Squeak-dev January 2001

squeak-dev@lists.squeakfoundation.org

112 participants
318 discussions

Re: Tablets and Squeak
by Bert Freudenberg 31 Jan '01

31 Jan '01

On Wed, 31 Jan 2001, Ian Trudel wrote: > I'm using a little Wacom Graphire Tablet with Squeak time to time, > it works really great. I had lot of fun with the handwritting thing > (though I don't really know how to use it, I've been doing by tries > and failures). Actually, the Windows driver looks to handle > everything, I haven't really notice differences than using a mouse. So I take it you're just using a pen as mouse replacement? I think what Wiebe was refering to is the tablet primitive support - here you get pressure and tilt and all nice Pen stuff. Check out Pen simplePressurePen (well, you *need* the tablet prims for this to work...) -- Bert

2 1

Apple Events in Squeak
by Splice31415927＠aol.com 31 Jan '01

31 Jan '01

Hi, Can anyone point me to sample code for sending and receiving apple events with Squeak? I've found classes like AEDesc and DescType, but I can't find any implementations of them that demonstrate their use. Help greatly appreciated, thanks. Randy

2 1

Young ImageSegments
by ajh18＠cornell.edu 31 Jan '01

31 Jan '01

I'm writing an experimental distributed computing package for Squeak. When I send a message to a remote object, I want to send all "free" arguments along with it (free arguments are those that have no other pointers to them). Thankfully, ImageSegment extracts exactly what I want, however, it scans the entire image, both tenured and young objects, taking half a second or more. This is great performance for extracting whole Projects, but for my small and frequent Message objects this is a problem. Since my Message objects and its arguments are usually young, I would like another primitive method that only scanned young space and moved any pointers into tenured space into the outPointers array. Ted Kaehler is the author of Interpreter>>primitiveStoreImageSegment. To change it into a new primitiveStoreYoungImageSegment, I think I only have to get rid of youngStart _ self startOfMemory. "process all of memory" and add and: [fieldOop >= youngStart] to the internal pointer test ((self longAt: fieldOop) bitAnd: MarkBit) = 0. Can any GC/ImageSegment experts out there confirm this? Do you think it will be much faster? Thanks, Anthony Hannan P.S. I will post my Distributed Squeak package to this list when it is ready, hopefully in a month or two.

1 0

Re: newbie. morphic is sluggish on my PC
by Doug Way 31 Jan '01

31 Jan '01

Bob Arning wrote: > > On Tue, 30 Jan 2001 14:53:10 +0500 "Lex Spoon" <lex(a)cc.gatech.edu> wrote: > >... > >It's probably the same bug, a #< where #<= would be appropriate: > > > > wait < MinActivityLapse "big waits happen after a snapshot" > > ifTrue: [DisplayScreen checkForNewScreenSize. > > For those having the cpu hogging behavior in MVC, try changing the #< to a #<=. I have added an update to that effect and it sould appear in a matter of days, but I don't know how much it will help. Seems to work well for Windows, at least. (I'll also check my Mac at home.) The screen size updates are still working, too. - Doug Way dway(a)riskmetrics.com

1 0

Re: newbie. morphic is sluggish on my PC
by Jeff Szuhay 30 Jan '01

30 Jan '01

> I found this on a Mac which doesn't have quite the same processor >monitoring facilities Have you tried Peek-a-Boo? <http://www.clarkwoodsoftware.com> 8-) -- ************************************************** Jeff Szuhay A randomly-directed www.szuhay.org chaotical wetware pattern jeff(a)szuhay.org recognizer/generator. <this space for rent>

1 0

[QUESTION?] Squeak & SQL databases
by jchludzinski 30 Jan '01

30 Jan '01

I've seen postings concerning Squeak and MySQL (courtesy of a port from VisualAge). How well does this work? Where/how can I find out more about this? Has anything been done to support Oracle? If so, what? Thanks in advance, John

1 0

re: [enh] celestetweaks
by Lex Spoon 30 Jan '01

30 Jan '01

> > 1. The custom menu button always presents a menu; matching filters are > > listed at the top of the the menu. > That sounds good, but your changeset puts them under the fixed filter > management options. The tiny CS below reorders that to be as advertised, > which I for one would prefer. Well, okay, whatever makes people happy. I must admit it makes it very fast to select that top filter now! > > 2. The auto-move button is removed. Instead, the regular TOC menu will > > give the option of an automove. > These two features you mentioned are not equivalent - note what (still, > with your changes applied) happens when you press alt-m. The automove I > put in automatches the appropriate filter and then *moves all* matchers > to the folder of same name. Of course, whether this is a good thing or > bad is a matter of taste. I'm happy as it is, still having the keyboard > shortcut. If people find that too dangerous an option to have (and it > might be, to have hidden only in keyboard shortcut), I don't > particularly mind having my own customization changeset. > Ah, okay. So this is similar to doing a custom filter and then a "move all", except that it all happens automagically with one button press? I know I got messed up the first time I pressed this, because I mistook it for the "custom filter" button. Sadly, the message I'd selected disappeared from sight and I still don't know where it went to. Hopefully it wasn't an important message. (and hopefully it was only one!) It still seems good to leave this off of the main button bar, I think. New users will get stubbed by it, and ultra-efficient users will be using hotkeys anyway. > > 3. Huge message lists are not shown, but a menu item is added to > > forcibly show all messages. > This introduces a bit of redundancy with the existing folder list's menu > item "view all messages"... I'd keep any one of them. Crud, you're right. Never noticed that.... > > > 4. The From Filter is switched to a Participant Filter."! > Hey, cool! It's much more useful this way. Cool! -Lex

1 0

Morphic performance on handhelds
by Kevin Fisher 30 Jan '01

30 Jan '01

Hi: This is a very subjective question...to anyone that's using Squeak on your handheld of choice, I ask: How responsive does Morphic feel? Although I'm still struggling with an unoptimized VM on my iPaq, I find Morphic to be a bit...sluggish. It's not horrible by any means... but it is a bit slower than I was hoping. Dragging morphs particularly slow (due to the drop shadow perhaps?). Stuff like character recognition seems a bit slow as well, but it's certainly usable (just not as fast as my Palm IIIc). I haven't had any time yet to look at potential bottlenecks. I'm just curious, is all...I'm still looking in to getting the build chain fixed on handhelds.org (I've already brought it up on their mailing list).

7 8

Re: Tablets and Squeak
by Bert Freudenberg 30 Jan '01

30 Jan '01

On Tue, 30 Jan 2001, Wiebe Baron wrote: > Hi, > Is anyone using tablets with Squeak? I am, and Alan K. mentioned it's used at SqC, too. > I'm considering WACOM tablets (as a complete substitute for the > mouse) I'm using a Wacom Intuos GD and it's *great*. > and would like to know what kind of (windows) support Squeak > currently has, and if there are any mayor issues that should be > considered. AFAIK there is no tablet support for Windows yet. The implementation should not be too hard, though. > Bert had mentioned 100% CPU loads in a post several moths ago when he > was adding Unix support. Really? Can't remember. It works fine now. I implemented a TabletHandMorph which is controlled by the tablet prims, and fixed the software cursors so you can actually have multiple of them. And then there is a subclass of TabletHandMorph called TabletPenMorph that displays a virtual pen with shadow instead of the cursor, including the correct tilting - immensely cool :) > Also, the windows 2.8 VM has a Joystick/Tablet plug-in - is this based > on the WinTab API? It's just a stub implementation - no tablet support at all. -- Bert

1 0

Re: [RFI] autogenerated type info in class comments
by Lex Spoon 30 Jan '01

30 Jan '01

Doug Way <dway(a)riskmetrics.com> wrote: > Type collecting is nice because it can run quickly, whereas a thorough type inferencer can be pretty slow. The downside to type collecting is that sometimes there will be no instances of a given class in the image to collect from. (Although this is often not a problem because as a general rule you're usually editing classes which you're actually running in the image.) I was thinking about adding Dan Ingalls' simple type inferencer to Whisker as a backup method for when the type collector can't find any instances. Also, I think Lex Spoon is working on a more thorough type inferencer... I'm not sure what the status of that is. Anyway, this is sort of a digression from your post. > Yeah. The problem is that my inferencer gets bogged way down in too many cases. Another observation on the difference between what Whisker does and what tools like mine do is that Whisker's is a *lower* bound on the types, and mine gives *upper* bounds. I believe Dan's just makes a complete guess, based on how the code seems to use the variable, so it's not known to be an upper or lower bound. It sounds like the guesses are very good, though. So, a good combination for people who want results *right now*, would probably be a Whisker-style query, followed by a Dan-style query if there are no current instances of the class. -Lex

1 0

Jump to page: