Squeak-dev July 1998

squeak-dev@lists.squeakfoundation.org

80 participants
111 discussions

Squeak 2.0 for Debian GNU/Linux
by Marcus Denker 08 Jul '98

08 Jul '98

Hi! There is an experimental package of Squeak2.0 (without updates) for Debian GNU/Linux 2.0 available at: ftp://ftp.ira.uka.de/pub/squeak/squeak-2.0.deb (it is compiled using the new libc6 (aka glibc2), so it is not possible to install it on Debian 1.3 without upgrading). Marcus -- Marcus Denker marcus(a)ira.uka.de fon@home:(0721)614235 @work:(0721)608-2749 Having the source code is the difference between buying a house and renting an apartment -- Behlendorf

1 0

Re: unification algorithm in Smalltalk?
by Les Tyrrell 08 Jul '98

08 Jul '98

As I recall, Prolog/V made use of a logic compiler that was already built into V/DOS- I don't believe having the Prolog/V Smalltalk source is sufficient to get it up and running in another dialect of Smalltalk. I still have Smalltalk/V DOS, and can check on this later. les > At 13:44 06/07/98 -0600, Kyle Pierce wrote: > > I am looking for Smalltalk AI tools in general, and any Smalltalk > > Can anyone > >provide pointers to such a thing, in a dialect that might be fairly > >Squeak-compatible? > > > Digitalk used to sell goodies for Smalltalk/V DOS. Among them was > a "Prolog/V" (maybe part of Carleton tools), which I was told was rather > buggy. Its syntax was an odd mixture of Smalltalk and Prolog. > > I cannot find any hint at it on the Web. However, I bought these goodies > long ago, and I might be able to find them back, if this is of interest to > you. > > Philippe Loève >

1 0

Redux Re: Questions On Proposed ANSI Standard DateAndTime
by R. A. Harmon 08 Jul '98

08 Jul '98

Hi James Foster: Thank you for taking the time to answer my post. Your reply was very helpful. I think I may understand it (mostly) now. At 09:27 AM 7/6/98 -0500, you wrote: [SNIP] >I believe that California is eight hours later than GMT. You're right. [SNIP] >>Is the following result of the Duration [SIC]>>#asLocal message correct? Is "[SIC]" used to indicate containing some error? If so, what? It just piqued my interest. [SNIP] >Or, treating the internal value as GMT: Sounds like an excellent suggestion. I'll do it. [SNIP] >>Should the Duration>>#< message given an operand other than an instance of >>class Duration generate an error or return false? > >Asking a duration of ten minutes if it is less than the color orange does >not seem to be false unless you can come up with a way to make 10 minutes >greater than orange. I would expect an error. [SNIP] I'm not sure what you mean. The color orange is always less that any other object in Smalltalk. See page 107 of "Smalltalk-80: The Language and its Implementation". I even amaze myself sometimes. I hope to finish up the ANSI additions to Smalltalk Express (I was more familiar with it) in a couple of weeks, then I can add them to Squeak. I'll post it and be happy to finally get some feedback on it. You really helped e out. Thanks -- Richard A. Harmon "The only good zombie is a dead zombie" harmonra(a)webname.com E. G. McCarthy

1 0

Re: Squeak 2.1 now available
by Matt Rosen 07 Jul '98

07 Jul '98

Dan Ingalls wrote: > If you want to keep working in an old image, and yet track the updates to 2.1, then > execute the following line after fetching updates in your 2.0 image... > > EToySystem eToyVersion: 'Squeak2.1' date: Date today > > ... and save it. Poof, you now have a system that thinks it is 2.1. Note however, that > the next update you download will want the new VM, so you should definitely retrieve > that asap (although almost everything will continue to run just fine on an old VM). > I just followed the instructions for making my Squeak 2.0 Win32 image 'think' it was 2.1, and then auto updated the latest changes. As we do not yet have an updated VM for 2.1 on Win32 (do we?), I used the existing VM. Bad things happened immediately. (In particular, the image hangs when entering a project, for example. A manual interrupt shows that the image is repeatedly halted with 'bad bitblt arg (fraction?) proceed to convert'). Given the nature of the released 2.1 changes, I suspect that either I have missed a step somewhere, or that we really do need the 2.1 VM before incorporating the 2.1 updates on Win32. I might suggest that others on Win32 hold off a bit until a 2.1 VM is available, or the problem is otherwise resolved. -- Matt Rosen, Director for Technical Services San Joaquin Delta College

3 2

Re: Squeak 2.1 now available
by Matt Rosen 07 Jul '98

07 Jul '98

Andreas Raab wrote: > Sorry for the delay. I was in the midst of refactoring the Win32 sources > (making them UNICODE clean and restructuring so that ports to Alpha and CE > should work "out of the box") when the official deadline for my thesis > struck me. Since it is due tomorrow you can expect an updated Win32 > version during the next week. Plenty soon enough for me. Whenever you get a chance is just fine. I _really_ appreciate the wonderful contributions you make to our little community, and my intent is not to pressure you in any way. Your thesis obviously comes first. I just thought that I would try to save others who might try to incorporate the 2.1 updates a little time and trouble. Thanks again! -- Matt Rosen, Director for Technical Services San Joaquin Delta College

1 0

Questions On Proposed ANSI Standard DateAndTime
by R. A. Harmon 07 Jul '98

07 Jul '98

Hi Squeakers: I'm trying to implement the <DateAndTime> protocol messages, but can't get a feel for what their definitions mean in terms concrete enough to implement. I posted a question about this on the comp.lang.smalltalk news group, but didn't get a response. I hope someone in the Squeak community might be able to help. I think I could figure out how <DateAndTime> works if I could just see what the various messages do with examples of real dates and times. Any help would be greatly appreciated. I'll use a short-hand notation of the class name with example values in parens. Duration(d:hh:mm:ss) DateAndTime(yy/mm/dd, hh:mm:ss, d:hh:mm:ss) date time offset As I understand it, a date and time in UTC means the date and time in GMT. So a DateAndTime class>>#year:month:day:hour:minute:second: message given the values for 8:00 am in California on April 26, 1997 return an instance DateAndTime(97/04/26, 02:00:00, 0:06:00:00). What does DateAndTime return from the following: DateAndTime(97/04/26, 02:00:00, 0:06:00:00) + Duration(0:00:00:11) ???-> DateAndTime(97/04/26, 02:00:11, 0:06:00:00) DateAndTime(97/04/26, 02:00:00, 0:06:00:00) + Duration(-0:00:00:11) ???-> DateAndTime(97/04/26, 01:59:49, 0:06:00:00) I think the implementation of DateAndTime>>#< should be as follows: < operand self offset = operand offset ifTrue: [ self theDate < operand theDate ifFalse: [ ^true "receiver #year, #month, & #day < operand #year, #month, & #day." ]. (self theDate = operand theDate) & (self theTime < operand theTime) ifTrue: [ ^true "receiver #hour24, #minute, #second < operand #hour24, #minute, #second." ]. ^false ]. ^self asUTC < operand asUTC Is the following result correct in that any of the operands are negative, it reduces the duration as the -1 in the second example reduces the one minute duration to 59 seconds? Duration days: 0 hours: 0 minutes: 1 seconds: 1 ???-> Duration(0:00:01:01) Duration days: 0 hours: 0 minutes: 1 seconds: -1 ???-> Duration(0:00:00:59) Is the following result of the Duration>>#* message correct? (Duration days: 1 hours: 0 minutes: 0 seconds: 1) * 9 ???-> Duration(9:00:00:09) Is the following result of the Duration>>#asLocal message correct? DateAndTime(97/04/26, 02:00:00, 0:06:00:00) asLocal ???-> DateAndTime(97/04/26, 08:00:00, 0:00:00:00) Is the following result of the Duration>>#asUTC message correct? DateAndTime(97/04/26, 02:00:00, 0:06:00:00) asUTC ???-> DateAndTime(97/04/26, 02:00:00, 0:06:00:00) DateAndTime(97/04/26, 08:00:00, 0:00:00:00) asUTC ???-> DateAndTime(97/04/26, 02:00:00, 0:06:00:00) Should the Duration>>#< message given an operand other than an instance of class Duration generate an error or return false? Is the correct global name DateAndTime or DateTime? I haven't this much trouble understanding a concept since trying to understand "present value of future dollars". Or was it "future value of present dollars"? Or maybe it was "dollar value of future presents"? -- Richard A. Harmon "The only good zombie is a dead zombie" harmonra(a)webname.com E. G. McCarthy

2 2

First set of updates for 2.1
by Dan Ingalls 07 Jul '98

07 Jul '98

Folks - I've just put out 9 new updates for 2.1, as follows... 140quickFixes-sw Moves the detent feature of rotation to the Halo, so that scripted rotations don't show the detent behavior. Performance improvement in Morph>>world. Fix to a bug in the copy-to-paste-buffer that showed up when the morph to be copied was flexed. Moved #stringForReadout from Integer to Number, as per Georg Gollman bug report. Warning issued upon attempt to duplicate an object bearing any anonymous script 141EmbedInTextFix-di Makes textAnchor offset work correctly within TransformMorphs. Corrects a problem with emphasis being ignored following a textAnchor. Amends = test for textAnchors to use == with the anchoredMorph. 142FloatTruncateFix-di Fixes Float truncate to return accurate values when they are too large to express as a SmallInteger (ie when the primitive fails). Also fixes LargeInteger asFloat which had been missing an occasional low-order bit. 143Weasel-HG-tk.cs 144DragOverChanges-di This changeSet removes the conflict between clients of mouseEnter: and mouseLeave:, regarding whether they are interested in mouseUp activity, or mouseDown (really hand-carrying) activity. It does this by establishing that, if you wish to receive enter and leave events, handlesMouseOver: will do so when the hand is empty and the mouse is not pressed, and that handlesDragOver: will do so when the hand is carrying something. Returning true to the first works well for most browser-like components. Returning true to both works for dragging tiles into scriptors, and probably dragOver only would suffice there. It adds dragOverMorphs to HandMorph, parallel to mouseOverMorphs, and ensures that a given morph is never in both at once. The successes achieved include: * Browsers no longer show two scrollbars at once. * Auto-scrolling now works, even when you pull the mouse over other panes. * Scrollbars do not disappear if you drag the elevator outside the scrollbar. * Resize handles do not conflict with adjacent scrollbars. * Time is not wasted on mouseOver support when you are dragging a morph. 145DragTweaks-di Restored the brown handle to morphic command halo, allowing a morph to be dragged within its owner's bounds without extracting it or changing its front-to-back order. Put in a feature that allows this constrained drag to be accessed merely by dragging more than 5 pixels with the cmd-key held down. Made handle size parametric in existing halo code. Adjusted placment logic of a few other command handles to work better on very small morphs. Fix a performance bug that repainted everything between 0@0 and a halo at pop-up time. 146FlexRefPosFix-di.st Makes the rotation handles respect center of rotation (can be set in paintBox). 147TwoTweaks-di Fixes the initial bounding box when bringing up a halo on a flexed morph. Renames 'window...' to 'open...' in the Morphic world menu, as in MVC. 148ScrollPane-scrollBy.st Fixes a bug that gave divide by zero after certain edits that affected scroll position.

1 0

Re: Squeak 2.1 now available
by Tim Rowledge 07 Jul '98

07 Jul '98

On Wed 01 Jul, Dan Ingalls wrote: > I have just placed a Mac self extracting archive in Smalltalk/Squeak/2.0 at st.cs.uiu > c.edu, called > Squeak2.1.sea.bin. Hooray! The first 'standard' image to run out-of-the-box on Acorn! (sfx:- happy dancing, scampering, cheering noises etc) tim -- Useful random insult:- Diarrhea of the mouth; constipation of the ideas. Tim Rowledge: rowledge(a)interval.com (w) +1 (650) 856-7230 (w) tim(a)sumeru.stanford.edu (h) <http://sumeru.stanford.edu/tim>

2 2

Re: unification algorithm in Smalltalk?
by Mark Wai 06 Jul '98

06 Jul '98

At 01:44 PM 7/6/98 -0600, Kyle Pierce wrote: >I am looking for Smalltalk AI tools in general, and any Smalltalk >implementation of a unification algorithm, in particular. Can anyone >provide pointers to such a thing, in a dialect that might be fairly >Squeak-compatible? For reference, you might want to take a look at Francois Pachet's NeOpus(http://www-poleia.lip6.fr/~fdp/). There are a also a few papers presented in OOSPLA as well. Check ACM electronic library for a list of such related topics. NeOpus is based on Atkinson and Laursen's OOSPLA 87 Opus paper. NeOpus should fairly compatible with Squeak since it is build on Smalltalk-80 derived dialect (ParcPlace). I also remember a guy (I forgot his name though) from Virginia, USA posted a message in comp.lang.smalltalk several years ago that he has a Smalltalk AI tool in the making. Do a search in Deja-news engine with keywords like AI, forward chaining etc and I am sure you will find a few useful tips. Good luck -- Mark Wai Frontier Systems Architecture Inc. mailto: mwai(a)ibm.net or:[ mwai(a)frontiersa.com] __

1 0

football
by David Farber 06 Jul '98

06 Jul '98

just out of curiosity: since Squeak has attracted a fair bit of international attention i was wondering what kind of representation we've got here for the remaining World Cup teams. for those not in the know, the World Cup semi-finals are tuesday and wednesday. Brazil plays Holland on tuesday, and France plays Croatia on wednesday. we got any Holland or Croatia fans out there? myself, i'm a firm supporter of Brazil's Selecão, and i know we've got few other Brasileiros on the list. sadly, both Argentina and Germany were eliminated in the quarter finals. i'm sure our fellow Argentinian and German Squeakers were not happy about that. dave -- j. david farber oo architect+mentor numenor labs incorporated in sunny boulder colorado dfarber(a)numenor.com www.numenor.com

3 2

← Newer
1
...
6
7
8
9
10
11
12
Older →