Squeak-dev September 2023

squeak-dev@lists.squeakfoundation.org

27 participants
82 discussions

The Inbox: Morphic-ct.2130.mcz
by commits＠source.squeak.org 12 Oct '23

12 Oct '23

Christoph Thiede uploaded a new version of Morphic to project The Inbox: http://source.squeak.org/inbox/Morphic-ct.2130.mcz ==================== Summary ==================== Name: Morphic-ct.2130 Author: ct Time: 11 September 2023, 6:18:59.713063 pm UUID: 3f9268df-f04d-4d4c-bfee-7651b41c1443 Ancestors: Morphic-mt.2122 Avoids installing a new editor in TextMorph when receiving #textStyle:. Regression from Morphic-ct.2060. An editor is not expected before the TextMorph has been added to a world/owner. For instance, #fillsOwner: before #openInWorld was previously broken, and thus the BannerMorph required for Objectland. Please review. I'm not entirely sure whether I identified the correct invariant or whether TextContainer>>#bounds should instead check for a nil owner. Do BannerMorph example openInWorld to reproduce. =============== Diff against Morphic-mt.2122 =============== Item was changed: ----- Method: TextMorph>>textStyle: (in category 'accessing') ----- textStyle: aTextStyle "Change the receiver's set of fonts to aTextStyle. You can access those fonts via the TextFontChange text attribute. If you want to enfore a specific font face or point size, use #font: instead. NOTE THAT you must provide either a freshly created instance of TextStyle or a copy of an existing one. NEVER use, for example, TextStyle class >> #default directly. Also see senders and implementors of #asNewTextStyle." | previousSelection | textStyle := aTextStyle. + previousSelection := editor ifNotNil: [editor selectionInterval]. - previousSelection := self selectionInterval. self releaseParagraph; changed. previousSelection ifNotNil: [self selectionInterval: previousSelection].!

3 2

The Inbox: OcompletionSqueakCompatibility-mad.4.mcz
by commits＠source.squeak.org 12 Oct '23

12 Oct '23

A new version of OcompletionSqueakCompatibility was added to project The Inbox: http://source.squeak.org/inbox/OcompletionSqueakCompatibility-mad.4.mcz ==================== Summary ==================== Name: OcompletionSqueakCompatibility-mad.4 Author: mad Time: 22 September 2023, 4:41:48.24935 pm UUID: b361db4b-3e6f-48b0-bdd2-3527680ea77c Ancestors: OcompletionSqueakCompatibility-JohnMcKeon.3 Fixes the position of the OMenuMorph for (Pluggable)TextMorphs inside TransformMorphs. I know this doesn't belong to trunk, could someone please merge this into OCompletion? Thank you. ==================== Snapshot ==================== SystemOrganization addCategory: #OcompletionSqueakCompatibility! StandardToolSet subclass: #ECToolSet instanceVariableNames: '' classVariableNames: '' poolDictionaries: '' category: 'OcompletionSqueakCompatibility'! ----- Method: ECToolSet class>>codeCompletionAround:textMorph:keyStroke: (in category 'completion') ----- codeCompletionAround: aBlock textMorph: aTextMorph keyStroke: evt | completionAllowed editor stringHolder | editor := aTextMorph editor. stringHolder := editor ifNotNil: [editor model]. completionAllowed := ECPreferences enabled and:[ ECController allowModel: stringHolder ]. completionAllowed ifTrue: [(stringHolder completionController handleKeystrokeBefore: evt editor: editor) ifTrue: [^ self]]. aBlock value. "sends to keyStroke event handler, if any" "Narrow the completion with any of the keys" completionAllowed ifTrue: [stringHolder completionController handleKeystrokeAfter: evt editor: editor]! ----- Method: ECToolSet class>>register (in category 'initialization') ----- register "Asks the system to use me when launching completion" ToolSet register: self! ----- Method: SmalltalkEditor>>isCaretBehindChar (in category '*OcompletionSqueakCompatibility') ----- isCaretBehindChar "Return true if the cursor position is after an alphanumeric character, otherwise false." | cursorPosition | ^(cursorPosition := self startIndex) >= 2 and: [ (self text at: cursorPosition - 1) isAlphaNumeric ]! ----- Method: SmalltalkEditor>>selectionPosition: (in category '*OcompletionSqueakCompatibility') ----- selectionPosition: aString | startIndex | startIndex := self startIndex. ^[ | bottomLeft topLeft index | index := startIndex - aString size. self selectInvisiblyFrom: index to: index - 1. bottomLeft := paragraph selectionRects first bottomLeft. topLeft := self morph boundsInWorld topLeft. topLeft + bottomLeft ] ensure: [ self selectInvisiblyFrom: startIndex to: startIndex - 1 ]! ----- Method: SmalltalkEditor>>wordAtCaret (in category '*OcompletionSqueakCompatibility') ----- wordAtCaret ^paragraph text asString wordBefore: self startIndex - 1 ! ----- Method: ParagraphEditor>>isCaretBehindChar (in category '*OcompletionSqueakCompatibility') ----- isCaretBehindChar "Return true if the cursor position is after an alphanumeric character, otherwise false." | cursorPosition | ^(cursorPosition := self startIndex) >= 2 and: [ (self text at: cursorPosition - 1) isAlphaNumeric ]! ----- Method: ParagraphEditor>>selectionPosition: (in category '*OcompletionSqueakCompatibility') ----- selectionPosition: aString | startIndex | startIndex := self startIndex. ^[ | bottomLeft topLeft index | index := startIndex - aString size. self selectInvisiblyFrom: index to: index - 1. bottomLeft := paragraph selectionRects first bottomLeft. topLeft := self morph boundsInWorld topLeft. topLeft + bottomLeft ] ensure: [ self selectInvisiblyFrom: startIndex to: startIndex - 1 ]! ----- Method: ParagraphEditor>>wordAtCaret (in category '*OcompletionSqueakCompatibility') ----- wordAtCaret ^paragraph text asString wordBefore: self startIndex - 1 !

2 1

interpreter VM on RISC-V with negative bytecodes per second
by Jecel Assumpcao Jr 05 Oct '23

05 Oct '23

Yesterday I managed to compile the intepreter VM on a StarFive VisionFive 2 board (after replacing the config.guess with a current one having gone through several dead ends before figuring that out). The images in the download page have no indication about being V3 or Spur, so I first tried a 5.0 image and when that didn't work I got a 4.0 one that did. I was a bit surprised at how snappy the interpreted Squeak felt on a 1.5GHz reasonably simple RISC-V. Most other applications on that machine (web browser, Open Office, a digital simulator in Java) run much slower than on my work computer (Intel Core i5-8259U CPU @ 2.30GHz) so I didn't expect that. Running "0 tinyBenchmarks" got me a result of about -660 thousand bytecodes per second and 47 thousand sends per second. I will try to find out how this negative result came about and it is very likely that the other result is completely wrong as well. Here are the results on my work machine (with a Cog VM): 0 tinyBenchmarks '1150561797 bytecodes/sec; 32470183 sends/sec' Having the email client able to evaluate Smalltalk expressions can be very convenient at times ;-) -- Jecel

5 10

Emacs & vim text editing in Squeak
by Tim Rowledge 30 Sep '23

30 Sep '23

Somebody was wanting more emacsy stuff for text editing. Whilst idly looking through the SqueakMap errr, map, I noticed a package that might provide the beginnings of an answer to that call - "Name: SVI Summary: Text editor w/ extensive support for VIM and Emacs commands. Author: Steven Swerling Owner: Steven Swerling <sswerling(a)yahoo.com> Description: Now works in 3.9, go <a href="http://www.squeaksource.com/SVI.html">here</a> For Squeak 3.7 or 3.8, use <a href="http://www.swerlingphoto.com/squeak/svi2/SVI-v2.1beta1.1.sar">this</a>. If you want a version that works in Squeak 3.6, go <a href="http://www.swerlingphoto.com/squeak/svi2/svi2.html">here</a>. SVI is a text editor for squeak that can be used in place of the PluggableTextMorph. SVI is designed to behave, as much as possible, like a standard text editor, so as not to confuse users that are not familiar with modal text editors. Nevertheless, SVI features extensive support for VIM and Emacs modes and commands, including: - incremental search - interactive search/replace - split panes - multilevel undo/redo - navigation, using keyboard driven menus and bookmarks, between classes, methods, files, and system windows - context sensitive help and command completion - customizable keybindings - ability to paste recently deleted, inserted, or copied text. Also, VIM-style "registers" - support for VIM's Normal, Visual and Linewise Visual modes, with extensive support for the VIM commands, motions, registers. Mark implementation supports method-to-method jumps , method-to-file, file-to-file, etc. By default, SVI supports many VIM commands. If you want SVI to support Emacs commands instead, be sure to take the tour of Emin (SVI's "mini-emacs"). After loading SVI, execute "SVI openEminDocs" to take a tour of Emin. Patches to make SVI work with StarBrowser and Celeste are available <a href="http://www.swerlingphoto.com/squeak/svi2/appPatches.html">here</a>. The Celeste version is provided as an example of how to fully integrate an application with SVI (with a recent message list, bookmarking, opening a celeste message in a split pane, etc). Some Links: <a href="http://www.swerlingphoto.com/squeak/svi2/sviChanges.txt">Release Change Log</a> <a href="http://www.swerlingphoto.com/squeak/svi2/standardBindings.html">A list of SVI's default keybindings</a> <a href="http://www.swerlingphoto.com/squeak/svi2/standardBindings.html">A list of EMin's default keybindings</a> Categories: Package format/Squeak Archive - A '.sar' file which is a zip-file following certain conventions. Package group/Non official package - Just a package for Squeak, no community guarantees. :) Squeak distribution/Squeak - The official distribution of Squeak. Squeak versions/Squeak3.9 - Released 8 Dec 2006. Published Version: SVI-3.9v2-sps.115 Created: 29 December 2006 2:16:10 pm Release Note: - Added "snippets". See SVI class>>docSnippets. - Another fix for preventing too many repeating keystrokes. This should eliminate dropped keystrokes. - Emin can be turned on/off through the SVI menu - Changed the default help key to <c-h>. SVI's docs are not completely up to date, so where SVI's documentation refers to ''<m-?>'', please use ''<c-h>'' instead. To switch the help key back, you can execute the following (you might want to put it in your svirc file too): SVIKeyMapper helpKey: ''<m-?>''. Homepage: http://www.squeaksource.com/SVI.html Package Info: SVI" I'd be surprised if it "just worked" after this long but it probably has a bunch of good stuff to steal. tim -- tim Rowledge; tim(a)rowledge.org; http://www.rowledge.org/tim Useful Latin Phrases:- Te audire no possum. Musa sapientum fixa est in aure. = I can't hear you. I have a banana in my ear.

2 2

Advice on Best Linux for old pentimento pc
by Edgar De Cleene 29 Sep '23

29 Sep '23

Folks I wish to install squeak on top of Linux in very old pc Which distribution could be ? Thanks in advance Sent from my iPhone

2 2

random selection of an element in an Inspector or Explorer ?
by gettimothy 28 Sep '23

28 Sep '23

Hi Folks. Use case is a collection of PGRows in an explorer or inspector. I would like to do random selection in the inspector/explorer for some initial QA. Is there a way to do this in the Inspector/Explorer? thx in advance.

2 1

Squeak 6.0: Collections-mt.1016.mcz
by commits＠source.squeak.org 28 Sep '23

28 Sep '23

Marcel Taeumel uploaded a new version of Collections to project Squeak 6.0: http://source.squeak.org/squeak60/Collections-mt.1016.mcz ==================== Summary ==================== Name: Collections-mt.1016 Author: mt Time: 28 September 2023, 10:26:14.4145 am UUID: 12073f29-4cbd-7444-8898-f0702a30bb4c Ancestors: Collections-eem.1015 Fixes re-hashing issue with objects from disk. Always #rehash. #compact will not work if the sizes match. =============== Diff against Collections-eem.1015 =============== Item was changed: ----- Method: HashedCollection>>comeFullyUpOnReload: (in category 'objects from disk') ----- comeFullyUpOnReload: smartRefStream "Symbols have new hashes in this image." + self rehash! - self compact!

1 0

SqueakMap 'Update' fails on Squeak trunk but ok on 5.3
by Tim Rowledge 28 Sep '23

28 Sep '23

Whilst updating the SM entries for MQTT etc I went to update the SM tool and it fails when reading the update map. The issue appears to be that deep in the process a Dictionary exists that has a tally of 4162 but an array of nil. Which doesn't seem quite proper somehow. The starting point is SMLoaderPlus>loadUpdates but a long way later we get to SmartRefStream>>#readInstanceSize:clsname:refPosn: with a refPosn of 1270478. A couple of methods later we try to rehash the Dictionary This can never end well. There's no way I have time to do anything much with this, so volunteers needed... tim -- tim Rowledge; tim(a)rowledge.org; http://www.rowledge.org/tim Littergators resolve disputes about rubbish

4 7

The Trunk: System-mt.1425.mcz
by commits＠source.squeak.org 28 Sep '23

28 Sep '23

Marcel Taeumel uploaded a new version of System to project The Trunk: http://source.squeak.org/trunk/System-mt.1425.mcz ==================== Summary ==================== Name: System-mt.1425 Author: mt Time: 28 September 2023, 10:22:20.27702 am UUID: c446d3d6-39ea-e74f-84b2-dcf1487575eb Ancestors: System-mt.1424 Complements Collections-mt.1051 =============== Diff against System-mt.1424 =============== Item was changed: ----- Method: SmartRefStream>>next (in category 'read write') ----- next + "Really write three objects: (version, class structure, object). But only when called from the outside. NOTE THAT we do not have to #rehash anything here. See implementors of #comeFullyUpOnReload:. " - "Really write three objects: (version, class structure, object). But only when called from the outside. " | version ss object | ^ topCall == nil ifTrue: [topCall := #marked. version := super next. version class == SmallInteger ifFalse: [^ version]. "version number, else just a regular object, not in our format, " self checkCrLf. ss := super next. ss class == Array ifFalse: [^ ss]. "just a regular object" (ss at: 1) = 'class structure' ifFalse: [^ ss]. + structures := ss at: 2. - structures := (ss at: 2) rehash. superclasses := (ss size > 3 and: [(ss at: 3) = 'superclasses']) + ifTrue: [ss at: 4] "class name -> superclass name" - ifTrue: [(ss at: 4) rehash] "class name -> superclass name" ifFalse: [Dictionary new]. (self verifyStructure = 'conversion method needed') ifTrue: [^ nil]. object := super next. "all the action here" self restoreClassInstVars. "for UniClasses. version 4" topCall := nil. "reset it" object] ifFalse: [super next] !

1 0

The Trunk: Kernel-mt.1527.mcz
by commits＠source.squeak.org 28 Sep '23

28 Sep '23

Marcel Taeumel uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-mt.1527.mcz ==================== Summary ==================== Name: Kernel-mt.1527 Author: mt Time: 28 September 2023, 10:21:53.11702 am UUID: d4fc65b6-ef98-b64b-af96-464763affa9d Ancestors: Kernel-mt.1526 Complements Collections-mt.1051 =============== Diff against Kernel-mt.1526 =============== Item was changed: ----- Method: Object>>readDataFrom:size: (in category 'objects from disk') ----- readDataFrom: aDataStream size: varsOnDisk "Fill in the fields of self based on the contents of aDataStream. Return self. Read in the instance-variables written by Object>>storeDataOn:. NOTE: This method must send beginReference: before reading any objects from aDataStream that might reference it. Allow aDataStream to have fewer inst vars. See SmartRefStream." | cntInstVars cntIndexedVars | cntInstVars := self class instSize. self class isVariable ifTrue: [cntIndexedVars := varsOnDisk - cntInstVars. cntIndexedVars < 0 ifTrue: [ self error: 'Class has changed too much. Define a convertxxx method']] ifFalse: [cntIndexedVars := 0. cntInstVars := varsOnDisk]. "OK if fewer than now" aDataStream beginReference: self. 1 to: cntInstVars do: [:i | self instVarAt: i put: aDataStream next]. 1 to: cntIndexedVars do: [:i | self basicAt: i put: aDataStream next]. "Total number read MUST be equal to varsOnDisk!!" - self rehash. ^ self "If we ever return something other than self, fix calls on (super readDataFrom: aDataStream size: anInteger)"!

1 0

← Newer
1
2
3
4
5
6
7
8
9
Older →