Squeak-dev May 2001

squeak-dev@lists.squeakfoundation.org

143 participants
597 discussions

RE: Q: Is the time stamp for an morphic event set by OS?
by Peter Crowther 02 May '01

02 May '01

> From: Dan Ingalls [mailto:Dan.Ingalls@disney.com] > On Macs, I believe mouse events are polled by the hardware > every 16ms, and this will limit your resolution as Mark says. Quick calculation: 2400 bits/sec (typical mouse, I think), 10 bits/byte inc. framing, 3 bytes/sample = 60 samples/sec or >16 ms between samples, limited by the bit rate. I wrote a driver for a tablet that used the same protocol, so I'm pretty certain of these values. - Peter

1 0

Re: latest and greatest iPAQ Squeak?
by Kevin Fisher 02 May '01

02 May '01

On Tue, May 01, 2001 at 09:45:46PM -0700, Tim Rowledge wrote: [snip] > e) logged out of gnome (good grief what piece of garbage _that_ is) > after much hassle WindowMaker all the way, baby! :) [snip] > > Anyway, the webpage is > <URL:http://www.netwinder.org/~scottb/notes/FP-Notes.html> and I hope > one of you might have bit more luck. Why nobody on the NW > team/newsgroups could have pointed us to this when I first reported the > problem months ago, I don't know. Assuming of course that it actually > has anything whatsoever to do with the problem. > Wow, thanks Tim! This certainly is interesting...I wonder how it will affect performance? Do you know if this affects the infamous optimization bug at all? There still doesn't seem to be any solution to the optimization problem...the build machines I use have been upgraded to GCC 2.95.3 and the bug is more insidious on this version of GCC (it doesn't fail to assemble as before...it compiles happily and then segfaults when run). > tim > -- > Tim Rowledge, tim(a)sumeru.stanford.edu, http://sumeru.stanford.edu/tim > Any given program will expand to fill available memory. >

1 0

Identifying hardware Hand
by Jesse Welton 02 May '01

02 May '01

Is there a reliable method for identifying whether a given Hand is following the hardware Cursor? I've identified two possibilities: HandMorph>>needsToBeDrawn can only be false for the hardware Hand, but this can give a false negative (on a return value of true). It might work to test 'aHand class == HandMorph', since there are HandMorph subclasses for all other Hands I'm aware of, but I'm not sure. Is a direct instance of HandMorph guaranteed to be following the hardware Cursor? Or is there yet some other possibility which I have missed? -Jesse

1 0

Re: What happened to the conversion method prototypes?
by Jesse Welton 02 May '01

02 May '01

Ned Konz wrote: > > On Tuesday 01 May 2001 11:15, Jesse Welton wrote: > > > The old conversion methods have been completely supplanted by > > #convertToCurrentVersion:refStream:. Comments explaining the change > > are available in SmartRefStream>>aComment. Unfortunately, there are > > still a few references the the old-style conversion mechanism, like > > those hanging around in the ChangeSorter. > > OK, I wrote it. However, it doesn't seem to be used for changing existing > instances in an image (I'd like for a file-in of my change set to edit > existing instances). I've removed two instance variables and added one; the > value of the new one depends on the two removed ones. > > How do people do this? I thought that maybe I could collect all the > appropriate instances (with copies of their soon-to-be-removed instVars) in > the preamble along with their required data, then fix them in the postscript. > Does this make sense? Yes, this makes sense. When I was doing layout changes I just did the conversion in the postscript, but I wasn't removing any instance variables with information I needed. Besides collecting them and caching their data in the preamble, you might consider making the change in two steps: First, add the new instance variable and fill it from the soon-to-be-deleted instance variables. Second, remove the old instance variables. (Presumably, this is how things actually progressed as you were making the changes in your image.) This could either be in two change sets, or (if I understand things correctly, though I've never tried it) in one hand-modified change set. If you can get the preamble-based version to work, that may be the more easily manageable option. A two-step change would be safer if there's any chance of instances being created during the file-in. -Jesse

1 0

[ANN] Lightweight Classifications
by Roel Wuyts 02 May '01

02 May '01

We recently implemented a lightweight classification model that allows to classify any kind of objects (items) in generic 'bags' (classifications). The actual idea stems from Koen De Hondt, who implemented it in VW and constructed a very powerful browser on top. However, Stéphane Ducasse and I wanted something much simpler, so we revised the model and the browser for simplicity and came up with *lightweight classifications*. We also implemented it in VW, and put a browser on top to show some of the functionality of the model. Of course we know that the current implementation is in VisualWorks. However, the idea is to port the model to other Smalltalk dialects as well (we or anybody interested). Therefore, if you think a model or a tool like this is useful or interesting, or if you want to co-operate or port it: let us know something! If you are interested in more information take a look at the website at http://www.iam.unibe.ch/~wuyts/lightweightClassifications/. There is detailed information and links with more explanations of the features, some screenshots to give you a feeling of the browser and the possibility to download. Of course, any feedback or ports are welcome :-) -- Roel Wuyts Programming Technology Lab rwuyts(a)vub.ac.be Vrije Universiteit Brussel http://prog.vub.ac.be/~rwuyts Webmaster of European Smalltalk User Group: www.esug.org

1 0

Subclassing OrderedSequence
by Richard A. O'Keefe 02 May '01

02 May '01

I have a subclass of OrderedCollection, which adds some instance variables. Poking around in the collection classes, I find that sometimes self copyEmpty is used to make an empty collection, and sometimes self species new: 0. The basic problem is that if you have a subclass of OrderedCollection (or any other SequenceableCollection) that adds instance variables, *sometimes* a selection or slicing method will return an empty result that has its additional instance variables appropriately set up, and simetimes it will return a non-empty result that _doesn't_ have its additional isntance variables appropriately set up. The problem is that using self species new: (size) to return a copy of part of a sequence doesn't provide information about the additional instance variables. I was wondering about a refactoring where there would be a new method copySansContentsForSize: size "Subclasses of existing sequenceable collections that add instance variables should override this to set up those variables in the new object." ^self species new: size and having the methods that now call self species new: newSize call self copySansContentsForSize: newSize instead. Before doing this, is there any agreed way to decide which selection/ filtering/slicing methods should return OrderdCollections and which should return objects of the new subclass?

3 2

[BUG] BookMorph Viewer Scripting Menu
by John Hinsley 02 May '01

02 May '01

SuSE 6.4 Squeak 3.1alpha update 3960 Current Change Set 3880 HaltUpdate.tk Ok, so I'm re-doing the Rolodex Tutorial. Or, at least, re-doing the gifs, which means I have to re-write it. I want to tweak the Search function of BookMorph so I open up the viewer. (I'm looking for "Book do menu item"). I note, first, that the viewer's top menu looks more like a menu, less like an unfurled Roman scroll. I click on the menu and note that IIRC the menu has changed. Completely. No big deal, I'll just click on "basic" ans ask for "miscellaneous" from the menu. Immediately the debugger pops up "MessageNotUnderstood: AllMorphsAndBookPagesInto:" (hey, I only asked for a menu!). And "Book do menu item" is nowhere to be found. This seems wierd. I could probably (and it looks like I'll have to!) find another way of doing what I wanted, but an update which *reduces* functionality? Cheers John -- ****************************************************************************** Marx: "Why do Anarchists only drink herbal tea?" Proudhon: "Because all proper tea is theft." ******************************************************************************

2 4

Re: Problem with TextMorph
by Raymond Asselin 02 May '01

02 May '01

Le Mardi 1 mai 2001, à 03:22, Andrew P. Black a écrit : > Raymond, I have some questions for you. (Is this what they all > "teledebugging"?) > >> Using TextMorph and entering a diacritic character like 'È' caused a >> walkBack > > This works fine on my Mac, but some of the keyboard stuff seems to be > platform specific. What platform are you using? Do you get the same > problem in a clean image (i.e., one into which you have not loaded the > Accufont stiff, or anything else. > > What character did you type? Was it one of > > $È (character value 142) > $Ë (character value 143) > $Í (character value 144) > $Î (character value 145) > > It looks as though the keycode stored in your event is the correct code > -256. Note that 144-256 = -112. If you select the stack frame (in the > top window of your debugger) for SmallInteger(Integer)>>asCharacter you > will see (in the middle pane) the code for that method, In the bottom > panes are the instance variables and temporaries in scope at that > point. self will be, I think -112. > > The keycodes should always be in the range 0-255, so something may be > rotten in your VM. Looks like something is being converted to a signed > 8-bit int when it should be an unsigned 8-bit int. > > This could perhaps be a bad C compiler, or a bad option to the > compiler. What's weird is that it only shows up in a textMorph. Or > does it? What happens if you type the same character in a Workspace? > > Andrew > > >> ================== >> Image version: Squeak3.1alpha >> latest update: #'Squeak3.1alpha' of 7 March 2001 update 3960 >> Current Change Set: Unnamed12 >> >> VM version: >> for: >> >> Receiver: #($> $ $ $ >> $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $! $" $# $$ >> $% $& $' $( $) $* $+ $, $- $. $/ $0 $1 $2 $3 $4 $5 $6 $7 $8 >> $9 $: $; $< $= $> $? $@ $A $B $C $D $E $F $G $H $I $J $K $L $M $N $O >> $P $Q $R $S $T $U $V $W $X $Y $Z $[ $\ $] $^ $_ $` $a $b $c $d $e $f >> $g $h $i $j $k $l $m $n $o $p $q $r $s $t $u $v $w $x $y >> $z ${ $| $} $~ $ $§ $? $í $ü $ì $ä $› $˝ $ñ $î $S $– $ë $? >> $Z $? $? $' $' $" $" $* $- $- $ó $Å $s $ $¶ $? $z $ç $ $° $¢ $£ $? >> $• $| $ß $® $© $™ $´ $¨ $- $Æ $Ø $∞ $± $2 $3 $¥ $µ $∂ $∑ $∏ $1 $∫ >> $ª $? $? $? $ø $¿ $¡ $¬ $√ $ƒ $≈ $∆ $« $» $… $ $À $Ã $Õ $Œ $œ $? $— >> $“ $” $‘ $’ $÷ $* $ÿ $Ÿ $⁄ $€ $‹ $Y $? $ﬂ $‡ $· $‚ $„ $‰ $Â $Ê $Á $Ë >> $È $Í $Î $Ï $Ì $Ó $Ô $? $Ò $Ú $Û $Ù $ı $ˆ $˜ $¯ $˘ $˙ $˚ $¸ $y $? $ˇ) >> >> Instance variables: >> >> Method (temp) variables: >> index: -113 >> >> Stack: >> Array(Object)>>error: >> Array(Object)>>errorSubscriptBounds: >> Array(Object)>>at: >> Character class>>value: >> SmallInteger(Integer)>>asCharacter >> KeyboardEvent>>keyCharacter >> TextMorph>>handleKeystroke: >> KeyboardEvent>>sentTo: >> TextMorph(Morph)>>handleEvent: >> TextMorph(Morph)>>handleFocusEvent: >> [] in HandMorph>>sendFocusEvent:to:clear: >> PasteUpMorph>>becomeActiveDuring: > Hi Andrew, you find the problem. In fact I use almost only MacOS 10, and two SqueakVM: a CarbonSqueakVM = 'Squeak 3.0Alpha11MT', and a CocoaSqueakVM from Marcel Weiher. The problem is linked to the CocoaSqueakVM, I will try to reload a new copy to see if the problem persist. Thank you very much for your Help.

1 0

Re: Squeak for 3-year-olds
by Alan Kay 02 May '01

02 May '01

I think Andreas is planning to set up localization of the etoy and other stuff in Squeak sometime in the next few months. Cheers, Alan ----- At 7:38 PM +0200 5/1/01, Cees de Groot wrote: >Sidetracking a bit on the topic of Squeak in education: my kids don't >happen to read English. Has anyone looked into i18n for Etoys?

1 0

Helio (prepacked) port?
by Bijan Parsia 02 May '01

02 May '01

Hey folks, I just download the Helio emulator and got Linux running on it, but I'm a bit baffled as to what to do to get Squeak loaded. Are there any prepackaged Helio "images" around? Or step by painful step instructions? :) Cheers, Bijan Parsia.

1 0

Jump to page: