Squeak-dev March 2023

squeak-dev@lists.squeakfoundation.org

32 participants
124 discussions

How do I use MenuItemMorph>>#debugAction?
by Eduardo Ochs 04 Mar '23

04 Mar '23

Hi all, short version of my question: How do I use MenuItemMorph>>#debugAction? Now the long version. I was trying to use the debugger to understand how some "demo morphs" are created, starting by this one: World -> new morph... -> from alphabetical list -> L-N -> LineMorph I used the debug tool of the halo to copy the MenuItemMorph of that menu entry to a globally-acessible place, and then in a workspace I copied that to a variable local to that workspace called itemLineMorph. I was playing with that variable, and I did a "browse protocol" on it and I was reading its results. One of them, that looked especially interesting, was: MenuItemMorph >> #debugAction I tried this, itemLineMorph debugAction. and it opens the debugger - really neat! But if I tell the debugger to "proceed" it halts a few steps ahead, with this error: UndefinedObject(Object)>>doesNotUnderstand: #hand TheWorldMenu>>newMorphOfClass:event: [] in MenuItemMorph>>invokeWithEvent: CursorWithMask(Cursor)>>showWhile: MenuItemMorph>>invokeWithEvent: MenuItemMorph>>doButtonAction [] in MenuItemMorph>>debugAction [] in FullBlockClosure(BlockClosure)>>newProcess the problem is that in TheWorldMenu>>newMorphOfClass:event: the argument evt is nil... if I set evt by hand there to a keyboard event that holds a CR and then tell the debugger to proceed, everything works - I get a LineMorph on the screen. So: what is the _right_ way to run MenuItemMorph>>#debugAction? And what is the right way to write something similar to this, itemLineMorph debugAction. but that works out of the box? Thanks in advance! Eduardo Ochs http://anggtwu.net/eev-squeak.html

2 2

Very slow copy/ primitives
by karl ramberg 03 Mar '23

03 Mar '23

Hi Mines game got really sluggish on a old laptop with i3 processor. I tracked it down to this method where we copy protoTile instance variable MineBoard>>adjustTiles "reset tiles" | newSubmorphs count r c | submorphs do: "clear out all of the tiles." [:m | m privateOwner: nil]. newSubmorphs := OrderedCollection new. r := 0. c := 0. count := columns * rows. 1 to: count do: [:m | newSubmorphs add: (protoTile copy *"Change this to MinesTile new to speed up"* position: self position + (self protoTile extent * (c @ r)); actionSelector: #tileClickedAt:newSelection:modifier:; arguments: (Array with: (c+1) @ (r+1)); target: self; privateOwner: self). c := c + 1. c >= columns ifTrue: [c := 0. r := r + 1]]. submorphs := newSubmorphs asArray. Here is MessageTally: [ 67.5% {8826ms} MinesBoard>>adjustTiles [ 67.5% {8824ms} MinesTile(Morph)>>copy [ 67.5% {8824ms} MinesTile(Object)>>veryDeepCopy [ 66.6% {8713ms} DeepCopier>>fixDependents [ 66.6% {8713ms} WeakIdentityKeyDictionary(WeakKeyDictionary)>>associationsDo: [ 64.3% {8404ms} primitives [ 2.2% {289ms} [] DeepCopier>>fixDependents [ 1.7% {222ms} DependentsArray>>do: [ 1.1% {139ms} primitives Best, Karl

2 2

Where does the compiler handle the "local bindings" of workspaces?
by Eduardo Ochs 01 Mar '23

01 Mar '23

Hi list, first some context: I was trying to put some blocks of executable notes in something like the "Terse Guide to Squeak"... the Terse Guide is implemented as a class called TerseGuideHelp, and I was working on a modified copy of that class, that I called EdrxGuideHelp... but then I saw that the window that shows the content of a topic in a help browser treats local variables in a way that is different than the one in workspaces - if I run "a := 42" in the window with the contents of a topic I get a dialog window whose title is: Unknown variable: a! please correct, or cancel: Now a more basic question. I sort of know where the "local variables" of a workspace are stored... if I start with a single workspace and I run this on it, w := Workspace new. w buildAndOpen. and then I run this on the second workspace, a := 2. b := a + 3. and then go back to the first workspace and "print it" this, w bindings. I get: a Dictionary('a'->2 'b'->5 ) My first guess was that "do it"s and "print it"s in a workspace use a dictionary of bindings, but "do it"s and "print it"s in the mini-workspace of a help browser do not. I tried to understand "real" workspaces first, and after setting a breakpoint in TextEditor >> #printIt and doing lots of single-steppings I arrived at this method, TextEditor >> #evaluateSelectionAndDo: that has this: result := [ rcvr class evaluatorClass new evaluate: self selectionAsStream in: ctxt to: rcvr environment: (model environment ifNil: [Smalltalk globals]) notifying: self ifFail: [self flash. ^ nil] logged: true. ] on: OutOfScopeNotification do: [ :ex | ex resume: true]. but when that code calls this method Compiler >> #evaluate:in:to:environment:notifying:ifFail:logged: the "rcvr class evaluatorClass new" is simply a "Compiler new", "self selectionAsStream" is "a := 2. b := a + 3." as a stream, the ctxt and the rcvr are nil, and the environment is just "Smalltalk"... where does the dictionary of local bindings enter the story? Am I doing something wrong in the debugger?... Thanks in advance! Eduardo Ochs http://angg.twu.net/eev-squeak.html

2 10

hb := HelpBrowserB openOn: TerseGuideHelp. hb model. --> a HelpBrowser
by Eduardo Ochs 01 Mar '23

01 Mar '23

Hi list, earlier today, in this other thread, http://lists.squeakfoundation.org/pipermail/squeak-dev/2023-February/223584… I was trying to create a subclass of HelpBrowser, called HelpBrowserB, that would behave in a slightly modified way... to make things trivial to test, suppose that we create it like this: HelpBrowser subclass: #HelpBrowserB instanceVariableNames: 'bindings' classVariableNames: 'LookupPools' poolDictionaries: '' category: 'Category-Edrx' Then if we run this with "print it", hb := HelpBrowserB openOn: TerseGuideHelp. hb model. the result is "a HelpBrowser", not "a HelpBrowserB". Is that intentional or a bug? Thanks in advance! Eduardo Ochs http://anggtwu.net/eev-squeak.html

1 1