Squeak-dev November 1999

squeak-dev@lists.squeakfoundation.org

164 participants
545 discussions

[BUG] Can't play SWF files
by Mark Guzdial 05 Nov '99

05 Nov '99

I should have tested this before sending out my earlier email, but now that I have some SWF files on disk (Thanks, Ivan, for the method, and David Cramer for the pointer), I can't play them from 2.6 or 2.7a. When I try to "open as Flash" from the fileList, I get an error that "Message not understood: +" in FlashSpriteMorph>>step method. lastStepTime is nil. I tried setting it to Time millisecondClockValue in initialize, but that just opens up an empty FlashPlayerMorph which does nothing. Any suggestions? Thanks! Mark -------------------------- Mark Guzdial : Georgia Tech : College of Computing : Atlanta, GA 30332-0280 (404) 894-5618 : Fax (404) 894-0673 : guzdial(a)cc.gatech.edu http://www.cc.gatech.edu/gvu/people/Faculty/Mark.Guzdial.html

3 4

Re: Self 4.1 for Mac OS Its too slow to be used for anything useful.
by alban 05 Nov '99

05 Nov '99

After reading the blurbs on the self site about how fast self was (half as fast as C & so on.) I downloaded it and tried it out on my iMac 233. Its really very interesting but it is way to slow to use for anything. I find it too slow to even mess about with. Reading the release notes it seems that most of the optimizations that make it so fast on SPARCs do not apply to the MAC OS port. As an example I get a frame rate of about 95fps in squeak and 14 in self - and that is the fastest I have seen. This is a shame. This self 4.1 port is an Alpha release (for friends of self) so maybe someone who understands it will accelerate things (at least 10 times faster would be a nice start.) It really puts the speed of squeak in perspective though. squeak is fast. Just out of curiosity does anyone know if self actually is fast on a SPARC?

3 2

Re: Seeking Good Flash Files
by Mathieu vanEchtelt 05 Nov '99

05 Nov '99

http://www.exclaim.ca/surreal/ani/swfs/ >From this site you can download different flash files. The great thing of running these files in Squeak is, that you can pickup the different flashchararcters out of the scene and treat them as morphs! (very, very cool and powerful). A week ago I also gave a demo with Squeak to promote Swiki. It's so easy to make an really cool presentation. The presentation was based on the BookMorph, every page had besides TextMorphs all kind of more 'Funny' and interactive morphs to make the presentation more entertaining. (Morphs can really be like stargates to totally (new) interactive worlds). Also the 3d alice-bunny was dropped into a pageMorph. It was without the background and it was following the cursor (see the example code in the play-with-me project). At the end my bookmorph was about 1.6 mb big. I didn't show my audience everything I made. It was 'only' a presentation about 'why we should use Swiki' and I didn't wanted to be as confusing as possible. hope this helps and good luck, mathieu >From: Mark Guzdial <guzdial(a)cc.gatech.edu> >Reply-To: squeak(a)cs.uiuc.edu >To: squeak(a)cs.uiuc.edu >Subject: Seeking Good Flash Files >Date: Fri, 5 Nov 1999 09:31:23 -0500 > >I'm giving a big Squeak demo next Thursday (to a crowd of Java zealots -- >gulp!), and I'm gathering cool resources. I'm having a heckuva time >finding good flash files to demo. Most of those that I find on the net are >embedded in an HTML file already, and I can't figure out how to download >them directly. I was able to get some from downloading Flash, but flash >movies of menus and dialog boxes are less than exciting. > >Could anyone, please,: >- either send me a cool Flash movie or two to demo, or >- point me to where you can actually DOWNLOAD good flash file demos? > >I'm particularly eager to show Andreas' cool trick of selecting a >sub-element of the Flash movie and pull it out. But again, pulling out a >menu item out of the provided Flash demos just isn't much fun...:-) > >Thank you! > Mark > >-------------------------- >Mark Guzdial : Georgia Tech : College of Computing : Atlanta, GA 30332-0280 >(404) 894-5618 : Fax (404) 894-0673 : guzdial(a)cc.gatech.edu >http://www.cc.gatech.edu/gvu/people/Faculty/Mark.Guzdial.html > > > ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com

1 0

modifying Process
by Stephen Pair 05 Nov '99

05 Nov '99

In an attempt to add additional context information to a Process, I have added an instance variable to Process. I got a warning about such shenanigans, but everything seems to have worked. All processes now have this additional instance variable and everything seems to work as designed. Is there danger lurking? Or, am I ok? Thanks, Stephen

1 0

Re: Self 4,1 for MacOs?
by Bert Freudenberg 05 Nov '99

05 Nov '99

On Thu, 4 Nov 1999, Jan Theodore Galkowski wrote: > What do you mean by "elegance" here? Is it something other than the > sheer basicness of lambda calculus? Or something else? "Elegance. Beauty that springs from the union of power and simplicity." (Gary H. Anthes) see http://www.computerworld.com/home/features.nsf/all/980420id and also the /. discussion http://slashdot.org/askslashdot/980718/1321229.shtml This is not Self-specific, but a lot of that applies to it. -Bert-

1 0

RE: TT800 (was RE: PRNGs (was [Q] Project: Better performance fo r LargeIntegers))
by Baveco, Dr. J.M. 05 Nov '99

05 Nov '99

Thanks for the code! However, I think the attachment missed a last few lines (the instantiation category at the class side). I guess you redefined <new> there, to return <super new initialize>? hans > OK, I got it and have converted it to Squeak (with some testing > assistance from Jeff). The code is attached. It's basically a > straight conversion from C. It uses, by its nature, 32-bit long > integers and is slow by a factor of about 30 relative to Squeak's > Random class. > > References and abstracts to the papers are included in the class > comment. The original C code (less its tiny main program) is in a > method as a comment. There are class methods that test the generator > and also assure that it answers the same results as the C version. > > This is a good example of why long integers need to be fast, and how > easy it is to hit that performance brick wall. > > If you are using Random, you can try this by changing the class name > to RandomTT800. The #next method answers the next pseudo-random > number and #seed: resets the seed. The generator automatically self > seeds from the millisecond clock. > > Dave << File: Random-TT800.4Nove333pm.cs >> << File: ATT1337399.txt >>

1 0

Re: Behaving Selfishly? Was: [Re: Self 4,1 for MacOs?]
by JArchibald＠aol.com 05 Nov '99

05 Nov '99

=> 11/5/99 1:44:20 AM EST, merlyn(a)stonehenge.com => Sorry, I took a closer look. Both #allClasses and #allObjectsDo: have been pruned by #majorShrink. And I'm not sure how #allClassesDo: may have simplified things. :-{o{{ ==> Oh! from this bearded man. Jerry. Jerry Archibald systemObjectivesIncorporated

1 0

Behaving Selfishly? Was: [Re: Self 4,1 for MacOs?]
by JArchibald＠aol.com 05 Nov '99

05 Nov '99

=> 11/4/99 9:00:59 PM EST, Helge.Horch(a)munich.netsurf.de => << I'm very interested in seeing Squeak become a little more Selfish. >> To present a contravalent point of view~ The interesting interchange that has been going on in the Mailing List parallels the broad and deep investigations of Morphic within Squeak. The general attraction of such an approach also created much enthusiasm and software within Digitalk's PARTs development. The attraction of the Selfish approach seems to have its best expression with the view that "it would make things much simpler." I would like to argue that this simplicity is an illusion. In a class based arena, the number of distinct ideas for which an individual is required to map a conceptual graph is precisely the number of classes in the domain which one is interested in. For example, Smalltalk in an MVC Squeak incarnation (in a Mini version), can be understood in approximately 100 classes (more or less). (I am working with a Mini Swiki 2.6 with a pruned image <roughly Dan Ingall's #majorShrink> but which also includes the 'System-Network' and 'PluggableWebServer' categories which has exactly 257 classes in it) So I claim that there are roughly 100 ideas which must be known, some perhaps mastered, others perhaps regulated to "over there," "up there," or "down there," in order to claim to have knowledge of basic Smalltalk. Some of these ideas may be overly complex, I will agree, but just the finiteness of the bound of how many are there determines a rough conceptual "size" for the set of ideas which an informed developer must master. Even the entire of distributed Squeak 2.6 has only 1089 classes. Again, I would assert that this is a rough measure of the number of ideas which a developer must master in order to get their arms around the system. In a Self oriented environment, however, the number of distinct behaviors is not governed by the number of classes which form the system, but the number of instances. Each instance will have a distinct behavior determined by the signature which is "pertinant" to that object. If you argue that "Well, many of them have the same (or similar) behavior" I would argue in return that "Well, you are just reintroducing the Class concept by doing that." In any case, in distributed Squeak 2.6, there are 125,678 instances of objects. While one can argue about the size of 100 vs. 257 vs. 1089 classes as a measure of Squeak's conceptual complexity, it is hard to argue that 100 (or 257, etc.) is in any way of the same order of complexity as 125,678. Enough said. I would be interested in hearing the responses of the Self proponents. The numbers I quote come from: (Smalltalk select: [ :i | i isKindOf: Class ]) size => 1089 | count | count := 0. (Smalltalk select: [ :i | i isKindOf: Class ]} do: [ :i | count := count + i allInstances size ]. count => 125,678 With best counting, Jerry Archibald systemObjectivesIncorporated -------------------------------------------- A belated Halloween message: "One day of the year, you get to dress up and pretend that you are someone else, 364 days of the year, you get to dress up and pretend that you are you."

1 0

Re: Self 4,1 for MacOs?
by Duane Maxwell 05 Nov '99

05 Nov '99

>On Thu, 4 Nov 1999 13:35:13 Duane Maxwell wrote: >>Stefan writes: > >>>Sun says that their new way of describing >>>objects is better than classes. Now I have >>>always thought that a "class" is, in the >>>first place, nothing more than a naming >>>convention. > >>>Can anybody explain to me what they are >>>talking about then? &BTW: do you agree with >>>Sun, or is this just another "dot in com" >>>talk? > >>What Self does is rather interesting - you >>basically clone existing objects and add your >>own "slots" (which roughly correspond to >>instance variables and methods) to the copy. >>You then use it to make copies, and so on. >>The idea of classes sort of becomes >>irrelevant - to create a new instance of >>something, you instead just find an object >>with the correct slots and copy it. You can >>also choose to make a very primitive object, >>make its parent an existing object, and >>simply override and/or add slots to do what >>you want. > >So, Duane, this "very primitive object" would >be the analog of a "virtual class" I guess? The most primitive sort of object in Self would be one that has no slots. I wouldn't know how to make one except clone something else and remove all of the slots, including the "parent" (which is the object to defer to if you're asked for something you don't have a slot for, kind of like a superclass object). In Self you can have more than one "parent" object, so multiple inheritance is possible. One version of Timothy Budd's "Little Smalltalk" did something similar to Self - object instances were really a linked list of smaller subobjects, one per class in the hierarchy up to Object. Inefficient, of course, but it was simply an academic exercise, not a production environment. > >What I don't understand is how, with the >additional dynamics, why doesn't this then >become just a SCHEME with syntactic sugar? I suppose an analogy of "Smalltalk:Self is as LISP:Scheme" is not a big stretch. > >>For instance, to make a 3D Point, I'd clone a >>prototype 2D Point and add the extra slot for >>the z value. Since the original and the >>close share the same parent object, the 3D >>Point inherits all of the slots (including >>methods) of the original. Now, to make more >>3D Points, I clone the one I just made. > >But isn't it important, for purposes of >professional software development, to have >a framework to use Meyer's principles for >OO? I know one can be dynamic in Smalltalk, >but that isn't always a good thing. Who said anything about professional software development? :) I guess one of the interesting points that Self makes is that you can have objects in a classless system, something I doubt many folks know that. You're welcome in Self to layer anything you want on top of it, including, I suppose, destroying its dynamic nature if you want. >>Morphic, which came from Self, simulates >>these additional slots with the "extensions" >>object which hangs off of Morph. Squeak uses >>them most frequently to store balloon help >>text. The "duplicate" halo would be critical >>in the Self world, and so would the >>functionality we see as "flaps". > >>Frankly, Self is a very elegant system, which >>takes Smalltalk-like languages towards being >>even simpler and easier to understand. You >>can artificially impose a more familiar class >>structure if you want, as evidenced by the >>Smalltalk implementation in Self. Being >>simpler, it lends itself pretty well to >>interesting implementations, including >>runtime translation to native code. I think >>I saw a note from one of the Squeak >>luminaries (Alan?) that Self could have been >>Smalltalk-80's natural successor. > >What do you mean by "elegance" here? Is it >something other than the sheer basicness of >lambda calculus? Or something else? I guess it's the simplicity of the design and language I like - all of the fluff removed to leave a pure center on which you can build anything you like. Don't get me wrong, I _love_ Squeak, but I have to admit that Self is alluring in concept, even if not in implementation. I'm glad Squeak has been influenced in part by some of the Self project. > >>I would encourage people to take a look at >>it. I got tremendous insight into Morphic's >>architecture and philosophy by examining Self >>and reading the various white papers. > =================================================== Duane Maxwell dmaxwell (at) entrypoint.com CTO http://www.entrypoint.com EntryPoint (619) 578-8500 x226 Information contained herein is my personal opinion and not necessarily that of EntryPoint. ===================================================

1 0

RE: Self 4,1 for MacOs?
by Helge Horch 05 Nov '99

05 Nov '99

At 17:04 04.11.99 -0500, Stephen Pair wrote: >A while back, someone implemented some exploratory code that allowed one to >create and clone objects ala Self. Anyone know if that stuff has >progressed? Hans-Martin Mosner's "ProtoObjects"? Dunno, it should be available at <http://www.heeg.de/~hmm/squeak/System-Prototypes.st> Another interesting approach was outlined in Kent Beck's "Guide to Better Smalltalk". (The Digitalk model: Basically, instances wouldn't carry a class reference anymore. Instead, they'd refer to an array of MethodDictionaries. These would know which class - if any - coined them.) >I'm very interested in seeing Squeak become a little more Selfish. <g> I guess it's on several folks' ToDo lists. HTH, Helge

1 0

← Newer
1
...
47
48
49
50
51
52
53
54
55
Older →

Jump to page: