Squeak-dev June 2001

squeak-dev@lists.squeakfoundation.org

122 participants
362 discussions

Color accuracy
by Karl Ramberg 29 Jun '01

29 Jun '01

Why is color internally represented with 1023 possible values for each red green or blue, when a color at 32 bit screen depth is limited to 255 for each? That's 4 times the precision needed. Karl

1 0

RE: Icons
by Norton, Chris 29 Jun '01

29 Jun '01

Duane Maxwell wrote: "I think there are a few changesets lurking around that add XPM reading to Squeak. I know I released one a while back, and was told that somebody had written a better one before me." You can find Chris Reuter's and Duane Maxwell's XPM code here: http://swiki.gsug.org/sqfixes/ Go to the bottom of the page, click on "Show all messages", then search the page for XPMReadWriter. Cheers, ---==> Chris

2 1

Re: Cassowary Demo Question
by Ross Boylan 29 Jun '01

29 Jun '01

On Fri, Jun 29, 2001 at 07:59:26AM -0400, Joshua 'Schwa' Gargus wrote: > Hi Ross, > > On Fri, Jun 29, 2001 at 01:35:06AM -0700, Ross Boylan wrote: > > I hate to be dense, but what do the demos do? > > The demos are broken. Apparently, the demo morphs are not well-behaved > morphs. When they are stored into a project, they don't properly store > their Cassowary solver state. I'm looking into this right now... > > > Or where does one find what > > the demos do? I see if I click on things stuff happens, though it seems to > > be mostly one way--things disappear after awhile. > > Try "ClCrossedQuadDemo new openInWorld", etc. This will give you fresh, > well-behaved demos. Yes, that works much better! With that and your pointer to the right method, I have a fighting chance.

1 0

Icons
by Ross Boylan 29 Jun '01

29 Jun '01

If I want to make a little icon, what's the best way to do it? Can it be done with the internal painting tool? I have a hazy recollection of seeing a recommendation to use an external paint program and then importing the result (into one of the image or button morphs?). While poking around I found the forms dictionary in Scripting System. Can anyone tell me if the code that populates it is in the system, or are we left only with the artifact? (I found some code for a few items, but not most). Any recommended sources of icons on the net? Thanks.

4 4

flash
by rhian lewis 29 Jun '01

29 Jun '01

?Does any one have any examples of using the Balloon MMFlash code, or notes relating to it. If so I would really appreciate having a look at them. at

1 0

InterpreterSimulator
by Anthony Hannan 29 Jun '01

29 Jun '01

Attached are a few changes I made to get the InterpreterSimulator to open a mini image. It works but it won't accept keyboard input, only mouse input. Anyone know how to fix this?

1 0

New Squeak Swiki page color feature
by Jochen F. Rick 29 Jun '01

29 Jun '01

Hi, I've been trying to follow the list, but a paper deadline is making me largely unavailable. Also, I will be going on vacation on July 11th, so I will largely be unavailable for a month after that. Now here's the good things: minnow is now a linux machine. Expect stability and speed. Yeah!!! I changed the look on the site. This time, I changed it so that it looks more compatible with the squeak.org site. In the last year or so, the Squeak web presence has started to look more professional. This new look is to go along with that trend. Also, there is a new feature. You can change the color of the page (or rather the color of the button banner). Go to a page, press the edit button, and scroll to the bottom of the page to check it out. Peace and Luck! Je77

1 0

New Block Closures, New Compiled Methods, and System Tracer
by Anthony Hannan 29 Jun '01

29 Jun '01

I saw on the squeak swiki two Squeak Central projects: http://minnow.cc.gatech.edu/squeak/BlockClosures, and http://minnow.cc.gatech.edu/squeak/NewCompiledMethodFormat. Is there any code for these, particularly any SystemTracer code. I already have most of the compiler and interpreter re-written using the new CompiledMethod format and BlockClosures, but I can't test it until I convert an image using SystemTracer. I ran the basic SystemTracer that comes with the latest image. It finishes creating an image file, but when I start it up the VM crashes (Exception code: C0000005) in Windows 98. Has anyone created a working clone lately? Tim Rowledge <tim(a)sumeru.stanford.edu> is widely believed to have written: > I hope you've remembered to make it possible to manipulate the closures > from the image side. It should remain possible to alter the return > context, the receiver etc. I have. > Don't forget to modify the exception handling related prims > to be able to scan up your frames rather than the Contexts they > currently chase. I'm working on this.

2 1

Re: [BUG] SyntaxMorph testClass:andMethod:
by Steve Wart 29 Jun '01

29 Jun '01

Hi Bob, Sorry if I was a bit unclear with this message, I was a bit tired when I wrote it. 1. The confusing bit is how to resize the scriptor. I read the class comments on SyntaxMorph and couldn't get past the first step. After spending about and hour and consulting with someone who was using it with 3.0 we were able to track down the change and get the command to work with the earlier version of the method. Since this method is only used from StringHolder>>openSyntaxView, a hypothesis was put forth that this change made it easier to drop SyntaxMorphs into books for interactive demos. 2. Very cool. Thanks 3. The 'nasty' method is testClass:andMethod: -- in the sense that the original window-based method was pretty short and easy to understand. The new version of the method has what appear to be UI parameters hard-coded into it, and it's not clear why the authors made these changes. Cheers, Steve Bob Arning <arning(a)charm.net> wrote: > Steve, > > First, it is possible (although certainly not ideal) to resize the scriptor you get in this fashion. If you move the scriptor a bit away from the top left corner, then cmd-click just to the left of the blue square that is visible inside the scriptor, you will get the halo. Click until the label on the halo is "TwoWayScrollPane", then drag the yellow handle and the scriptor will resize. > > Second, there is a newer way to do this that you may find more useful - from the menu icon at the top of the browser, turn ON the "tiles" option. Then you will see the tiles right in the browser. > > Third, I'm not clear on which method it is that you think is confusing and nasty. > > Cheers, > Bob > > On Thu, 28 Jun 2001 21:47:28 -0800 Steve Wart <thecows(a)home.com> wrote: > >If you click on a method in a browser and select 'tile scriptor' the > >SyntaxMorph comes up in a little window in the upper left hand corner of > >the screen. Can't resize, pretty confusing. > > > >If you revert to the 3.0 version of this method, everything is fine. In > >fact the 3.0 version of the method is much simpler. It is sending > >#openInWindow to the SyntaxMorph whereas the 3.1 version creates an > >AlignmentMorph and sends it #openInWorld. > > > >The 3.1 version of the method is quite confusing and nasty (IMHO), in > >addition to the fact that it doesn't seem to work correctly. I would > >suggest that the first version is 'right', or at least that some > >refactoring is in order.

1 0

Re: Fwd: BankAccount Tutorial
by Dan Ingalls 29 Jun '01

29 Jun '01

Hello Michael - This is the result of the Alignment rewrite last fall. A bunch of protocol has changed, and we don't even use alignment morphs any more, although they *should* still work. >>I have been working through the tutorial but when I try to "accept" the >>following code, >> >>historyMorph >> |bars m| >> bars_history collect: [:v| Morph new extent: 5@v]. >> m_AlignmentMorph newRow centering: #bottomRight. >> m addAllMorphs: bars. >> ^m >> >>Squeak tells me that "centering:" is an unknown selector. Of course, when I >>told it to confirm and then tried to do a "printIt" in the workspace of the >>following, I get an error. I believe I know enough about Squeak that I >>shouldn't have had to ask it to confirm "centering:" Well, I disagree that you "shouldn't" have been asked to confirm. What it is saying is "You will get an error if you try to execute this (so you might want to stop now and fix it), but if you intend later to define #centering:, then you may want to proceed as is. >>, and just because I did >>confirm it, doesn't mean that their is an object to communicate with, am I >>right? There is an object, but it does not understand #centering:. Note that, if you really knew what you were doing, you could have confirmed and then, when you got the error, you could have defined #centering:, and then proceeded, and all would have worked. This is why you are allowed the option of proceeding. >>Is the tutorial wrong, or is my version of Squeak (3.0) inappropriate, or >have I made some error in copying it? The tutorial is out of date because the alignment rewrite did not preserve the existing protocols. Here is how I figured out the problem to answer your dilemma: 1. I confirmed your experience. 2. I knew there must be some new protocol similar to #centering:, so I selected bbottomRight, and used cmd-N for find all references to it (all seNders for that message. 3. This did not find anything I hoped for (users of a new centering message), except that it located the wonderfully revealing method, convertOldAlignmentsNov2000: varDict using: smartRefStrm which makes it quite clear that the old #centering: message has been replaced by a new #wrapCentering: message. 4. When I made that change to the above code from the tutorial, it did something reasonable (I haven't tested the rest of the tutorial). If you wish to know more about the new alignment protocol, you might look through the TableLayouts project on Bob's Superswiki, which documents a lot of the changes that were made. Hope this helps. - Dan --

2 1

← Newer
1
2
3
4
5
6
7
...
37
Older →

Jump to page: