFlowCtrl, int outFlowCtrl, int xOnChar, int xOffChar); +int serialPortOpenByName(char *portName, int dataRate, int stopBitsType, int parityType, int dataBits, + int inFlowCtrl, int outFlowCtrl, int xOnChar, int xOffChar); int serialPortReadInto(int portNum, int count, void *bufferPtr); int serialPortReadIntoByName(const char *portName, int count, void *bufferPtr); int serialPortWriteFrom(int portNum, int count, void *bufferPtr); diff -ur Squeak-4.10.2.2614-src-no-mp3.orig/unix/plugins/AsynchFilePlugin/sqUnixAsynchFile.c Squeak-4.10.2.2614-src-no-mp3/unix/plugins/AsynchFilePlugin/sqUnixAsynchFile.c --- Squeak-4.10.2.2614-src-no-mp3.orig/unix/plugins/AsynchFilePlugin/sqUnixAsynchFile.c 2006-10-18 19:10:25.000000000 +0200 +++ Squeak-4.10.2.2614-src-no-mp3/unix/plugins/AsynchFilePlugin/sqUnixAsynchFile.c 2022-11-29 12:30:44.062730997 +0100 @@ -142,7 +142,7 @@ return fp; } -INLINE static allocateBuffer(struct FileBuf *buf, int size) +INLINE static int allocateBuffer(struct FileBuf *buf, int size) { if (buf->capacity >= size) return 1; A few more changes like these seem to be needed. But now I'm hitting this error: …/unix/src/vm/intplugins/MiscPrimitivePlugin/MiscPrimitivePlugin.c:118: implicit function declaration: isBytes This error is harder to deal with because it seems to be a bug in the code generator. There probably should be a macro definition for isBytes similar to arrayValueOf, like this: #define isBytes(oop) (interpreterProxy->isBytes(oop)) Otherwise, the global function definition is picked up instead, and that does not seem to be the right pattern. Thanks, Florian

4 9

What it is the expected behavior of Rectangle (class) #intersect: if there is no interesection?
by LawsonEnglish 02 Dec '22

02 Dec '22

I did this and created 2 small rectangles quite far from each other trying to figure out what the method returns when there is no intersection: r1 := Rectangle fromUser. r2 := Rectangle fromUser. r2 3072@993 corner: 3203@1136 r1 2710@632 corner: 2835@752 r2 intersect: r1 3072@993 corner: 2835@752 r1 intersect: r2 3072@993 corner: 2835@752 Could someone explain these results? Not in terms of the math involved (I assume that there’s no math bug as the routine was last edited 22 years ago and it is used everywhere in Morphic), but what the LOGIC is for returning what is returned and what it means in the context of the comments: intersect: aRectangle "Answer a Rectangle that is the area in which the receiver overlaps with aRectangle…. “ I was trying to figure out how to look to identify when there is no intersection, and that isn’t at all obvious here. Thanks. L

2 6

Can anyone add me as a committer to http://www.squeaksource.com/INIFile?
by Eliot Miranda 29 Nov '22

29 Nov '22

The package contains underscore assignments. _,,,^..^,,,_ best, Eliot

2 2

The Trunk: Compiler-eem.483.mcz
by commits＠source.squeak.org 29 Nov '22

29 Nov '22

Eliot Miranda uploaded a new version of Compiler to project The Trunk: http://source.squeak.org/trunk/Compiler-eem.483.mcz ==================== Summary ==================== Name: Compiler-eem.483 Author: eem Time: 29 November 2022, 1:46:01.772575 pm UUID: 08cfac57-1240-4bff-bfc9-4af67efb8a5f Ancestors: Compiler-eem.482 Oops! The SistaV1 encoder has been including unnecessarily in the literal frame literals that are encoded in bytecodes (integers and characters in the -32k/32k, 0/64k range). Fix this and recompile. =============== Diff against Compiler-eem.482 =============== Item was changed: ----- Method: LiteralNode>>sizeCodeForValue: (in category 'code generation') ----- sizeCodeForValue: encoder - self reserve: encoder. ^(encoder isSpecialLiteralForPush: key) ifTrue: [encoder sizePushSpecialLiteral: key] + ifFalse: + [self reserve: encoder. + encoder sizePushLiteral: index]! - ifFalse: [encoder sizePushLiteral: index]! Item was changed: (PackageInfo named: 'Compiler') postscript: '"below, add code to be run after the loading of this package" + "recompile any and all SistaV1 compiled methods that refer to special literals. These literals were mistakenly included in the literal frame. saves about 200k bytes in a normal 64-bit Squeak image" + | e nm | + e := EncoderForSistaV1 new. + nm := 0. + ''Recompiling'' + displayProgressFrom: 1 + to: CompiledMethod instanceCount + during: + [:bar| + self systemNavigation allSelect: + [:m| | r | + bar value: (nm := nm + 1). + m encoderClass == EncoderForSistaV1 ifTrue: + [ + r := false. + m codeLiteralsDo: + [:cl| + cl literalsDo: [:l| (e isSpecialLiteralForPush: l) ifTrue: [r := true]]]. + r ifTrue: + [nm := nm + 1. + m methodClass recompile: m selector]]. + false]]'! - "Recompile senders of caseOf:" - self systemNavigation allSelectorsAndMethodsDo: [ :behavior :selector :method | - (method hasLiteral: #caseOf:) - ifTrue: [behavior recompile: selector]]. - - "Treat non-breaking space (16ra0) as delimiter in parser/scanner (Compiler-ct.473)." - Scanner initialize.'!

1 0

UK Smalltalk User Group meeting - 30 November 2022
by Giovanni Corriga 29 Nov '22

29 Nov '22

For this month's UKSTUG meeting, Dave Mason will talk about Zag Smalltalk, a new Smalltalk VM written in Zig. Dave Mason ( http://sarg.ryerson.ca/dmason/ ) has been a professor of Computer Science at Toronto Metropolitan University (previously known as Ryerson) for 41 years. He has done research on operating systems, software reliability and programming languages. Current research is mostly around Smalltalk and other dynamic languages. If forced to program very low level projects such as virtual machines, he is willing to use Zig or Rust, but for any other purpose, he insists on using higher productivity languages - primarily Smalltalk. Zag Smalltalk ( https://github.com/dvmason/Zag-Smalltalk ) is a principle-based Smalltalk VM. "Principled" means that the only 3 operations are: message send, assignment, and return. 1) it is designed from the ground up to use multi-processing, to leverage multi-core systems. 2) There is no special-casing of methods like ifTrue:ifFalse or whileTrue:, although of course some methods are implemented by primitive methods. It does aggressive inlining of methods and blocks. 3) “Source” code is maintained as ASTs. 4) Compiled code runs in a dual form of threaded code and JIT’ed machine code that seamlessly interoperate. 5) It has a partially-copying and partially-non-moving garbage collector. 6) It keeps many more values as immediate, including symbols. 7) It uses a single-level dispatch mechanism to implement Smalltalk dispatch semantics The research question is, "Can this be made fast enough to be competitive?” Preliminary results are encouraging. This will be an online meeting from home. If you'd like to join us, please sign up in advance on the meeting's Meetup page ( https://www.meetup.com/ukstug/events/290036186 ) to receive the meeting details.

1 0

wrt debugger flexibility
by Lauren P 29 Nov '22

29 Nov '22

Hi List, In reference to my question about the debugger in regards to Chronology-Core-dtl.80, I have a proposal. I am halfway or so to stealing RESTART from Common Lisp for Squeak. It lets the programmer define alternatives for exception handling. Restarts would be responsible for making the Proceed, Abandon, etc. buttons in the debugger window, with every #on:offer: defining a button. It's an extra layer of indirection and complexity in the evaluator, but it is an extremely powerful tool for writing clean code. If I had to choose one of an interactive debugger or restarts I'd pick restarts. Implementation is easy but tedious... restarts have lots of optional parameters. If List thinks RESTART would be a good thing to have I will prioritize it. PS: Whoever wrote the signalling code, thank you. It is very clean and easy to read. I learned everything about dynamic extent in Smalltalk from reading it. -- Obviously, I vote yes, and the effort to not proselytize the Cult Of Restart here is great indeed.

4 6

Any progress on a GridLayout LayoutPolicy?
by LawsonEnglish 28 Nov '22

28 Nov '22

Or if it is working in any substantial way, are there any working examples? Thanks. Lawson

2 1

Review Request: browseMethodReference.1.cs
by christoph.thiede＠student.hpi.uni-potsdam.de 28 Nov '22

28 Nov '22

=============== Summary =============== Change Set: browseMethodReference Date: 28 November 2022 Author: Christoph Thiede Adds support for method references in editors's 'Browse it' and 'iMplmentors' shortcuts. Method references can have different formats such as String findMethodReference or String>>#findMethodReference. Includes tests and proper support for message traces. Performance: <1.1 ms per run if you select an entire line, which should be fine. :-) Also makes sure that no debuggers opens up when you browse the iMplementors of an incomplete literal such as a single quote. =============== Diff =============== MessageTrace>>browseMethod:requestor: {actions} · ct 11/28/2022 18:00 + browseMethod: aMethodReference requestor: requestor + "Overwritten to modify the trace if the request origins from a model-menu command such as the message-list menu (shortcut)." + + (Preferences traceMessages and: [ self hasUnacceptedEdits not ]) + ifTrue: [ self + addChildMessages: (OrderedCollection with: aMethodReference) + autoSelectString: '' ] + ifFalse: [ super browseAllImplementorsOf: aMethodReference requestor: requestor ]. Model>>browseMethod: {*Tools-MessageSets} · ct 11/28/2022 18:58 + browseMethod: aMethodReference + "A tool's widget/view (i.e., 'requestor') requested to browse aMethodReference. By default, dispatch back to the reference." + + ^ aMethodReference browse Model>>browseMethod:requestor: {*Tools-MessageSets} · ct 11/28/2022 18:58 + browseMethod: aMethodReference requestor: requestor + "A tool's widget/view (i.e., 'requestor') requested to browse aMethodReference. By default, dispatch back to the reference." + + ^ self browseMethod: aMethodReference ParagraphEditor>>selectedLiteral {menu messages} · ct 11/28/2022 18:18 (changed) selectedLiteral "Try to make a Smalltalk literal out of the current text selection." - ^ self selection string findLiteral + ^ [self selection string findLiteral] ifError: ["e.g., incomplete string"] String>>findMethodReference {converting} · ct 11/28/2022 19:03 + findMethodReference + "Find and parse a method reference within the receiver. Answer a MethodReference or nil if none is found. Heuristic." + + | str substrings | + str := self copyReplaceAll: '>>' with: ' >> '. + str := str copyWithRegex: '(?=(?<![^\p{L}\:\s])[^\p{L}#\:\s]|(?<=[^\p{L}#\:\s])[\p{L}#\:\s])' matchesReplacedWith: ' '. "beginning or end of token" + substrings := str substrings. + substrings size - 1 to: 1 by: -1 do: [:classIndex | | classStr | + classStr := substrings at: classIndex. + (self environment classNamed: classStr) ifNotNil: [:class | + (classIndex + 1 + ((substrings at: classIndex + 1) = '>>') asBit clampHigh: substrings size) to: classIndex + 1 by: -1 do: [:selIndex | + | ref selStr | + selStr := substrings at: selIndex. + (selStr size > 1 and: [selStr first = $#]) ifTrue: [selStr := selStr allButFirst]. + (Symbol lookup: selStr) ifNotNil: [:sel | + ref := MethodReference + class: class + selector: sel. + ref isValid ifTrue: [^ ref]]]]]. + ^ nil StringTest>>testFindMethodReference {tests - finding} · ct 11/28/2022 18:11 + testFindMethodReference + + { + Number >> #negated. 'Number negated'. + Number >> #negated. 'Number>>negated'. + Number >> #negated. 'Number>>#negated'. + Number >> #negated. 'Number >> #negated'. + + Number >> #negated. 'Number >> #negated.'. + Number >> #negated. '"Number >> #negated"'. + Number >> #negated. 'For more information, see Number>>#negated or run the tests.'. + + nil. 'Number absentSelector'. + nil. ''. + + Morph >> #drawOn:. 'Morph drawOn:'. + nil. 'Morph drawOn'. "Do not add colons." + + Number >> #+. 'Number +'. + Number >> #+. 'Number>>+'. + Number >> #+. 'Number>>#+'. + Number >> #+. 'Number+'. + + Behavior >> #>>. 'Behavior >>'. + Behavior >> #>>. 'Behavior>>'. + Behavior >> #>>. 'Behavior>> >>'. + } pairsDo: [:method :string | + self + assert: (method ifNotNil: [method asCodeReference]) + equals: string findMethodReference]. TextEditor>>browseIt {menu messages} · ct 11/28/2022 17:53 (changed) browseIt "Launch a browser for the current selection, if appropriate." + self selectedMethodReference ifNotNil: + [:methodReference | ^self model browseMethod: methodReference requestor: morph]. Preferences alternativeBrowseIt ifTrue: [^ self browseClassFromIt]. self lineSelectAndEmptyCheck: [^ morph flash]. "First, try to show all accesses to instance or class variables." self selectedInstanceVariable ifNotNil: [:nameToClass | self systemNavigation browseAllAccessesTo: nameToClass key from: nameToClass value]. self selectedClassVariable ifNotNil: [:binding | self model browseAllCallsOn: binding]. "Then, either browse the class (from a binding) or all implementors of a selector." self selectedBinding ifNotNil: [:binding | ^ self systemNavigation browseClass: binding]. self selectedSelector ifNotNil: [:selector | ^ self model browseAllImplementorsOf: selector requestor: morph]. morph flash TextEditor>>implementorsOfIt {menu messages} · ct 11/28/2022 17:05 (changed) implementorsOfIt "Open an implementors browser on the selected selector" self lineSelectAndEmptyCheck: [^ self]. + self selectedMethodReference ifNotNil: + [:methodReference | ^self model browseMethod: methodReference requestor: morph]. self selectedSelector ifNotNil: [:aSelector| ^self model browseAllImplementorsOf: aSelector requestor: morph]. self selectedLiteral ifNotNil: [:aLiteral| ^self model browseAllImplementorsOf: aLiteral requestor: morph]. morph flash. TextEditor>>selectedLiteral {menu messages} · ct 11/28/2022 18:18 (changed) selectedLiteral "Try to make a Smalltalk literal out of the current text selection." - ^ self selection string findLiteral + ^ [self selection string findLiteral] ifError: ["e.g., incomplete string"] TextEditor>>selectedMethodReference {menu messages} · ct 11/28/2022 17:06 + selectedMethodReference + "Try to make a method reference out of the current text selection." + + ^ self selection string findMethodReference --- Sent from Squeak Inbox Talk ["browseMethodReference.1.cs"]

1 0

← Newer
1
2
3
4
5
6
7
8
9
10
Older →