Squeak-dev May 2004

squeak-dev@lists.squeakfoundation.org

152 participants
288 discussions

components versus classes
by Richard Staehli 04 May '04

04 May '04

We have prototyped a simple component architecture in Squeak (see http:/www.simula.no:8888/QuA), but the current mechanism for loading component "blueprints" has problems. Here is a motivating scenario: 1) Alice implements a class Chess for chess games and serializes this code to a "blueprint" that is stored with the name "ChessBlueprint". 2) Alice advertises "ChessBlueprint" as a blueprint for a factory (class) for chess games. 3) Bob needs a chess game and loads a chess game factory from "ChessBlueprint". Here's my problem. I use the following code to serialize aClass: | serialization | serialization := ReadWriteStream on: String new. serialization header; timeStamp. aClass fileOutOn: serialization moveSource: false toFile: 0. serialization trailer; reset. Then I rewrite the code to replace the class name with a special token 'QuATemplate'. This allows us to fileIn the code without the risk that the class name will conflict with other user code. When Bob loads the blueprint, the following code recreates the class: (ReadStream on: blueprint) fileInAnnouncing: 'loading ',aBlueprintName printString. then the resulting 'QuATemplate' class is renamed to a unique name generated for the current image. This mechanism is too complex, wastes time with rewriting and renaming, fills up the change set with lots of useless changes, and sometimes results in obsolete classes hanging around. Anyone care to suggest a better approach, particularly if it is easy to implement in our prototype? Since our component architecture must keep track of which components are active and support binding, we don't need Smalltalk to allocate a global name at all. Ideally, we could create a read a class from a stream just like any other object, without assuming that it is a change to the code in the image. Richard Staehli

2 1

FFI question
by liste.squeak 03 May '04

03 May '04

Hi, I am working on a robotics simulation written in C++. I would like to link thios simulation to squeak to experiment some artificial intelligence algorithms on it. I have written a C++ shared library and when I attempt to use it in squeak with FFI, I have the classical "Error : can't find function adress". Do you know whether it is possible to use C++ librarries with FFI in the same way that I can use C libraries ? Thank you. Julien Bourdon, trainee in a french computer science laboratory. Accédez au courrier électronique de La Poste : www.laposte.net ; 3615 LAPOSTENET (0,34/mn) ; tél : 08 92 68 13 50 (0,34/mn)

3 3

Copying file creation and modification time
by Noury Bouraqadi 03 May '04

03 May '04

Hi, I would like to make a copy of a file in another directory while keeping its creation time and modification time unchanged. Any hint? Thanx, Noury -- ------------------------------------------ Dr. Noury Bouraqadi - Enseignant/Chercheur Ecole des Mines de Douai - Dept. G.I.P http://csl.ensm-douai.fr/noury European Smalltalk Users Group http://www.esug.org Squeak: an Open Source Smalltalk http://www.squeak.org ------------------------------------------

3 2

Re:FFI question
by .: Julien BOURDON :. 03 May '04

03 May '04

> Hi, > > I am working on a robotics simulation written in C++. I would > like to link thios simulation to squeak to experiment some > artificial intelligence algorithms on it. I have written a C++ > shared library and when I attempt to use it in squeak with > FFI, I have the classical "Error : can't find function adress". > > Do you know whether it is possible to use C++ librarries with > FFI in the same way that I can use C libraries ? > > Thank you. > > Julien Bourdon, trainee in a french computer science laboratory. > I've heard of an old project "panda3d" (panda3d.org) that used to use squeak and FFI. Does anyone have the sources or know where I can get them ? Julien Bourdon, trainee in a french computer science laboratory Accédez au courrier électronique de La Poste : www.laposte.net ; 3615 LAPOSTENET (0,34/mn) ; tél : 08 92 68 13 50 (0,34/mn)

1 0

[ANN] Preview of Squeak interface to wxWidgets
by Rob Gayvert 03 May '04

03 May '04

Greetings, I have posted a snapshot of my efforts to blend Squeak with the wxWidgets GUI library. Downloads are available at http://homepage.mac.com/rgayvert/wxsqueak.html. Notes: - Currently Win32 only - Wrapper methods are implemented only for core framework and demo classes - Uses version 2.5.1 of the wxWidgets library - Supports (in theory) most of the wxPython API - Uses synchronous callbacks to Squeak via interpret() - Built using a single internal plugin - Most of the plugin methods are generated by parsing the wxWidgets docs - Uses a modified VM (minor tweaks to ioProcessEvents() and interpret() are needed to provide synchronous callbacks to wx events). Any comments or suggestions are appreciated. .. Rob

5 7

[ANN] OperaORB-native on SM
by Masashi Umezawa 03 May '04

03 May '04

Hi, I've recently updated OperaORB, an lightweight-ORB. It has now a 'native'-mode. It means OperaORB runs on RMT (not SOAP). Since RMT is a Squeak specific messaging layer, it is not portable. But the strong point is its performance. So, if you seek a fast ORB, please take a look at it. http://www.mars.dti.ne.jp/~umejava/smalltalk/soapOpera/operaORB.html Cheers, --- [:masashi | ^umezawa]

1 0

[Q][BUG] 5905 double initialize ?
by Lic. Edgar J. De Cleene 03 May '04

03 May '04

I have this weirdness and wish know someone have it Class method: start super new initialize "This execute twice instance initialize method" If I change to start super new start "This execute only once" Can someone check this or explain ? Never see this before 5905. Edgar

2 2

Developer initials on SM
by goran.krampe＠bluefish.se 03 May '04

03 May '04

Hi people! Just wanted to mention that a few people have elected to enter pretty "long" developer initials on SM. I haven't thought much about it - but looking back in history we have tried to keep the initials short. First choice would probably be lower case, two letters. Second choice another extra letter etc. These initials show up in the tools here and there and perhaps it is a good idea to try to keep them short. Given SM we can of course gradually make the tools "SM aware" and have them showing full names if the map is available etc. Thoughts? regards, Göran

1 0

[BUG]Array(Object)>>doesNotUnderstand: #markExteriorEdges
by tim＠sumeru.stanford.edu 02 May '04

02 May '04

PoohTestMorph new openInWorld Opens a morph - drag the mouse around a little an release mouse button. boom. TWO notifiers appear, first one for this stack, second one when the display safely fails to display safly... 2 May 2004 12:17:26 pm VM: RiscOS - a SmalltalkImage Image: Squeak3.7beta [latest update: #5878] SecurityManager state: Restricted: false FileAccess: true SocketAccess: true Working Dir ADFS::HardDisc4/$/Squeak/sq31 Trusted Dir ADFS::HardDisc4/$/Squeak/sq31/secure Untrusted Dir ADFS::HardDisc4/$/Squeak/sq31/untrusted Array(Object)>>doesNotUnderstand: #markExteriorEdges Receiver: #(222.0(a)253.0 233.0(a)253.0 245.0(a)253.0 258.0(a)253.0 270.0(a)252.0 280.0(a)249.0 288.0(a)245.0 294.0(a)240.0 297.0(a)233.0 300.0(a)224.0 3...etc... Arguments and temporary variables: aMessage: markExteriorEdges Receiver's instance variables: #(222.0(a)253.0 233.0(a)253.0 245.0(a)253.0 258.0(a)253.0 270.0(a)252.0 280.0(a)249.0 288.0(a)245.0 294.0(a)240.0 297.0(a)233.0 300.0(a)224.0 3...etc... PoohTestMorph>>triangulate Receiver: a PoohTestMorph(4002) Arguments and temporary variables: Receiver's instance variables: bounds: 393@277 corner: 793@677 owner: a PasteUpMorph<page>(1248) submorphs: #() fullBounds: 393@277 corner: 793@677 color: Color white extension: a MorphExtension (3088) [other: (borderStyle -> a SimpleBorder) (errorOnDraw -> true)]...etc... borderWidth: 2 borderColor: Color black points: #(213.0(a)252.0 213.0(a)252.0 222.0(a)253.0 222.0(a)253.0 222.0(a)253.0 233.0(a)253.0 233.0(a)253.0 233.0(a)253.0 245.0...etc... subdivision: #(222.0(a)253.0 233.0(a)253.0 245.0(a)253.0 258.0(a)253.0 270.0(a)252.0 280.0(a)249.0 288.0(a)245.0 294.0@240.0...etc... lastPoint: 177@117 firstPoly: nil time: nil [] in PoohTestMorph>>subdivide {[points := self simplify: points. points := self smoothen: points length: 1...]} Arguments and temporary variables: Time class>>millisecondsToRun: Receiver: Time Arguments and temporary variables: timedBlock: [] in PoohTestMorph>>subdivide {[points := self simplify: points. points := self smoothen: points length: 1...]}...etc... initialMilliseconds: 47714203 Receiver's instance variables: superclass: Magnitude methodDict: a MethodDictionary(#<->a CompiledMethod (1795) #=->a CompiledMethod...etc... format: 134 instanceVariables: #('seconds' 'nanos') organization: ('ansi protocol' < = duration hash hour hour12 hour24 meridianAbb...etc... subclasses: nil name: #Time classPool: nil sharedPools: an OrderedCollection(ChronologyConstants) environment: a SystemDictionary(lots of globals) category: nil --- The full stack --- Array(Object)>>doesNotUnderstand: #markExteriorEdges PoohTestMorph>>triangulate [] in PoohTestMorph>>subdivide {[points := self simplify: points. points := self smoothen: points length: 1...]} Time class>>millisecondsToRun: - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - BlockContext>>timeToRun PoohTestMorph>>subdivide PoohTestMorph>>mouseUp: PoohTestMorph(Morph)>>handleMouseUp: MouseButtonEvent>>sentTo: PoohTestMorph(Morph)>>handleEvent: PoohTestMorph(Morph)>>handleFocusEvent: [] in HandMorph>>sendFocusEvent:to:clear: {[ActiveHand := self. ActiveEvent := anEvent. result := focusHolder han...]} [] in PasteUpMorph>>becomeActiveDuring: {[aBlock value]} BlockContext>>on:do: PasteUpMorph>>becomeActiveDuring: HandMorph>>sendFocusEvent:to:clear: HandMorph>>sendEvent:focus:clear: HandMorph>>sendMouseEvent: HandMorph>>handleEvent: HandMorph>>processEvents [] in WorldState>>doOneCycleNowFor: {[:h | ActiveHand := h. h processEvents. capturingGesture := capturingGest...]} Array(SequenceableCollection)>>do: WorldState>>handsDo: WorldState>>doOneCycleNowFor: WorldState>>doOneCycleFor: PasteUpMorph>>doOneCycle [] in Project class>>spawnNewProcess {[[World doOneCycle. Processor yield. false] whileFalse. nil]} [] in BlockContext>>newProcess {[self value. Processor terminateActive]}

1 0

[Q[ What is the correct VM for OS X 10.3.3 ?
by Lic. Edgar J. De Cleene 02 May '04

02 May '04

John: This morning I have a meeting with local Apple Dealer Mario Verardi and try to get Squeak working in latest 10.3.3 without success. I remember something you say about a problem in this version and what we must send mail to Apple. What I don't know is if this was fixed by you , what the correct VM is or what steps we must follow to have Squeak working in 10.3.3. Very, very thanks in advance. Edgar

4 5

Jump to page: