Squeak-dev June 2013

squeak-dev@lists.squeakfoundation.org

45 participants
215 discussions

Re: [squeak-dev] Re: Squeak on tablets
by JohnReed Maffeo 23 Jun '13

23 Jun '13

> ----- Original Message ----- > From: OpenSlate ChalkDust > Sent: 06/17/13 07:34 PM > To: The general-purpose Squeak developers list > Subject: [squeak-dev] Re: Squeak on tablets > > A couple more focused questions. > > 1. How does the CogDroid vm compare with the Windows and OS X versions? > Will it run the same image? I am running CogDroid on an inexpensive Next Book 7 inch tablet. I can run current versions of Squeak, Pharo, and Cuis. For development, I can run the image from the tablet in USB mode if I want, or I can download images from my computer. The only actual development work I have done is to make minor modifications to the SameGame program. It runs well and looks like a tablet game. DR Geo is a program that has been ported to run on Android using CogDroid, but I don't know any more than that. I am pretty sure that gestures have not been implemented. > > 2. How well does the Samsung Galaxy S-pen work in place of a three button > mouse? > > > On Mon, Jun 17, 2013 at 11:42 AM, OpenSlate ChalkDust < > openslateproj(a)gmail.com> wrote: > > > As some of you know I spent quite a bit of time getting nowhere trying to > > get Squeak running on top of FreeBSD in a tablet form factor.The situation > > has changed considerably, in that the world has gone tablet crazy. > > Typically using fingers rather than a mouse or stylus. > > > > I would appreciate any recommendations on what tablets work well with > > squeak, and what to avoid. What has been done to adapt the three-button > > mouse squeak expects to have to a fingertip input? Then of course there is > > keyboard input in general, including keyboard shortcuts. > > > > I'd be happy to collect the results in a table on Google Drive. > > > > -- > > Gary Dunn > > Open Slate Project > > http://openslate.org/ > > > > > > -- > Gary Dun > Open Slate Project > http://openslate.org/ jrm

5 4

Decompiler reordering temps
by Frank Shearar 23 Jun '13

23 Jun '13

I happened to have Metacello loaded when I ran the DecompilerTests. A method (MetacelloCommonMCSpecLoader >> #retryingResolvePackageSpecReferences:gofer:) has these locals declared: | references repositoryString retryCount repositoryError | and when decompiled the list looks like this: | references retryCount repositoryError repositoryString | frank

2 2

No subject
by Azka Niazi 23 Jun '13

23 Jun '13

Treat me like a beginner. I have two targets:1. I want to document the source code of Scratch.2. I want to modify Scratch. Below the #'s is what I have already been told through this mailing list. Now, the source code provided for Scratch [1] has two files with code inside them; a .SOURCES file and a .CHANGES file. I have two simple questions:What is the purpose of the two files?If I want to change the code in one of these file and see it how does it come out, (how) can that be done? I have been able to understand the format of the source code provided. I would like to ask is there is any code which is responsible for creating the OO world of Scratch. The system categories for Scratch objects hold a lot of classes in them. Those classes, along with a lot of other new and old classes have been used to construct Scratch. These classes didn't preexist. Can a file of code not be used to create new classes and system categories? Squeak by Example [8] talks about a 'hierarchy browser' which is easy to find with the files provided with it. Is there a similar kind of thing to explore Scratch objects? If yes, how to get to it. If no, how do can I trace the hierarchy of inheritance objects used to make Scratch. Thanks,Azka ##################################Scratch uses the Morphic UI on a Squeak image with a virtual machine underneath. The Squeak version is, I think, 2.8 (as opposed to 4.4 which is about to come out) and came out in 2005. The important difference is that the virtual machine has changed. Modern vms are closure complete and have a JIT. To see Squeak 2.8 when it was released. [7]For some videos on Scratch from a Squeak perspective [3-6].An introductory book on Squeak. [8]Source code on the virtual machines. [2].The Scratch image that will not work on a modern Squeak virtual machine. [1]But will work on a virtual machine from a Squeakland Etoys release. [10]As you first expressed interest with how Scratch interfaces with the hardware you may need to join the VM-Beginners list [9]. [1] http://info.scratch.mit.edu/Scratch_Source_Code_Licensed_Code[2] http://squeakvm.org/index.html[3] http://www.youtube.com/watch?v=wqV3nGD9N7U[4] http://www.youtube.com/watch?v=XwmQ1DoEjsA[5] http://www.youtube.com/watch?v=uGyfzw_gePo[6] http://www.youtube.com/watch?v=YmTk9t71jZ4[7] http://ftp.squeak.org/2.8/[8] http://squeakbyexample.org/[9] http://lists.squeakfoundation.org/pipermail/vm-beginners[10] http://www.squeakland.org/ Courtesy of Chris Cunnington; when I said I need to understand Scratch (get about its source code) and how does it interface with the hardware.

5 5

Understanding Smalltalk for Modifying Scratch (Squeak version)
by Darius 23 Jun '13

23 Jun '13

Hi Azka, Thank you for helping us improve Scratch. :) Perhaps I can guide you into the Smalltalk way of thinking and how different it is from the usual application development paradigm and workflow. Imagine you're a Java developer and you use Eclipse for your IDE. Imagine you keep your source code in a SVN or similar repository which runs on a database server. Imagine your development team keeps your entire setup ... Eclips, SVN, repository, application data, operating system, and configurations, in an Virtual Operating System ram-memory-space-image-in-a-file which VirtualBox, or VMWare can run ... so there is no "installation" to get one's environment running. The VirtualBox ram-memory-space-image-in-a-file has a "snapshot" to a single disk file of all the applications installed an running. In the setup described above, one simple takes an VirtualBox snapshot ram-memory-space-image-in-a-file and starts it with VirtualBox. This has the Eclipse IDE, the application running in Eclipse, and the SVN server running all at the same time from the moment the VirtualBox image snapshot was taken. One just continues to make changes to the code in Eclipse and the running application in Eclipse immediately reflects those changes. In (almost) all Smalltalk distributions, the Smalltalk development environment has everything mentioned above, and works like everything mentioned above. The squeak.exe application is like the VirtualBox application which runs snapshot saved ram-memory-space-image-in-a-file. The squeak.image file is like the ram-memory-space-image-in-a-file which has Eclipse and the developer's application all running all the time. The .source file is like the SVN repository in a database. You'd never, never edit the SVN source code database file directly, even if in theory you could. It'd ruin the versioning and indexing among other problems. The .changes file is like the SVN repository database journaling file. It's just there for emergency recovery reasons. Not to be edited directly either. Making snapshots of the ram-memory-space-image-in-a-file is like selecting "save" or "save as" in the Squeak World menu to create a new .image file. Once the squeak.exe (in Windows, or the Linux, or OS X equivalent virtual machine app) is running the .image file, one uses the World Menu (like clicking on the Smalltalk desktop background to bring up the menu, this might be disabled in the scratch version) one can select the Browser menu option to start the Hierarchal Browser (Smalltalk's Eclipse IDE equivalent running in the app). Edit the code in the Hierarchal Browser with each method save, automatically saves it in the .changes file. Everything else automatically gets save when the .image is saved from within the Smalltalk running environment. This means you can't use your favorite text editor to edit Smalltalk code. One must use the built-in Hierarchal Browser. However, all the source code for running the Hierarchal Browser is in the Hierarchal Browser itself, so you can edit the behavior of the IDE any way you like to add your favorite IDE features ... in real time. It's all one running app. There's no separate programs for the IDE or viewing graphics or editing text or playing music or animating a game. It's all one app, running in one memory space. So, every "application" can see every other "applications" classes, instances, and field values. A Vulcan mind meld of all applications into each other. That's Smalltalk. Does this help? Cheers, Darius _________________ On Fri, Jun 21, 2013 at 4:57 AM, Azka Niazi <azka.akn(a)live.com> wrote: Thanks for the feedback Ron and Chris, but I have some more questions now. 1. I am sorry I didn't fully understand what did you mean by "act as storage space for the code you write in Smalltalk". Also please specify whether by Smalltalk you meant Smalltalk-80 (i.e. Squeak) or did you mean Smalltalk only. As in is there a custom to refer to Squeak as Smalltalk. Doesn't it create am

1 0

No Subject
by Chris Cunnington 23 Jun '13

23 Jun '13

I wouldn't worry about what the changes and sources files do right now. You want to know how to open the image and look at the code. These videos will help with that. [1] You need to go to Squeakland [2] and get yourself an application. You can open the Scratch image from the Scratch website with that. Finally, there is a beginners list [3] you can ask questions at. This list isn't the best for this kind of question. Once you learn some basics of Smalltalk, you should be able to explore the source code of Scratch. Chris [1] http://smalltalk.org/videos/video_AbsoluteBeginner_DownloadOpen.html [2] http://www.squeakland.org [3] http://lists.squeakfoundation.org/mailman/listinfo/beginners

2 1

Daily Commit Log
by commits＠source.squeak.org 23 Jun '13

23 Jun '13

Changes to Trunk (http://source.squeak.org/trunk.html) in the last 24 hours: http://lists.squeakfoundation.org/pipermail/packages/2013-June/006066.html Name: Files-tpr.123 Ancestors: Files-fbs.122 Fix bug reported in mantis ttp://bugs.squeak.org/view.php?id=7740 FileTestUrl should now pass ============================================= http://lists.squeakfoundation.org/pipermail/packages/2013-June/006067.html Name: Tests-fbs.223 Ancestors: Tests-fbs.222 Test for reported Decompiler bug http://bugs.squeak.org/view.php?id=7771 Marked as an expected failure because we don't know when/if we'll fix this edge case. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2013-June/006068.html Name: Protocols-fbs.41 Ancestors: Protocols-fbs.40 Typo. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2013-June/006069.html Name: Tools-fbs.474 Ancestors: Tools-fbs.473 Convert PointerFinder to use ToolBuilder. #openInMorphic preserves the old behaviour in case I've messed up. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2013-June/006070.html Name: SUnit-fbs.95 Ancestors: SUnit-fbs.94 Actually display meaningful information in the event of #shouldnt:raise: failing. =============================================

1 0

The Trunk: SUnit-fbs.95.mcz
by commits＠source.squeak.org 22 Jun '13

22 Jun '13

Frank Shearar uploaded a new version of SUnit to project The Trunk: http://source.squeak.org/trunk/SUnit-fbs.95.mcz ==================== Summary ==================== Name: SUnit-fbs.95 Author: fbs Time: 22 June 2013, 9:59:35.152 pm UUID: ab8a9b79-cb02-4d4f-854d-0a940e0f6e20 Ancestors: SUnit-fbs.94 Actually display meaningful information in the event of #shouldnt:raise: failing. =============== Diff against SUnit-fbs.94 =============== Item was added: + ----- Method: TestCase>>fail: (in category 'extensions') ----- + fail: aString + + ^self assert: false description: aString.! Item was changed: ----- Method: TestCase>>shouldnt:raise: (in category 'accessing') ----- shouldnt: aBlock raise: anExceptionalEvent + ^ [ aBlock value ] + on: anExceptionalEvent + do: [:e | self fail: 'Block raised ', e className, ': ', e messageText].! - ^self assert: (self executeShould: aBlock inScopeOf: anExceptionalEvent) not - !

1 0

Re: [squeak-dev] The Trunk: Tools-fbs.474.mcz
by Frank Shearar 22 Jun '13

22 Jun '13

On 22 June 2013 21:32, <commits(a)source.squeak.org> wrote: > Frank Shearar uploaded a new version of Tools to project The Trunk: > http://source.squeak.org/trunk/Tools-fbs.474.mcz > > ==================== Summary ==================== > > Name: Tools-fbs.474 > Author: fbs > Time: 22 June 2013, 9:33:27.224 pm > UUID: 6711307b-55fe-45d4-ada0-a1ff7aeda855 > Ancestors: Tools-fbs.473 > > Convert PointerFinder to use ToolBuilder. #openInMorphic preserves the old behaviour in case I've messed up. > > =============== Diff against Tools-fbs.473 =============== Please try this guy out right away, so I can remove #openInMorphic and cut one more thread in the Tools -> Morphic dependency. Thanks! frank

1 0

The Trunk: Tools-fbs.474.mcz
by commits＠source.squeak.org 22 Jun '13

22 Jun '13

Frank Shearar uploaded a new version of Tools to project The Trunk: http://source.squeak.org/trunk/Tools-fbs.474.mcz ==================== Summary ==================== Name: Tools-fbs.474 Author: fbs Time: 22 June 2013, 9:33:27.224 pm UUID: 6711307b-55fe-45d4-ada0-a1ff7aeda855 Ancestors: Tools-fbs.473 Convert PointerFinder to use ToolBuilder. #openInMorphic preserves the old behaviour in case I've messed up. =============== Diff against Tools-fbs.473 =============== Item was changed: + ----- Method: PointerFinder>>arrowKey:from: (in category 'pointer-list') ----- - ----- Method: PointerFinder>>arrowKey:from: (in category 'morphic ui') ----- arrowKey: key from: aController key = $i ifTrue: [^ self inspectObject]. ^ super arrowKey: key from: aController! Item was added: + ----- Method: PointerFinder>>buildWith: (in category 'ui-building') ----- + buildWith: builder + | listSpec windowSpec | + windowSpec := builder pluggableWindowSpec new + model: self; + label: 'Pointer Finder'. + + listSpec := builder pluggableListSpec new + model: self; + frame: (0@0 corner: 1@1); + list: #pointerList; + getIndex: #pointerListIndex; + setIndex: #pointerListIndex:; + menu: #menu:shifted:; + keyPress: #arrowKey:from:; + doubleClick: #inspectObject; + yourself. + + windowSpec children: {listSpec}. + + ^ builder build: windowSpec.! Item was changed: + ----- Method: PointerFinder>>initialExtent (in category 'ui-building') ----- - ----- Method: PointerFinder>>initialExtent (in category 'morphic ui') ----- initialExtent ^ 300 @ 300! Item was changed: + ----- Method: PointerFinder>>inspectObject (in category 'pointer-list') ----- - ----- Method: PointerFinder>>inspectObject (in category 'morphic ui') ----- inspectObject pointerListIndex = 0 ifTrue: [^ Beeper beep]. (objectList at: pointerListIndex) inspect! Item was changed: + ----- Method: PointerFinder>>menu:shifted: (in category 'pointer-list') ----- - ----- Method: PointerFinder>>menu:shifted: (in category 'morphic ui') ----- menu: aMenu shifted: shifted + ^ aMenu + addList: #( + ('Inspect (i)' #inspectObject 'Live long and prosper!!') + - + ('Search again' #searchAgain 'Search again\for the same object' withCRs)); + yourself.! - ^ MenuMorph new - defaultTarget: self; - add: 'Inspect (i)' action: #inspectObject; - balloonTextForLastItem: 'Live long and prosper!!'; - addLine; - add: 'Search again' action: #searchAgain; - balloonTextForLastItem: 'Search again\for the same object' withCRs; - yourself! Item was changed: + ----- Method: PointerFinder>>open (in category 'ui-building') ----- - ----- Method: PointerFinder>>open (in category 'morphic ui') ----- open + ^ ToolBuilder open: self.! - | window list | - window := (SystemWindow labelled: 'Pointer Finder') - model: self. - list := PluggableListMorph new - doubleClickSelector: #inspectObject; - - on: self - list: #pointerList - selected: #pointerListIndex - changeSelected: #pointerListIndex: - menu: #menu:shifted: - keystroke: #arrowKey:from:. - "For doubleClick to work best disable autoDeselect" - list autoDeselect: false. - window addMorph: list frame: (0@0 extent: 1@1). - list color: Color lightMagenta. - window openInWorld! Item was added: + ----- Method: PointerFinder>>openInMorphic (in category 'morphic ui') ----- + openInMorphic + | window list | + window := (SystemWindow labelled: 'Pointer Finder') + model: self. + list := PluggableListMorph new + doubleClickSelector: #inspectObject; + + on: self + list: #pointerList + selected: #pointerListIndex + changeSelected: #pointerListIndex: + menu: #menu:shifted: + keystroke: #arrowKey:from:. + "For doubleClick to work best disable autoDeselect" + list autoDeselect: false. + window addMorph: list frame: (0@0 extent: 1@1). + list color: Color lightMagenta. + window openInWorld! Item was changed: + ----- Method: PointerFinder>>perform:orSendTo: (in category 'message handling') ----- - ----- Method: PointerFinder>>perform:orSendTo: (in category 'morphic ui') ----- perform: selector orSendTo: otherTarget selector == #inspectObject ifTrue: [^ self inspectObject]. selector == #searchAgain ifTrue: [^ self searchAgain]. ^ super perform: selector orSendTo: otherTarget! Item was changed: + ----- Method: PointerFinder>>pointerList (in category 'pointer-list') ----- - ----- Method: PointerFinder>>pointerList (in category 'morphic ui') ----- pointerList ^ pointerList asArray! Item was changed: + ----- Method: PointerFinder>>pointerListIndex (in category 'pointer-list') ----- - ----- Method: PointerFinder>>pointerListIndex (in category 'morphic ui') ----- pointerListIndex ^ pointerListIndex ifNil: [0]! Item was changed: + ----- Method: PointerFinder>>pointerListIndex: (in category 'pointer-list') ----- - ----- Method: PointerFinder>>pointerListIndex: (in category 'morphic ui') ----- pointerListIndex: anInteger pointerListIndex := anInteger. self changed: #pointerListIndex! Item was changed: + ----- Method: PointerFinder>>searchAgain (in category 'pointer-list') ----- - ----- Method: PointerFinder>>searchAgain (in category 'morphic ui') ----- searchAgain self pointerListIndex: 0. self search. self changed: #pointerList!

1 0

The Trunk: Protocols-fbs.41.mcz
by commits＠source.squeak.org 22 Jun '13

22 Jun '13

Frank Shearar uploaded a new version of Protocols to project The Trunk: http://source.squeak.org/trunk/Protocols-fbs.41.mcz ==================== Summary ==================== Name: Protocols-fbs.41 Author: fbs Time: 22 June 2013, 9:28:53.268 pm UUID: f5b44683-d40a-4e61-a21f-86b0677ca194 Ancestors: Protocols-fbs.40 Typo. =============== Diff against Protocols-fbs.40 =============== Item was changed: Object subclass: #ObjectWithDocumentation instanceVariableNames: 'authoringStamp properties elementSymbol naturalLanguageTranslations' classVariableNames: '' poolDictionaries: '' category: 'Protocols-Kernel'! + !ObjectWithDocumentation commentStamp: 'fbs 6/20/2013 17:25' prior: 0! - !ObjectWithDocumentation commentStamp: '<historical>' prior: 0! ObjectWithDocumentation - an abstract superclass for objects that allows maintenance of an authoring stamp, a body of documentation, and a properties dictionary. + The Properties implementation has not happened yet -- it would closely mirror the implementation of properties in the MorphExtension, for example.! - The Properties implementation has not happened yet -- it would closely mirror the implemenation of properties in the MorphExtension, for example.!

1 0

← Newer
1
...
5
6
7
8
9
10
11
...
22
Older →

Jump to page: