Squeak-dev May 1998

squeak-dev@lists.squeakfoundation.org

82 participants
182 discussions

RE: Pluggability
by Alejandro F. Reimondo 30 May '98

30 May '98

Yes, yes, & yes. You are right. I will change Actions implementation soon. when I saw #evaluate (& #evaluate:) the first time, it sounds wrong to me... Later I say, why PPD people has used #evaluate and not #value ? Then I thought (provably more than needed to arrive a good answer ;-) ) " When I say value to anObject I want to know it's value and when I say evaluate I want to know the result of evaluating an object... For (runnable/)evaluable objects there is a difference (#value returns the object and #evaluates returns an object result of evaluation)... Provably blocks returns it's evaluation result as value... " Do you understand what happened? I found a way to justify #evaluate in my mind, because PPD didn't use #value. two times wrong. :-) Ale. ---------- Desde: Ralph E. Johnson[SMTP:johnson@cs.uiuc.edu] Enviado el: Viernes 29 de Mayo de 1998 10:46 Para: squeak(a)cs.uiuc.edu Cc: alereimondo(a)sugarweb.com Asunto: RE: Pluggability I looked at Actions, and liked what I saw. I've seen and done things like that before. My main complaint is that the method should be #value and #value:, not #evaluate and #evaluate: A block is a built-in Action, and it already defines a protocol, so we should use it. One of the interesting things is that Actions define both #value and #evaluate. So, you can use an Action instead of a block, but you can't use a block instead of an Action. Why? A block is just a built-in Action. Now, the fileIn defines #evaluate and #evaluate: for blocks, but that does not passify me! We should not proliferate protocols, and #value and #value: are well defined and short. They are not particular logical, but that is not so important for the most common messages. #at:put: is not particularly logical, either. It is a mistake to put #evaluate and #evaluate: in Object. ParcPlace did this with #value and #value: for awhile, and the problem is that you lose useful information because errors don't get caught right away but the system keeps on running for awhile before finally noticing some error. It is so easy to convert an object into an Action (just send #asAction to it) that you don't need these helper methods in Object. Further, using #asAction provides helpful information to the reader that an object is going to be used as an action, so it gives hints that can prevent the reader from having to keep on reading. BlockContext should define asAction to return self, not to make a new action that returns itself. A block is just a built-in action. (I'm hoping that if I say it enough, people will believe me!) Blocks should also have , defined to be the same as in Action. There is no reason for ActionSecuence, and its comment should be moved to ActionComposite. I like CompositeAction better as a name, by the way. It follows Kent Beck's pattern for naming subclasses, which I think describes the way most experienced Smalltalkers do it. -Ralph

2 1

Updates for today
by Dan Ingalls 30 May '98

30 May '98

075jhmNestedTests.cs Makes test-yes-no constructs nest properly in script editors. 076jhmOptionKeyAndDo.cs Eliminates tests for the option key (which isn't supported on all platforms) and renames #optionKeyPressed to #macOptionKeyPressed. Fixes a problem with transforms when invoking the yellow button menu from Morphic list views. Also replaces Object>do: with Object>in: allowing expression like: Float pi in: [:x | x * x] 077tk-bigString-EG.cs Fixes two bugs relating to writing long strings on streams Enjoy... The Squeak Team

1 0

Re: Yellow Button and Morphic
by Dan Ingalls 30 May '98

30 May '98

>I'd gotten tired of having to hit the little button at the top of the >scroll-bar in order to bring up menus for the SystemWindows in Morphic. >I thought I'd try to make the menus work with the "yellow" mouse button, >as they do in MVC. After browsing related senders and implementors for a >while, I realized that I was getting thoroughly lost. I can't seem to >see the forest for the trees. > >Where should I be looking to get a perspective on how input events, >particularly the mouse, are handled with Morphic? I did all this as part of the update: '061MorphicPaneMenus-di. >One thing that I noticed: it looks like any mouse button will act like >the "yellow" one if the "Option" key is pressed. Trouble is, I don't >have an option key. As I recall, that's something that's on the Mac? >Does the Mac still sport a single button mouse? Sorry, I'm culturally >deprived ;-) Unfortunately I was confused about how we work the Mac option key and the "yellow" button. Here is a QUICK FIX, that we will issue later as an update with probably some changes that make it clear that optionKey can only be tested on the Mac. !PluggableListMorph methodsFor: 'events' stamp: 'di 5/29/1998 11:13'! mouseDown: event onItem: aMorph event yellowButtonPressed ifTrue: [event shiftPressed ifTrue: [^ self shiftedYellowButtonActivity] ifFalse: [^ self yellowButtonActivity]]. model okToChange ifFalse: [^ self]. "No change if model is locked" ((autoDeselect == nil or: [autoDeselect]) and: [aMorph == selectedMorph]) ifTrue: [self setSelectedMorph: nil] ifFalse: [self setSelectedMorph: aMorph]! ! !TextMorphForEditView methodsFor: 'all' stamp: 'di 5/29/1998 11:14'! mouseDown: event event yellowButtonPressed ifTrue: [event shiftPressed ifTrue: [^ editView shiftedYellowButtonActivity] ifFalse: [^ editView yellowButtonActivity]]. ^ super mouseDown: event! !

2 2

ian, andreas, stp
by Dan Ingalls 30 May '98

30 May '98

Folks - The original version of SqueakV2.sources had a linefeed character inadvertently pasted into it. This is fine for Squeak, and will normally cause no problem. However, some text conversion and file transfer programs will treat linefeed specially, and this can cause a problem. Squeakers: If you find a character missing at the beginning of methods in classes that follow 'Server' alphabetically, you have this problem. The correct length for SqueakV2.sources is 5,602,467 characters. You should be able to correct it by re-fetching your sources in binary mode. [A quick fix if you have this problem, and your file is one character short, is to replace the unique string 'one ftp' by the same string with two spaces, as 'one ftp'] Squeak Site Maintainers: I have just placed a new version of SqueakV2.sources.sea.bin on the UIUC archive (write date May 29), that has no linefeed characters in it. At your convenience, please update your site so that all copies or derivatives of this file are replaced with the new version. Our apologies for the inconvenience. - Dan

2 1

re: location of Squeak party (Rosotti's / Alpine Inn / Zot's / etc.)
by Craig Latta 30 May '98

30 May '98

Hi-- It's at 3915 Alpine Road near Arestradero Road in Portola Valley, CA, USA. http://www.vicinity.com/yahoo/yt.hm?FAM=yahoo&CMD=GEO&SEC=geo&AD2=3915+Alpi… -C -- Craig Latta composer and computer scientist craig.latta(a)netjam.org www.netjam.org latta(a)interval.com Smalltalkers do: [:it | All with: Class, (And love: it)]

2 1

Squeak get together
by Steve Dekorte 29 May '98

29 May '98

Could someone post the address? Thanks, Steve

1 0

RE: Pluggability
by johnson＠cs.uiuc.edu 29 May '98

29 May '98

I looked at Actions, and liked what I saw. I've seen and done things like that before. My main complaint is that the method should be #value and #value:, not #evaluate and #evaluate: A block is a built-in Action, and it already defines a protocol, so we should use it. One of the interesting things is that Actions define both #value and #evaluate. So, you can use an Action instead of a block, but you can't use a block instead of an Action. Why? A block is just a built-in Action. Now, the fileIn defines #evaluate and #evaluate: for blocks, but that does not passify me! We should not proliferate protocols, and #value and #value: are well defined and short. They are not particular logical, but that is not so important for the most common messages. #at:put: is not particularly logical, either. It is a mistake to put #evaluate and #evaluate: in Object. ParcPlace did this with #value and #value: for awhile, and the problem is that you lose useful information because errors don't get caught right away but the system keeps on running for awhile before finally noticing some error. It is so easy to convert an object into an Action (just send #asAction to it) that you don't need these helper methods in Object. Further, using #asAction provides helpful information to the reader that an object is going to be used as an action, so it gives hints that can prevent the reader from having to keep on reading. BlockContext should define asAction to return self, not to make a new action that returns itself. A block is just a built-in action. (I'm hoping that if I say it enough, people will believe me!) Blocks should also have , defined to be the same as in Action. There is no reason for ActionSecuence, and its comment should be moved to ActionComposite. I like CompositeAction better as a name, by the way. It follows Kent Beck's pattern for naming subclasses, which I think describes the way most experienced Smalltalkers do it. -Ralph

4 3

Keyboard events in Morphic
by Martin Pammer 29 May '98

29 May '98

Dear Squeakers Tim Olsons SameGame showed us how to develop an entertaining game with the morphic framework. As for learning Morphic I decided to port a game I have written for VSE and Dolphin Smalltalk as an excercise. I encountered the following 2 problems. 1.) I have a class #MyGame that is a subclass of #AlignmentMorph. I want to add some submorphs (kind of tiles) and I want #MyGame to handle keybord input (cursor keys) to move those tiles. I tried #on:send:to: and #on:send:to:withValue: but I had no success. Overwriting keyStroke: did not get me further. It seemed that I did not get those #keyStroke events. Looking in the hierarchy for senders of #keyStroke: did not help me much either. Maybe someone can shed some light on this. 2.) When I create a morph like in the following example initialize | m | super initialize. self extent: 400@200. m := BorderedMorph new color: Color black; extent: 20@20; position: 30@30. self addMorph: m. and then I ask the morph m's position I get somthing like 183@53 and not the expected 30@30. What is wrong with this ? With best regards Martin Pammer mpammer(a)vrz.net

1 0

Squeak and the Internet
by Dan Shafer 29 May '98

29 May '98

Sorry if this has been discussed before; if it has, please just point me to the right place and I'll go read all of it. I'm contemplating doing a column for CNET's BUILDER.COM (where I write the "Master Builder" column every week for people who build Web sites) on the subject of Squeak as a possible tool/platform for Web developers. One place it seems to me that Squeak has some intriguing potential is in the VM universe, where Java has kind of staked out the turf. There would be some obvious advantages to a Smalltalk VM vs. a Java VM, not the least of which is the relative maturity and completeness of the whole Smalltalk world as evidenced by Squeak. The VM is clearly under 500K. What isn't clear to me is how much would have to be added to the VM to create a browser-usable (applet-like) or stand-alone, networkable application relying on the VM. Clearly that is to some extent application-specific, but what is the general feeling of the community? Does Squeak's VM really have the potential to be a replacement or alternative to Java VMs? >From the famous "Back to the Future" paper, I observe: "Squeak is a portable, malleable, full-service computing environment, including browsing, split-second recompilation, and source debugging tools, all in a 1-megabyte footprint. Though many of its strengths are rooted in the past, Squeak is suited to the intimate computing potential of PDAs and the Internet, and our work is, now more than ever, inspired by the future." This gives me a lot of hope, but is it hope I can safely convey to the hundreds of thousands of Web designers who will read my column on this subject? Is there any working code anywhere that would exemplify what I'm trying to get at here? -- |=================================================| | Dan Shafer, Executive Technical Producer and | | Editor-at-Large, CNET's Site for Site Builders | | dshafer(a)cnet.com - Preferred Method of Contact | | http://www.builder.com - Main Builder Site | |http://buzz.builder.com - Online Discussion Area | |http://web.builder.com - Semi-Annual Conferences| | Save the Dates - Dec. 7-9, 1998, Web.Builder | | New Orleans Strikes Again! | |=================================================|

2 2

RE: Mouse Buttons and Modifier Keys (was Yellow Button and Morphi c)
by Kevin Driedger 29 May '98

29 May '98

[Kevin Driedger] [snip] The Smalltalk image from which Squeak is derived was designed to > run with a three-button mouse on an Alto or Dorado computer. These > buttons were known as "red", "blue", and "yellow", and some > mice really had buttons of these colors. > > Win32 machines have only two mouse buttons, so the mapping is: > left-mouse -> "red" > alt-left-mouse -> "blue" > right-mouse -> "yellow" > > The Mac has only a single button, so the mapping is: > mouse -> "red" > cmd-mouse -> "blue" > option-mouse -> "yellow" > > I believe the mapping of modifier keys is: > Mac Win32 Unix > shift shift shift > ctrl ctrl ctrl > cmd alt meta(?) > option --- --- [Kevin Driedger] On my Sun workstation sitting in front of me there are the following keys: shift, ctrl, alt, meta and "Compose". On any newer Windows machine they have this "Windows" button with the Windows (TM) logo on it (something equivalent to say, the singer formerly known as Prince). On Linux machines there is a difference between the left and right alt keys where the left key is the compose key. As a possible solution, we could map the option modifier to say, Ctrl-Shift or Alt-Ctrl. > Re: > >On a more general note: Are there any efforts going on that are > related > >to making the user input interface more easily configured or > customized, > >without the need to change methods in the base system? > > Not that I know of. [Kevin] I guess that is what open source is for, eh?

1 0

Jump to page: