Squeak-dev February 2001

squeak-dev@lists.squeakfoundation.org

144 participants
725 discussions

Re: Project layout
by Jeff Szuhay 27 Feb '01

27 Feb '01

>Actually, a pitch of about 150/inch and using some color selection >trickery will do the job pretty darn well. So a pretty nice display >would be about 1.9 million pixels (8" * 10.5"). Microsoft has a >beautiful 150 pitch XVGA display for their e-Book (so you get a >paperback-sized screen that is very readable). Aw, darn. this must be one of those prototype systems not yet available. And, again, aw darn. It had to be Microsoft. Sorry, but I've despised Microsoft since they charged for their bug fixes in DOS 3.x when IBM was honorably giving the patches away. (And I'll stop with that.) Nevertheless, if the machine gives me a "slate" that I can do what I want with, I really don't care who makes it (Apple, take heed). Jeff Sz. 8-) -- ************************************************** Jeff Szuhay A randomly-directed www.szuhay.org chaotical wetware pattern jeff(a)szuhay.org recognizer/generator. When the effective leader is finished with his work, the people say it happened naturally. -- Lao-Tzu, Chinese philosopher (6th century B.C.)

1 0

Re: RE: RE: [BUG] Low Space problem in 3.0 (and 3.1.3696)
by Bob Arning 27 Feb '01

27 Feb '01

On Mon, 26 Feb 2001 11:35:45 -0500 "Norton, Chris" <chrisn(a)Kronos.com> wrote: >One thing that VSE does that is really cool is that it returns a walkback >quickly (stack overflow) when you have an infinite loop. I don't have the >faintest idea how it does this, but I think Squeak could benefit greatly >from this. Chris, Now that you mention it, I have the vague memory of there being some limit to the depth of the stack. Which can keep you out of trouble, but also prevents you from doing interesting recursive stuff. I shouldn't be too hard to hack up a limit to stack depth, but I'm not sure how much of a blessing that would be. Cheers, Bob

1 0

Re: Project layout
by Alan Kay 27 Feb '01

27 Feb '01

Actually, a pitch of about 150/inch and using some color selection trickery will do the job pretty darn well. So a pretty nice display would be about 1.9 million pixels (8" * 10.5"). Microsoft has a beautiful 150 pitch XVGA display for their e-Book (so you get a paperback-sized screen that is very readable). Cheers, Alan ------- At 10:27 PM -0500 2/26/01, Lex Spoon wrote: >Karl Ramberg <karl.ramberg(a)chello.se> wrote: >> When a project is loaded in a different size that it was created >> it often crops stuff off or leave stuff in the upper left corner >> depending on size of the user screen. Is there a way to >> layout the project so it scales resonably on diffrent screen sizes? >> > >Switch to 2000x2000 displays and start drawing with centimeters instead >of pixels. You can get an idea of what this would be like by viewing >PDF. Whether you love or hate the way PDF files end up looking on your >machine is a good indicator of whether you think your hardware is enough >for this. If you have less than 4 million pixels, I'm guessing you'll >think it sucks, but one day... > >In the meantime, it's hard. :) > > >-Lex

1 0

Re: Hyperstone (RISC/DSP) port .. TCP/IP?
by Lex Spoon 27 Feb '01

27 Feb '01

Tom Poliquin <poliquin(a)softcomp.com> wrote: > > There would still be a load of integration work to do, not least ripping the > > kernel dependencies out of that code, but at least you wouldn't have to > > worry as much about fixing the stack in the face of bug reports --- like the > > recent CERT advisory describing DoS attacks on a couple of the TCP states. > > I did consider the BSD stack (it's the subject of the entire TCP/IP Vol > II) but > you came up with the same objection that I did .. extracting the kernal > 'stuff' > looked daunting .. I also reviewed the Linux stack code commentary book > in > which that extraction looked even worse .. Your comment will send me > back > to review it one more time .. > > Last I checked, "slirp" claimed to be using the BSD stack to handle its TCP/IP. And "slirp" doesn't just *use* TCP/IP, and does strange things to it as it passes by. So you might glance through slirp and see how they managed the extraction. -Lex

1 0

Re: [GOODIE] WatchWindow-nk
by Joshua Channing Gargus 27 Feb '01

27 Feb '01

Wow! I haven't used it yet, but I know I would have plenty of times in the past. Thanks Ned, Joshua On Mon, Feb 26, 2001 at 05:05:39PM -0800, Ned Konz wrote: > from preamble: > > "Change Set: WatchWindow-nk > Date: 26 February 2001 > Author: Ned Konz > > This adds a 'watch it' choice to the context menus in > text editors. This will take the selection, make a BlockContext > out of it, and give you a Morph that will display the value > of that block every half second. > > Only works in Morphic; results in an inspector being put up > for the resultant block under MVC (you can evaluate 'self value' > in that inspector if you want)."!

1 0

Re: Zooming interface -- my 5 cents worth
by Alan Kay 27 Feb '01

27 Feb '01

Here's my plan. I have to give a talk at ACM1 in San Jose on March 13th plus we are trying to get a bunch of the kids programming stuff on the net by then. On the 14th, Kristin Nygaard is giving a talk on the foundations of OOP at the Computer History Museum. I do the donation then and start the process for digitizing the tapes. I wil arrange with the CHM to take donations for this particular project through their web site. Cheers, Alan At 10:04 PM +0100 2/26/01, Cees de Groot wrote: >Norton, Chris <chrisn(a)Kronos.com> said: >>An on-line digital library of these important, historical references is an >>excellent suggestion and you have made exceedingly generous offer. Thank >>you, from all of us! >> >Well, Alan, I'd say get yourself a PayPal wallet - I smell donations for >this project coming up ;-) > >-- >Cees de Groot http://www.cdegroot.com <cg(a)cdegroot.com> >GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD 1986 F303 937F E098 9E8B

1 0

Usenet client for Squeak
by Strugatsky, Alex 27 Feb '01

27 Feb '01

Hi Squeakers, Does anyone work on Usenet client for Squeak or on NNTP protocol implementation?

2 1

Unidentified subject!
by Dan Shafer 27 Feb '01

27 Feb '01

I ran into a problem running a 3.0 image on the Mac. I have downloaded all updates as of 2/21. I'm working on a bunch of documentation stuff for Squeak as a labor of love. And I _do_ love this stuff! One entry point I looked at was from the main Swiki (http://minnow.cc.gatech.edu/squeak/795) where a counter example not unlike that included in my earlier Smalltalk books appears. I decided the familiarity of the project might facilitate my making quick progress. The directions for doing a direct-manipulation project on that page were self-described as incomplete and outdated. But I attempted to recreate them. (I apologize for the length of this email. In helping people sort out bugs over the years, I've always tried to provide as much detail as possible even when some of the details didn't seem particularly relevant.) I placed an Alignment Morph from the morphs menu, resized it and changed its color. Then I looked for an LedMorph. It doesn't appear to be included in the standard morphs menu, so I rummaged about and found one in the SameGame morph. I duplicated it and dragged the copy into the Alignment Morph, and embedded it in the Pasteup Morph containing the Alignment Morph. (QUERY - Is there a reason the LedMorph isn't on the "Add a new morph" menu?) Next, I added a SimpleButtonMorph from the morphs menu, duped it twice, and edited the labels on the three buttons. I didn't like the default alignment I was getting so I turned off alignment (i.e., selected "no alignment") so I could move them at will. In the course of that step, I believe, the LedMorph wandered outside the bounds of the Alignment morph. I suspect that's where my trouble began. I rearranged the buttons and when I tried to return the LedMorph to its place inside the Alignment morph, I couldn't get a halo on it. I could get one on an individual Led digit but not on the Led morph. I tried several times, figuring I was just hitting the wrong target with the cursor. Suddenly, the menu I was trying to work with on the Led digit morphs didn't close after I selected an option. Instead, the option stayed highlighted in a kind of strange yellow color and the menu stayed up. opened another, selected the same choice, with the same effect. So I figure now I'm in trouble, so I click on the desktop to bring up the world menu so I can quit without saving and start over. To my surprise and chagrin, clicking on the desktop produced, not the world menu, but a menu I've never seen before designed to help me find a window. So now I have no way of getting out of Squeak cleanly. I experimented a bit and found I couldn't really do anything. So I quit from the Mac system menu, was not asked about saving, and relaunched the image. The Counter project has disappeared, but I still cannot get a world menu; clicking on the desktop (or what I've come to think of as the desktop at least) in my opening Squeak window produces the menu for finding a window. I quit and re-launched twice to no avail. By now it's 1 a.m. so I shut down for the night and I get up this morning to write this email. In the midst of it, I go back to the 3.0 image and click on the desktop to confirm that I still can't get a world menu. And, again to my surprise, things seem to be working again. Maybe this was a low-memory condition. I did restart my machine between those sessions (last night and this morning). But I thought I'd report it anyway in part because it's the first major sign of instability I've seen in 3.0. __________________________________________________ Do You Yahoo!? Yahoo! Auctions - Buy the things you want at great prices! http://auctions.yahoo.com/

7 8

Re: [ENH][VM][UNIX]generalized async IO
by Tim Rowledge 27 Feb '01

27 Feb '01

Bert Freudenberg <bert(a)isg.cs.uni-magdeburg.de> is widely believed to have written: > On Sun, 28 Jan 2001, Tim Rowledge wrote: > > > > While you're tidying up, you might like to remove the calls to > > joystickInit(), serialInit() etc in sqXWindow.c - they're not needed at > > all with plugins since those functions are called by the plugin-local > > init code.If you leave them in and build a vm with the appropriate > > plugins external you get problems. Hmm, I wrote this on the 28th of JANUARY and it's just shown up? Wierd... > > Exactly. I removed these too, for my Tablet support. Although I did not > yet test with the tablet module built external ... Same sort of crime as the security plugin being directly referenced from the file & sockets plugins... And... "Lex Spoon" <lex(a)cc.gatech.edu> is widely believed to have written: > Tim Rowledge <tim(a)sumeru.stanford.edu> wrote: > Hrm, I don't understand. What functions need different numbers > of arguments? > Last time I looked the sockets needed two arguments (microseconds and a spare file descriptor) and sound needed none. tim -- Tim Rowledge, tim(a)sumeru.stanford.edu, http://sumeru.stanford.edu/tim Useful random insult:- Guillotining him would make only an aesthetic difference.

1 0

emacs and squeak, again...
by Phil Mitchell 26 Feb '01

26 Feb '01

I have noticed several threads in the archives where people ask about integrating emacs into Squeak. The standard answer seems to be: why would you want to do that when the built-in editor is so nice? As a pre-newbie thinking about starting up in squeak, here's one reason: I'm looking for a platform where I can develop generic info management tools. The ability to plug in an editor with the power of emacs and the extensive package library of emacs is almost a sine qua non... So I'm wondering... how possible would this be to do? Does it even make sense to think about doing it?

5 7

← Newer
1
2
3
4
5
6
7
8
9
...
73
Older →

Jump to page: