Squeak-dev June 2002

squeak-dev@lists.squeakfoundation.org

173 participants
382 discussions

Re: The Voodoo of Inlined Methods
by Richard A. O'Keefe 30 Jun '02

30 Jun '02

My question: What's right? Squeak does not (yet) do type inference, so it *can't* make inlining decisions based on the type of the receiver. If it sees a to: b do: [:each | ... ] and the expression a is not some kind of literal, it hasn't a clue whether a is a number or a numbat. [A numbat is a small Australian marsupial carnivore.] (*) By the way, the intended ``upLeft to: bottomRight do: aBlock'' iterates over the obvious rectangular grid consisting of all points p: upLeft <= p <= bottomRight (with step 1 in each component and <= in the sense defined by Point). For the time being, I've resigned myself to using another selector. It's not _that_ obvious. >From the argument order I guess that up...bottom and left..right are the orders for the dimensions, but is it row major, column major, or some kind of diagonal? What does it do when the coordinates are floating-point numbers?

4 4

can current line be changed in debugger?
by Rob Whitfield 30 Jun '02

30 Jun '02

Is there a way to change the current line of execution within the debugger? For example, can I jump back a few lines to re-execute a section of code?

4 3

[Win32 VM] window placement oddity
by Ross Boylan 29 Jun '02

29 Jun '02

I recently started using the latest Win32 VM, and have noticed something strange. When I start up squeak, the resulting window frame is partly off the top of the screen. The lower edge is also slightly raised (i.e., there's space between the lower edge of the window and the top of the task bar). The window I am talking about here is the MS-Windows window that surrounds all of squeak, not one of the internal windows. I'm frankly surprised that any program is able to do this. I run the program with the window maximized under Windows 2000. Smalltalk vmVersion 'Squeak3.2gamma of 12 January 2002 [latest update: #4879]' The latest update in my image is 4889. P.S. It's also a little surprising that 12 jan 2002 is the date reported. I assume it just hasn't been bumped while the code was in development. squeak.exe is dated 5/26/02.

1 0

Dynapad R0.1 Released
by Aaron 29 Jun '02

29 Jun '02

Hello all, A new Dynapad release today. Biggest change is a migration to Squeak 2.8 (from 3.0). Also, a new look, broken PIM apps, and a new function-plotter. This is still very much alpha or pre-alpha! But it's useful to me even at this point. Download: http://dynapad.swiki.net/8 Screenshots: http://dynapad.swiki.net/7 Let me know what you all think- Thanks and Regards, Aaron

9 10

problems with OpenGL hardware acceleration on linux
by Felix Franz 29 Jun '02

29 Jun '02

Hi all, after installing the nvidia Accelerated Linux drivers (and watching a fast teapot :) I tried Balloon-3d in squeak: the performance is the same with or without "enable hardware acceleration" in the WonderlandCameraMorph-menu. I rebuilt the vm (I am using Ians 3.2.2 release). ldd said it linked to the right libGL (the one provided by the nvidia-driver). B3DAcceleratorPlugin is listed in "Smalltalk listBuiltinModules" (don't know if this is enough, it is the first time I had problems with a plugin). I tried " B3DHardwareEngine primitiveSetVerboseLevel: 5" which should print some infos, but I see no debugging-information. Now I don't know what to do ... Is anyone enjoying a hardware accelerated 3d-squeak and can give me some advice? If you need more information: just ask, I'll be happy to provide them. Thanks in advance, felix

2 2

Re: MenuMorph callback
by Bob Arning 29 Jun '02

29 Jun '02

Curiously, this is related to a recent discussion about #perform:orSendTo:. What' shappening is that MenuMorph>>setInvokingView: is manipulating your menu entry... "Re-work every menu item of the form <target> perform: <selector> to the form <target> perform: <selector> orSendTo: <invokingView>. This supports MVC's vectoring of non-model messages to the editPane." So if MyClass does not implement #perform:orSendTo:, then it will get the default behavior which is to pass the message to the invoking view (the PluggableListMorph in this case). You have several choices: - implement MyClass>>perform:orSendTo: to handle the message - use a DumberMenuMorph instead of MenuMorph and this problem is avoided Cheers, Bob On Sat, 29 Jun 2002 00:20:19 +0200 Bergel Alexandre <bergel(a)iam.unibe.ch> wrote: >I do not understand, I have defined : >-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=3D-=-= >MyClass>>getMenu: aMenu > aMenu defaultTarget: self. > aMenu > add: 'showdependancies' > target: self > selector: #showDependancies. > ^ aMenu >-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=3D-=-= > >And when I press the item in my menu, I have a message not understood #showDependancies for the PluggableListMorph related. >Of course, I have defined the method MyClass>>showDependancies > >If anyone has a suggestion...

1 0

MenuMorph callback
by Bergel Alexandre 29 Jun '02

29 Jun '02

Hello, I do not understand, I have defined : -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= MyClass>>getMenu: aMenu aMenu defaultTarget: self. aMenu add: 'showdependancies' target: self selector: #showDependancies. ^ aMenu -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= And when I press the item in my menu, I have a message not understood #showDependancies for the PluggableListMorph related. Of course, I have defined the method MyClass>>showDependancies If anyone has a suggestion... Regards, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._ Bergel Alexandre http://www.iam.unibe.ch/~bergel ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^

1 0

3.2, Help wanted. Now we're getting somewhere.
by John Voiklis 28 Jun '02

28 Jun '02

I would like to offer my help. One major bug (or, more accurately, one thing that bugs me) in version 3.2 and all versions of Squeak, is that, after a year of working with Squeak, I still don't know how to use 95% of the built in Morphs. Forget about how those Morphs are developed or how to customize them (this mailing list is actually great about that), I don't know how to even use them--what are they for and how do they work (e.g. how does one use all those chat tools). Let me back up a bit and lavish all of you with some praise. I love Squeak: I love the interface (it's much more in tune with how I and most of my colleagues work); I love that it functions like a guest OS (everything can be done in squeak); I love that Squeak can be user-programmed to be anything that user wants. In fact, I have based much of my recent professional work on Squeak: as a Web developer, I promote the use of Swiki and Seaside; and as the head of an educational technology NPO, I have made the Squeak Plugin the basis for all our projects. I have no complaints about how Squeak looks or functions. What I and, quite likely, the entire user population needs is documentation, i.e. a user manual. I agree, as many on the list have argued, that the rate at which Squeak changes may make the writing of a user manual a fools errand. While that may hold true for how various Morphs are implemented and extended, that is rarely true about what they are for and how they function, especially for those Morphs in the Morph Catalog. I find it important enough to have such a manual that I am willing to volunteer a life-time's worth of free time to compile and edit a Squeak User Manual. I envision something that both goes beyond yet is more basic than the current opening screen tutorials and bubble help: something akin to the Squeakland tutorials, though more prosaic. An illustrated catalog describing what each morph is, what problems it solves, it's basic functionality (version 1 of the manual should probably be limited to the Morph Catalog). The manual could be kept in a flap (preferably default), and, to assure that the manual is obvious to users, it could interject itself every time a Morph is instantiated--whether from the supplies/widgets flap, or from the Morph catalog, or from the World Menu, or, even, from the workspace. What do I mean by interject: every time a Morph is instantiated the User Manual flap opens and scrolls to the correct description. To keep this from annoying those who don't need or who outgrow such a help feature, I imagine a help on/of button in the Navigator bar. If the community and SqueakCentral agrees with such a project, this is how I see the writing/editing process working: COMPILATION: 1) Pick a Morph that you either authored or mastered and write a description detailing what the morph is, what problems it solves, it's basic functionality, and instructions on how to use it. 2) Send it to me. 3) I will try those instructions on the latest iterations of Squeak and, where applicable, the SqueakPlugin (I have access to boxes running Windows 2000, Red Hat Linux 7.1, and Mac OS9) EDITING: 1) I will write back reporting on my success/failure and appending any suggestions or requests for clarification. 2) You write back with clarifications. 3) I retest and, if necessary, resubmit requests for clarification 4) I compile the descriptions; clean up, correct, and create consistency in language; layout descriptions with some "pretty" graphics. 5) I post the User Manual for peer review and QA. 6) I correct as necessary. PUBLISHING: 1) I hand it over to Squeak Central for the coding of the pop-up functionality and for inclusion in the next build. As to my qualifications: I have earned a BA in English and MFA in Poetry. I have worked as an editor (at OUP) and freelance writer (a bunch of places). I have worked as a Web developer for 6+ years during which I have cultivated varying levels of expertise with a variety of technologies (Java, SQL, Python, and recently Squeak/Smalltalk). I spent much of my time during those 6+ years translating between the tech savvy and the tech hostile. In addition, I co-authored and edited the specification for the Open eBook Publication Structure (www.openebook.org), the XML format readable by Microsoft Reader, among others. I hope people will agree that this proposal addresses a dire need in the progress and evangelism of Squeak, and that we can begin work immediately. Thank you, John --------------------- John Voiklis, Executive Director Red Figure Foundation See our project site at: http://www.madeleinelengle.org/

3 2

Re: [newbie][q] How to create a new module?
by danielv＠netvision.net.il 28 Jun '02

28 Jun '02

Ok, that won't be a problem. The base level answer is - Module frmoPath: #(People LB BusCommander) forceCreate: true The meta level answer is read the relevant Swiki pages, like goran said, they describe many of the basic actions, like creating a new module. BTW, what's a BusCommander? Have fun, Daniel Leo Burd <leoburd(a)media.mit.edu> wrote: > - Hello everyone! > > I'm trying to port my code into a new module called > People-LB-BusCommander. The system is saying that I'm not allowed to > create a new top level module... What shall I do? > > Best, > > Leo

1 0

Help with T-Gen
by Jose HG 28 Jun '02

28 Jun '02

I am trying to install T-Gen, I do the following steps: 1. File in tgen.14Nov1237pm.cs into a 2.6 or later image. 2. Copy the files exampleN.* to your squeak directory. 3. To verify installation, open a Transcript window and in a workspace evaluate. TranslatorGenerator runAllTests But the result in transcript is: T-gen TEST SUITE (answer 'yes' to all prompters) START: T-gen LL Parser Generation Test 1 STOP: T-gen LL Parser Generation Test 1 failed. START: T-gen LL Parser Generation Test 2 and suceed this error: 'Error: A class named TgenTemporaryTestScanner already exists.' I haven't idea why? Thanks, Jose L. _________________________________________________________________ Hable con sus amigos en línea, pruebe MSN Messenger: http://messenger.msn.es

2 1

← Newer
1
2
3
4
5
6
7
...
39
Older →

Jump to page: