Squeak-dev September 1999

squeak-dev@lists.squeakfoundation.org

133 participants
357 discussions

Re: A warning on HandMorph|handleMouseOver:
by Bob Arning 30 Sep '99

30 Sep '99

On Wed, 29 Sep 1999 23:33:22 -0300 "Leandro Caniglia" <caniglia(a)dm.uba.ar> wrote: >IMO this behavior is opposite to the stability of the system for one simple >reason: the programmer who creates a Morph subclass must be aware of the >fact that each morph is going to be included in a Dictionary and thus she >must provide an infallible hash method. Well, the problem extends a bit beyond that. ANY object whose hash can change over time (i.e. its hash is a function of mutable attributes) is a candidate for bad behavior in hashed collections (Set, Dictionary, whatever). If it gets inserted when its hash is one thing and you later try to see if it is there when its hash is something else, you have problems. The obvious solutions are not very nice. Who has a good one? Cheers, Bob

1 0

Re: An image is a lonely place
by Jan Theodore Galkowski 30 Sep '99

30 Sep '99

On Wed, 29 Sep 1999 17:18:32 Josh Flowers wrote: >This has a bit of a Jini feel to it... > >The one big problem that I have with Java (and Squeak), is that >changes to the class libraries easily break the entire system. This >is not a very big problem when your dealing with one VM and one set >of class libraries, but in a community setting, this is far to >fragile. [snip] This is very interesting. We're treading, IMO, into Dostoyevskii territory here.... Eiffel, as an OOP/OOD formalism with strict discipline tries to overcome this problem by enforcing the precondition-postcondition type of specifications on its objects and methods. This certainly works, and has a virtue to it, but it has, IMO, a kind of impedance mismatch with the programming and user community. In particular, in my experience of 24 years in software, the chance of getting users to formalize their wants even applying various incentives, disincentives, etc, is negligible. So, in that case, if one, as a software technologist, wants to use this stuff, one either takes on the entire responsibility of translating their imprecise requirements to precise ones or one needs to be very mean and force such elucidation out of users. Either choice is not very much fun and is just plain unrealistic. The other problem with it, IMO, is that it is based upon a fundamentally elitist view of software and software development. IMO, the thing which motivates the Web and Smalltalk and especially Squeak is the idea that This is a Big Conversation, tolerant of imprecision, understanding that things will be imperfect, yet in it for the long term payoff. It is, IMO, a lot more like playing music as part of an ensemble or orchestra than it is a reply of Bourbaki's finished proofs. In the context of a music group, it's not that "mistakes" aren't made, it's that they are bounded and -- in some cases -- exploited towards a group end. Yeah, I agree that classes can be fragile. But, again IMO, the benefit to be achieved by making all open for development outweighs whatever risks are incurred. We aren't deliverying refrigerators, at least not yet. I feel we should think of them as the little things that go wrong in the opening scenes of "Honey, I Shrunk Ourselves" or of "Back to the Future". Without risk and the possibility and encountering actual failure there is no learning and no growth. Sorry for rambling so long. ________________________________________________ Jan Theodore Galkowski demiourgos(a)smalltalk.org home.stny.rr.com/algebraist/ squeak.org/ www.smalltalk.org/ Get your FREE, private e-mail account at http://www.salonmember.com.

1 0

newbie silly question
by Adriano 30 Sep '99

30 Sep '99

Hello evrybody. I'm in a morphic project; in a workspace, I type: " a_AlignmentMorph new. a openInWorld " I select all and type Option-P (Print it) >From now on, EVERY message I want to send to my morph must be sent to the "a" variable. Now, in the workspace, I type: " AlignmentMorph new openInSpace " and print it. In this case, who shall I send the messages to? Thanks Adriano

2 1

Re: An image is a lonely place
by Luciano Notarfrancesco 29 Sep '99

29 Sep '99

Tom, I've implemented: o. speech synthesis (text-to-speech and singing voices) o. simple gestures synthesis (2D faces, SouthPark-like toons) o. 3D facial animation (the Waters' muscles model) I haven't released it yet, but if you want anything of this just e-mail me and I will send you a changeset. I think this stuff might be useful for you if you want to implement avatars. Cheers, Luciano.- -----Original Message----- From: Tom Morgan <tmorgan(a)acm.org> To: squeak(a)cs.uiuc.edu Date: Wed, 29 Sep 1999 07:58:02 -0400 Subject: An image is a lonely place > I've been noticing how lonesome it is inside an image. > > There are lots of strong place based/community elements > which surround Squeak. These are currently formed > by the basic networked apps of email, the discussion Swiki, > the mail archives, and so on. > > There is now lots of machinery which is available which > would allow basic collaborative support right inside > a running image (and there are some spacey ones imagineable > using the 3D stuff that feel like 'Snowcrash') > > The kinds of things that I have in mind revolve around > establishing a sense of 'place', which is shared, 'embodiment' > of you the user explicitly some how in the interface and > some amount of 'awareness' of the presence of > others in the shared place, all done with a light enough > touch that we don't end up with avatar fashion shows. > > In this picture, an Morphic book wouldn't be found by > reaching out across the network and loading it; instead, there'd be > a more or less far away part of your image (maybe > dressed up as an Alice world) that would let you wander > around and find the book you wanted. Perahaps the > author would hand it to you, if he happened to be around. > > While working, instead of the discussion Swiki, there'd be > small knots of others, gathered around places of interest > to them. You could glance around and see if anyone > was nearby, possibly interested in what you were up to. > > Over the course of a few years, I have seen naive users > comfortably navigate incredibly complex graph structures, when > they are presented with a place based metaphor. > > I am not sure how universally appealing these interface > metaphors are, but they seem to be captivating for > at least some portion of the population. > > I remember one posting about trying to incorporate the ICQ > protocols into Squeak. > > Are there others padding around this collaborative territory? > > ...Tom M > > >

3 2

Re: An image is a lonely place
by Josh Flowers 29 Sep '99

29 Sep '99

This has a bit of a Jini feel to it. I don't know if you've looked at the Jini Technology from Sun, but this is kind of what it's driving at. There's no UI implicit in Jini, but it's founded on the idea of a distributed community. The one big problem that I have with Java (and Squeak), is that changes to the class libraries easily break the entire system. This is not a very big problem when your dealing with one VM and one set of class libraries, but in a community setting, this is far to fragile. I was probing at an answer before when I brought up the idea of animate vs. inanimate objects, but I never came up with a solid solution. > I've been noticing how lonesome it is inside an image. > > There are lots of strong place based/community elements > which surround Squeak. These are currently formed > by the basic networked apps of email, the discussion Swiki, > the mail archives, and so on. > > There is now lots of machinery which is available which > would allow basic collaborative support right inside > a running image (and there are some spacey ones imagineable > using the 3D stuff that feel like 'Snowcrash') > > The kinds of things that I have in mind revolve around > establishing a sense of 'place', which is shared, 'embodiment' > of you the user explicitly some how in the interface and > some amount of 'awareness' of the presence of > others in the shared place, all done with a light enough > touch that we don't end up with avatar fashion shows. > > In this picture, an Morphic book wouldn't be found by > reaching out across the network and loading it; instead, there'd be > a more or less far away part of your image (maybe > dressed up as an Alice world) that would let you wander > around and find the book you wanted. Perahaps the > author would hand it to you, if he happened to be around. > > While working, instead of the discussion Swiki, there'd be > small knots of others, gathered around places of interest > to them. You could glance around and see if anyone > was nearby, possibly interested in what you were up to. > > Over the course of a few years, I have seen naive users > comfortably navigate incredibly complex graph structures, when > they are presented with a place based metaphor. > > I am not sure how universally appealing these interface > metaphors are, but they seem to be captivating for > at least some portion of the population. > > I remember one posting about trying to incorporate the ICQ > protocols into Squeak. > > Are there others padding around this collaborative territory? > > ...Tom M > >

1 0

Simple Morphic project
by Tim Olson 29 Sep '99

29 Sep '99

The following was posted in reply to a question on comp.lang.smalltalk; I thought I'd send it to the Squeak mailing list for those new folks who don't browse that newsgroup. "Andre M. (Mike) Maloney" wrote: > > Hi all, > > I'm just starting to learn smalltalk and am using Squak (2.4c). Welcome to Smalltalk and Squeak! You'll find a lot of things here to explore and have fun with. > I'm trying to build a simple counter and need some help. The counter > will have three buttons and a display. The displayed value will start > at one and the buttons will increment the displayed value by one, > decrement the displayed value by one, and reset the displayed value to > zero. > I've found the ValueHolder, button, and PluggableButtonController > classes, but how do I put the buttons and ValueHolder into a parent > frame (window?, view?)? Squeak has two different interface models: the older MVC (Model, View, Control), and the newer Morphic. The classes you found are part of MVC. You can use these to build an MVC interface, but it is a little more complicated than Morphic. If you want to try your example using Morphic, do the following: 1) make sure you are in the Morphic interface. If your menus are white with a blue frame around them, you are. Otherwise, select "open..." then "project (morphic)" from the main display menu, then enter the project by clicking on it. 2) Open up a Browser and have a look at the SameGame class; we're going to borrow heavily from it, here. 3) Open up a Workspace and type in the following: ---- cut here ---- "this is going to be the box containing our panel" frame := AlignmentMorph new orientation: #vertical. "create a 4-digit LED display to contain the count" display := LedMorph new digits: 4; extent: (4*10@15). "we want our buttons arranged in a horizontal row" controls := AlignmentMorph new orientation: #horizontal. "add a quit button" controls addMorph: (SimpleButtonMorph new target: frame; label: 'quit'; actionSelector: #delete). "add an increment button" controls addMorph: (SimpleButtonMorph new target: display; label: 'inc'; actionSelector: #increment). "add a decrement button" controls addMorph: (SimpleButtonMorph new target: display; label: 'dec'; actionSelector: #decrement). "add a clear button" controls addMorph: (SimpleButtonMorph new target: display; label: 'clear'; actionSelector: #clear). "add the controls and display to our panel" frame addMorph: controls. frame addMorph: display. "start playing with it!" frame openInWorld --- cut here --- We're not quite ready to run it yet. We don't have anything to handle the increment/decrement logic. To make this easy, we'll just add it to the existing LedMorph class, which we are using for the display. In the browser, find the LedMorph class, go to the "accessing" category, and find the "value:" method. We'll just tweak this a few times to add the new methods "increment", "decrement", and "clear". First, select the bold method header (value: aNumber) and change it to increment. Then replace the line "value := aNumber." with "value := value + 1.". Then accept your change (hit command-s or select "accept" from the menu in the lower pane. Now do the same thing for "decrement" and "clear" (doing the right thing with value). Now we're ready to give it a try. Select everything you typed in the Workspace and doIt (command-d or select doit from the menu). You should now have a light green panel with buttons and an LED display. -- Tim Olson

8 10

PNG in Squeak?
by Dwight Hughes 29 Sep '99

29 Sep '99

I'm wanting to create and manipulate PNG format images in Squeak, so before I go and re-reinvent the wheel, I wanted to check to see if anyone had already done something on this already. Anyone? -- Dwight

1 0

RE: An image is a lonely place
by Norton, Chris 29 Sep '99

29 Sep '99

Hi Tom & Squeakers everywhere... [Warning: the following is an opinion laden response!] Perhaps I am only addressing a fraction of what you wrote; I love science fiction, so forgive me if I go a bit off subject. :-) I really enjoyed the book "Snow Crash" (author: Neal Stephenson), and in particular, the whole concept of a Meta-verse. But, I would argue that we are still too far away from the concept of a replacement reality (virtual). Yes, we could perhaps model some of the 3D features of the Meta-verse with tools like Alice, but it would still be too hard to manipulate that universe to make it worth the effort. In the story, the lead character's computer tracks his eyes and fills them with a complete view of his place in the Meta-verse. As he moves about his room, his avatar (his virtual simulacrum) moves as he does. In order to create this kind of simulation, we would need some pretty fancy hardware that doesn't exist yet! I think that a bio-feedback suit would also be essential to navigation through a virtual world, as well as a "real" voice interface with your computer. We do have some limited voice synthesis & voice driven software out there, but it is too primitive to be truly practical in the way that it is used in the story. Also, I'm not sure that a stack of three dimensional representations of BookMorphs would be a particularly useful idea. We have libraries full of actual books in the physical universe, but we have to use card catalogues and computer indexes to find the books. A virtual rendering of a pile of books would be just as messy as a real pile of books. Don't get me wrong... From a fun & games perspective, I think a Meta-verse would be great! But from a practical, bang for your buck perspective, I'm not sold on the utility of Virtual Reality as a replacement of my office, library, etc. Now if we could create a voice operated Swiki, that might be an improvement in a useful direction! We would truly have a "chat" room! ---==> Chris

1 0

Change of e-mail address
by Jan Theodore Galkowski 29 Sep '99

29 Sep '99

The Disney Club Blast D-mail wasn't quite up to handling Squeak's 60 messages per day traffic, so I've changed the e-mail address I subscribe from. If anyone needs to reach me they can do so through the smalltalk.org address shown below, right. Thanks. ________________________________________________ Jan Theodore Galkowski demiourgos(a)smalltalk.org home.stny.rr.com/algebraist/ squeak.org/ www.smalltalk.org/ Get your FREE, private e-mail account at http://www.salonmember.com.

1 0

Squeak stability in large aquisition and data processing project ...
by Pedro Gomes 29 Sep '99

29 Sep '99

Hello, I'm involved in the development of a aquisition and data computation system in a research institute and was originally decided to use C++ to develop such complex system. I've been doing some experiments (fun!!) with Squeak for some time now and am considering the opportuinity to use version 2.5 for my project. My question is : Is Squeak fast and stable enougth to use in the development of such system.. Thanks in advance Pedro Gomes ------------------------------------------------------------------------------ Pedro Miguel Marrecas Gomes | 'Make it simple, not simpler' FCT/UNL Universidade Nova de Lisboa | Albert Einstein 1879-1955 Eng.Fisica | Fisico-Matematico Alemao email: pmmg(a)students.fct.unl.pt | http://students.fct.unl.pt/users/pmmg ------------------------------------------------------------------------------

5 4

Jump to page: