Squeak-dev April 2018

squeak-dev@lists.squeakfoundation.org

47 participants
159 discussions

Proof-of-concept: moving Smalltalk code editing support into SmalltalkEditor
by Tim Johnson 30 Aug '18

30 Aug '18

Hi all, I went on a journey this morning which could have been quixotic, but I believe I actually found and slayed a small windmill. I was a bit annoyed by how autoEnclose (and by extension autoIndent) were in effect for even the most basic of TextMorphs, e.g. those dragged out of the Supplies flap. I was also kind of disappointed that turning on/off the auto-enclose preference affected all text morphs, even those not specifically designated for Smalltalk code editing. I was emboldened when I saw the class comment for TextEditor: "[I] have no specific facilities for editing Smalltalk code. Those are found in SmalltalkEditor." I found myself able to address this by making a few changes: * moving autoEnclose and autoIndent functionality out of TextEditor and into SmalltalkEditor * changing the default editor class for TextMorph to TextEditor * specifying that the #editorClass for TextMorphForEditView should be a SmalltalkEditor * changing PreferenceWizard to recognize these preferences have a new home I'm attaching a changeset to this email and not uploading an .mcz to inbox (sorry). This is because I don't know how to represent changes across many projects in Monticello, and I sort of consider this a proof-of-concept rather than something I think warrants immediate inclusion in the image. Please do let me know what you think about this idea. I also fear my implementation of SmalltalkEditor>>#dispatchOnKeyboardEvent: might be lacking. Thanks, Tim ps - I think a next step of moving Smalltalk editing support into SmalltalkEditor might be to move the call to ToolSet>>#codeCompletionAround:textMorph:keyStroke: out of TextMorph>>#keyStroke: and into its subclass, in TextMorphForEditView>>#keyStroke:. I didn't complete that move and did not quite figure out how the subclass version of that method works.

9 19

Survey: what do you do with Squeak, what do you *want* to do?
by Benoit St-Jean 01 Aug '18

01 Aug '18

Q-What do you use Squeak for? Genetic algorithms, scripting (a LOT), experimenting, Euler Project code, a chess engine, random data generator, cleaning up/transforming crappy data from databases and/or other sources, keeping up with new ideas in Smalltalk. Q-If you don't use Squeak, why not? Lack of a decent GUI builder. And unstability of the base code and/or difficulty to find code/frameworks/applications that load in-one-shot without me having to debug in the first place! Q-If you used Squeak in the past and don't now, what pulled you away? Lack of a decent GUI builder. Lack of a decent Windows support. Q-What does Squeak lack that you think might make you use it for 'regular' development? A decent GUI builder. And a central code repository that wouldn't make loading code a crazy adventure every time. If I want Python code, all I have to do is to visit PyPI. If I want Ruby code, all I have to do is to visit RubyGems. If I want R code, all I have to do is to visit CRAN R. If I want VisualWorks code, all I have to do is to visit Cincom Public Repository... Now, if I want Squeak code, I have to visit SqueakSource, SmalltalkHub, GitHub, SqueakSource3, SqueakMap and use a GitHub tool, or file in some code, or use the package manager, or use Monticello, or use the Package Universe browser, or Gofer, or the SqueakMap Catalog Browser, or... Well, you get the picture. Besides, once I've found what I was looking for, often times the code doesn't load because it was ported to Pharo... or it's never been updated to be compatible with the current Squeak version. Another tool Squeak needs is an image packager (à la VW or VAST or Dolphin) that works! NOT scripts! Q-What things are too hard or annoying to do? Building GUIs, again. Package an image to deploy. Keeping up with the base code. Who need traits, really? Who uses traits, really? It would have been more beneficial to invest that time/effort into cleaning up morphic, modularizing the code, create a GUI builder and an Image Stripper ! Q-What would you like to be able to use Squeak for? Everything. I've always been a Smalltalker! ----------------- Benoît St-Jean Yahoo! Messenger: bstjean Twitter: @BenLeChialeux Pinterest: benoitstjean Instagram: Chef_Benito IRC: lamneth Blogue: endormitoire.wordpress.com "A standpoint is an intellectual horizon of radius zero". (A. Einstein)

11 15

Bug: SoundService default is not a class
by Nicolas Cellier 19 Jul '18

19 Jul '18

Hi Marcel, Tobias, I encountered a bug, when playing a sound, the default SoundService (class inst var) was not a class (BasedSoundService) but an instance (a BasedSoundService). Why? Because, there is a TileMorphTest>>setUp that remember which is the default, then tries to restore it in tearDown. It used to work. But, in between, SoundService default and defaultOrNil have been re-defined so as to answer an instance rather than a class. Thus, when we play the tests, we can no more play the sounds... I hesitate to open a Mantis issue, no one read nor close them, so the bug report is only here for the moment.

8 21

Dynamic essay project MorphLayoutArticle on Bob's SuperSwiki?
by H. Hirzel 28 Jun '18

28 Jun '18

Hello Is there a backup-copy/mirror available of the MorphLayoutArticle project which was on Bob's SuperSwiki? [1] Regards Hannes ---------------------------------- [1] http://wiki.squeak.org/squeak/2141 How to lay out submorphs Please read the excellent dynamic essay project MorphLayoutArticle (broken link) on Bob's SuperSwiki. Every Morph now has the capability to layout it's submorphs. (Previously, only the AlignmentMorph could implement layout policies. AlignmentMorph is still available because of compatibility reasons and some utility methods it implements.

10 35

Re: [squeak-dev] MorphicProject subclass: #EtoysProject
by LEnglish 08 Jun '18

08 Jun '18

Hmmm? What about if you want to implement a separate hierarchy in a separate System Window, or multiple textures embedded in OpenGL objects? I’m not sure how Craig Latta’s done some of his Caffeine demos, but I doubt if all those Morphs in various independent Canvases are embedded in a single WorldMorph. L "Debugging is twice as hard as writing the code in the first place. Therefore, if you write the code as cleverly as possible, you are, by definition, not smart enough to debug it." -Brian Kernighan > On Oct 3, 2017, at 06:03, Marcel Taeumel <marcel.taeumel(a)hpi.de> wrote: > > +1 :) > > And then later: Rename PasteUpMorph to WorldMorph, and keep an empty PasteUpMorph subclass around for compatibility reasons. So many ideas have been ported down to Morph class over the past years. New applications have no reason to ever use other instances of PasteUpMorph. > > Best, > Marcel >> Am 03.10.2017 14:57:55 schrieb H. Hirzel <hannes.hirzel(a)gmail.com>: >> >> On 10/3/17, H. Hirzel wrote: >> > Dave >> > >> > your change set contains the class EtoysProject with >> > >> > EtoysProject selectors >> > >> > #(#finalEnterActions: #restoreGlobalPreferences #saveGlobalPreferences >> > #initializeProjectPreferences #configureOnFirstEntry >> > #finalExitActions:) >> > >> > For complete configuration of a EtoysProject it might be necessary to do >> > >> > PasteUpMorph subclass: EtoysPasteUpMorph >> > >> > as well. http://wiki.squeak.org/squeak/6461 >> > >> > Then Etoys related methods may be pushed down to EtoysPasteUpMorph. >> >> See screen shot attached. >> >> > And probably an Etoys specific subclass of WorldMenu would be fine as well >> > http://wiki.squeak.org/squeak/6461 >> > >> > >> > there is a test project [2] and some more information about adaptions >> > needed because of the UI changes in the thread 'Etoys in 2017?' - UI >> > preferences [3]. And it would be good to have Etoys methods / >> > configuration separate [4]. >> > >> > I suggest that you start go ahead and start implementing this while >> > using a test Etoys project dropped onto the desktop. >> > >> > --Hannes >> > >> > >> > [2] > You simply drop it in. E.g. download this project >> >> http://etoys.laptop.org/src/Content/ExampleEtoys/CarAndPen.014.pr >> > >> > [3] Hannes Hirzel, 'Etoys in 2017?' mail, Wed, Feb 22, 2017 at 11:01 AM >> > >> > [4] David T. Lewis, Sep 4, 2016 at 3:34 PM >> > "I think it would be great if both Etoys and Scratch were easily >> > loadable and unloadable in trunk." >> > >> > On 10/2/17, David T. Lewis wrote: >> >> An EtoysProject is a project that is configured for running Etoys. On >> >> first entry to a new EtoysProject, the playground and project preferences >> >> are initialized to provide an environment similar to that of a >> >> traditional >> >> standalone Etoys image. >> >> >> >> Certain preferences that are required for Etoys are initialized on >> >> project >> >> entry, overriding their global preference values while this EtoysProject >> >> is active. On leaving the project, these preferences are restored to >> >> their >> >> previous values. >> >> >> >> "ProjectViewMorph openOn: EtoysProject new" >> >> >> >> Change set attached for a minimal implementation. >> >> >> >> Anyone with Etoys knowledge care to help? I do not know enough about >> >> Etoys >> >> to fill in the rest of the initialization that will be required, but it >> >> should not be hard to do. >> >> >> >> Dave >> >> >> >> >> > >> >

3 3

Pressing Esc on empty tree morph raises MNU
by Jakob Reschke 27 May '18

27 May '18

Hi all, I wanted to open the yellow button menu on an empty PluggableTreeMorph (the SqueakMap catalog's package list to be precise, but I assume it does not matter) and wanted to do so by pressing the escape key and got the stack trace below. The scroller of the PluggableTreeMorph is a TransformMorph, but it has no submorphs if the tree is empty. Thus, in numberOfItemsPotentiallyInView the localSubmorphBounds are nil, and even if they were not, the next error would be a division by zero. The TransformMorph does not expect to have no submorphs. However, the real problem might be (or might not be) that a PluggableTreeMorph behaves differently from a PluggableListMorph in response to pressing Esc. The latter opens the yellow button menu, while the tree morph does not treat Esc specially at all. It uses it for list filtering instead, like any typed character. So effectively, the Esc key in a tree morph is currently the "expand this node and select the first child" key. I like the menu behavior from list morphs better... What is your opinion? Kind regards, Jakob UndefinedObject(Object)>>doesNotUnderstand: #height TransformMorph>>numberOfItemsPotentiallyInView PluggableTreeMorph(ScrollPane)>>numSelectionsInView PluggableTreeMorph(SimpleHierarchicalListMorph)>>arrowKey: PluggableTreeMorph(SimpleHierarchicalListMorph)>>specialKeyPressed: PluggableTreeMorph>>specialKeyPressed: PluggableTreeMorph(SimpleHierarchicalListMorph)>>keyStroke: PluggableTreeMorph>>keyStroke: PluggableTreeMorph(Morph)>>handleKeystroke: KeyboardEvent>>sentTo: PluggableTreeMorph(Morph)>>handleEvent: PluggableTreeMorph(Morph)>>handleFocusEvent: MorphicEventDispatcher>>doHandlingForFocusEvent:with: MorphicEventDispatcher>>dispatchFocusEvent:with: PluggableTreeMorph(Morph)>>processFocusEvent:using: PluggableTreeMorph(Morph)>>processFocusEvent: [] in [] in [] in HandMorph>>sendFocusEvent:to:clear: BlockClosure>>ensure: KeyboardEvent(MorphicEvent)>>becomeActiveDuring: [] in [] in HandMorph>>sendFocusEvent:to:clear:

2 2

[Please Test] Refactoring for BorderedMorph to use BorderStyle
by Marcel Taeumel 22 May '18

22 May '18

Hi, there. Please find attached a refactoring of BorderedMorph to make use of BorderStyle like regular morphs do. Maybe, in the future, we can get rid of BorderedMorph. It includes: - No instVar access to borderColor and borderWidth but message sends. - No #inset or #raised anymore when asking a morph for its #borderColor. Just colors. - Copying the prototypical border styles from the UI theme when used in morphs. - A post-load script that updates all your morphs in the image. Not that important but good for keeping your current tools opened. After a few days of no complaints, I will merge it into Trunk. Best, Marcel

8 33

Improvements to wiki.squeak.org
by K K Subbu 20 May '18

20 May '18

Hi, wiki.squeak.org/squeak is an excellent resource for Squeakers. However, it's interface is a bit dated and could do with some improvements. I hope I am not burdening maintainers with my site suggestions. * search could be a text field instead of a button. This will ease the interface for beginners * The search results prints lots of text (including the IP address of the machine from which the author last edited the text!). Printing the page title and modification timestamp (ISO) should be sufficient. VM Command Line Options, 2006-11-12 9:12 pm * Is it possible to tag pages (say with #obsolete or #beginner #developer). Pages with #obsolete tags could be listed at end of the search results making them more relevant. A ton of thanks to the team that keeps the wiki running. Regards .. Subbu

17 52

The Inbox: Morphic-dtl.1360.mcz
by commits＠source.squeak.org 08 May '18

08 May '18

David T. Lewis uploaded a new version of Morphic to project The Inbox: http://source.squeak.org/inbox/Morphic-dtl.1360.mcz ==================== Summary ==================== Name: Morphic-dtl.1360 Author: dtl Time: 11 November 2017, 4:04:34.153784 pm UUID: 0e1f3870-5f57-4dc6-a1e7-5d8ce68b71b5 Ancestors: Morphic-dtl.1359 World global elimination. Once the current project has been entered, Project current world == World. Begin eliminating references to the global variable World in cases where it is not required. =============== Diff against Morphic-dtl.1359 =============== Item was changed: ----- Method: ComplexProgressIndicator>>backgroundWorldDisplay (in category 'as yet unclassified') ----- backgroundWorldDisplay self flag: #bob. "really need a better way to do this" "World displayWorldSafely." "ugliness to try to track down a possible error" + [Project current world displayWorld] ifError: [ :a :b | - [World displayWorld] ifError: [ :a :b | | f | stageCompleted := 999. f := FileDirectory default fileNamed: 'bob.errors'. f nextPutAll: a printString,' ',b printString; cr; cr. f nextPutAll: 'worlds equal ',(formerWorld == World) printString; cr; cr. f nextPutAll: thisContext longStack; cr; cr. f nextPutAll: formerProcess suspendedContext longStack; cr; cr. f close. Beeper beep. ]. ! Item was changed: ----- Method: Debugger class>>morphicOpenInterrupt:onProcess: (in category '*Morphic-opening') ----- morphicOpenInterrupt: aString onProcess: interruptedProcess "Open a notifier in response to an interrupt. An interrupt occurs when the user types the interrupt key (cmd-. on Macs, ctrl-c or alt-. on other systems) or when the low-space watcher detects that memory is low." | debugger | <primitive: 19> "Simulation guard" debugger := self new. debugger process: interruptedProcess controller: nil context: interruptedProcess suspendedContext. debugger externalInterrupt: true. Preferences logDebuggerStackToFile ifTrue: [(aString includesSubstring: 'Space') & (aString includesSubstring: 'low') ifTrue: [Smalltalk logError: aString inContext: debugger interruptedContext to: 'LowSpaceDebug.log'] "logging disabled for 4.3 release, see http://lists.squeak.org/pipermail/squeak-dev/2011-December/162503.html" "ifFalse: [Smalltalk logSqueakError: aString inContext: debugger interruptedContext]"]. + Preferences eToyFriendly ifTrue: [Project current world stopRunningAll]. - Preferences eToyFriendly ifTrue: [World stopRunningAll]. ^debugger openNotifierContents: nil label: aString; yourself ! Item was changed: ----- Method: Morph>>updateFromResource (in category 'fileIn/out') ----- updateFromResource | pathName newMorph f | (pathName := self valueOfProperty: #resourceFilePath) ifNil: [^self]. (pathName asLowercase endsWith: '.morph') ifTrue: [newMorph := (FileStream readOnlyFileNamed: pathName) fileInObjectAndCode. (newMorph isMorph) ifFalse: [^self error: 'Resource not a single morph']] ifFalse: [f := Form fromFileNamed: pathName. f ifNil: [^self error: 'unrecognized image file format']. + newMorph := Project current world drawingClass withForm: f]. - newMorph := World drawingClass withForm: f]. newMorph setProperty: #resourceFilePath toValue: pathName. self owner replaceSubmorph: self by: newMorph! Item was changed: ----- Method: MorphHierarchyListMorph>>createContainer (in category 'private') ----- createContainer "Private - Create a container" | container | container := BorderedMorph new. + container extent: (Project current world extent * (1 / 4 @ (2 / 3))) rounded. - container extent: (World extent * (1 / 4 @ (2 / 3))) rounded. container layoutPolicy: TableLayout new. container hResizing: #rigid. container vResizing: #rigid. container setColor: (Color gray: 0.9) borderWidth: 1 borderColor: Color gray. container layoutInset: 0. "container useRoundedCorners." "" container setProperty: #morphHierarchy toValue: true. container setNameTo: 'Objects Hierarchy' translated. "" ^ container! Item was changed: ----- Method: MorphicProject>>storeSegment (in category 'file in/out') ----- storeSegment "Store my project out on the disk as an ImageSegment. Keep the outPointers in memory. Name it <project name>.seg. *** Caller must be holding (Project alInstances) to keep subprojects from going out. ***" | is sizeHint | + (Project current world == world) ifTrue: [^ false]. - (World == world) ifTrue: [^ false]. "self inform: 'Can''t send the current world out'." world isInMemory ifFalse: [^ false]. "already done" world ifNil: [^ false]. world presenter ifNil: [^ false]. ScrapBook default emptyScrapBook. World checkCurrentHandForObjectToPaste. world releaseSqueakPages. sizeHint := self projectParameters at: #segmentSize ifAbsent: [0]. is := ImageSegment copyFromRootsLocalFileFor: {world presenter. world} "world, and all Players" sizeHint: sizeHint. is state = #tooBig ifTrue: [^ false]. is segment size < 2000 ifTrue: ["debugging" Transcript show: self name, ' only ', is segment size printString, 'bytes in Segment.'; cr]. self projectParameters at: #segmentSize put: is segment size. is extract; writeToFile: self name. ^ true! Item was changed: ----- Method: MorphicProject>>storeSegmentNoFile (in category 'file in/out') ----- storeSegmentNoFile "For testing. Make an ImageSegment. Keep the outPointers in memory. Also useful if you want to enumerate the objects in the segment afterwards (allObjectsDo:)" | is | + (Project current world == world) ifTrue: [^ self]. " inform: 'Can''t send the current world out'." - (World == world) ifTrue: [^ self]. " inform: 'Can''t send the current world out'." world isInMemory ifFalse: [^ self]. "already done" world ifNil: [^ self]. world presenter ifNil: [^ self]. "Do this on project enter" World flapTabs do: [:ft | ft referent adaptToWorld: World]. "Hack to keep the Menu flap from pointing at my project" "Preferences setPreference: #useGlobalFlaps toValue: false." "Utilities globalFlapTabsIfAny do: [:aFlapTab | Utilities removeFlapTab: aFlapTab keepInList: false]. Utilities clobberFlapTabList. " "project world deleteAllFlapArtifacts." "self currentWorld deleteAllFlapArtifacts. " ScrapBook default emptyScrapBook. World checkCurrentHandForObjectToPaste2. is := ImageSegment copyFromRootsLocalFileFor: {world presenter. world} "world, and all Players" sizeHint: 0. is segment size < 800 ifTrue: ["debugging" Transcript show: self name, ' did not get enough objects'; cr. ^ Beeper beep]. is extract. "is instVarAt: 2 put: is segment clone." "different memory"! Item was changed: ----- Method: TheWorldMainDockingBar>>startMessageTally (in category 'menu actions') ----- startMessageTally + | world | + world := Project current world. (self confirm: 'MessageTally will start now, and stop when the cursor goes to the top of the screen') ifTrue: [MessageTally spyOn: + [[Sensor peekPosition y > 0] whileTrue: [world doOneCycle]]]! - [[Sensor peekPosition y > 0] whileTrue: [World doOneCycle]]]! Item was changed: ----- Method: TheWorldMenu>>startMessageTally (in category 'commands') ----- startMessageTally + | world | + world := Project current world. - (self confirm: 'MessageTally will start now, and stop when the cursor goes to the top of the screen') ifTrue: [MessageTally spyOn: + [[Sensor peekPosition y > 0] whileTrue: [world doOneCycle]]]! - [[Sensor peekPosition y > 0] whileTrue: [World doOneCycle]]]!

11 38

[ANN] Boadebui, a simple MIDI sequencer
by Stéphane Rollandin 08 May '18

08 May '18

Hello, I have made a simple and easy to use MIDI sequencer for loop-based music: http://www.zogotounga.net/comp/boadebui.htm Stef

4 4

Jump to page: