Squeak-dev September 2001

squeak-dev@lists.squeakfoundation.org

168 participants
408 discussions

RE: Fifth french column available...
by Ivan Tomek 04 Sep '01

04 Sep '01

Stephane, You are getting spoiled - your last title reads Exemples of introspection en Squeak Is that Franglais? Ivan Ivan Tomek Jodrey School of Computer Science Acadia University Wolfville, Nova Scotia, Canada Plan to attend ED-MEDIA 2002 at http://www.aace.org/conf/edmedia/default.htm and I_KNOW 2002 at www.know-center.at/i-know02 -----Original Message----- From: Stephane Ducasse [mailto:ducasse@iam.unibe.ch] Sent: Tuesday, September 04, 2001 6:00 AM To: squeak-dev(a)lists.squeakfoundation.org Cc: olivier nano Subject: Fifth french column available... Hi I put on my web page http://www.iam.unibe.ch/~ducasse/WebPages/columns.html the last column I wrote. I hope sincerly that the journal will continue to exist but they are in trouble right now. So if you are french, buy Programmez ;) to continue to read my columns Stef

1 0

[ENH] PairedCollection class
by Richard A. O'Keefe 04 Sep '01

04 Sep '01

"Lex Spoon" <lex(a)cc.gatech.edu> wrote: By the way, there is a nice generalization of the above with:foo: pattern. It looks like this: (firstArray with: secondArray) collect: [ :combo | combo first + combo second ] This use of #with: (which doesn't currently exist), is identical to ML and Haskel's "zip" function, and it allows you to use all of collect:, select:, reject:, do:, detect:, etc. without writing with:collect:, with:select:, with:reject:, etc. In brief, the presence of a "with" clause can be orthogonal to which collection operation is being used. Now it does exist. Note that it can be combined, so ((a with: b) with: (c with: d)) collect: [:combo | ... combo first first "a" ... ... combo first second "b" ... ... combo second first "c" ... ... combo second second "d" ... ] gives you iteration over four collections, should you want that. 'From Squeak2.8 of 13 June 2000 [latest update: #2359] on 4 September 2001 at 1:52:50 pm'! Magnitude subclass: #Pair instanceVariableNames: 'first second ' classVariableNames: '' poolDictionaries: '' category: 'Collections-Support'! !Pair commentStamp: 'raok 9/4/2001 13:21' prior: 0! I represent an ordered pair of objects. Pairs exist to support iteration over two or more collections at once and have no semantics other than aggregation. As a rule, you should look for a problem-domain abstraction you can use rather than reusing me. Pair first: x second: y makes a new Pair object. The obvious accessors are defined. Comparison is defined as ledxicographic comparison. My instance variables 'first' and 'second' mean what they sound like. ! SequenceableCollection subclass: #PairedCollection instanceVariableNames: 'first second count size ' classVariableNames: '' poolDictionaries: '' category: 'Collections-Sequenceable'! !PairedCollection commentStamp: 'raok 9/4/2001 13:45' prior: 0! I act like the sequence you would get by doing a with: b collect: [:x :y | Pair first: x second: y] except that I do not allocate a new array or ordered collection, nor do I create all the pairs straight away. I create pairs lazily, as they are needed. That means that previously generated pairs can be reclaimed by the garbage collector while you are still traversing me. The essence of a PairedCollection is to delegate accesses to the collections it is built from, so you are not allowed to change one. (a with: b) collect: [:aPair| ... aPair first ... aPair second ...] is like a with: b collect: [:x :y| ... x ... y ...]! !Pair methodsFor: 'comparing' stamp: 'raok 9/4/2001 13:25'! < aPair first < aPair first ifTrue: [^true]. aPair first < first ifTrue: [^false]. ^second < aPair second! ! !Pair methodsFor: 'comparing' stamp: 'raok 9/4/2001 13:24'! = aPair self species = aPair species ifFalse: [^false] ifTrue: [^first = aPair first and: [second = aPair second]]! ! !Pair methodsFor: 'comparing' stamp: 'raok 9/4/2001 13:26'! hash ^first hash "quick and dirty"! ! !Pair methodsFor: 'accessing' stamp: 'raok 9/4/2001 13:21'! first ^first! ! !Pair methodsFor: 'accessing' stamp: 'raok 9/4/2001 13:22'! first: anObject first _ anObject! ! !Pair methodsFor: 'accessing' stamp: 'raok 9/4/2001 13:26'! first: x second: y first _ x. second _ y. ! ! !Pair methodsFor: 'accessing' stamp: 'raok 9/4/2001 13:22'! second ^second! ! !Pair methodsFor: 'accessing' stamp: 'raok 9/4/2001 13:23'! second: anObject second _ anObject ! ! !Pair class methodsFor: 'instance creation' stamp: 'raok 9/4/2001 13:28'! first: x second: y ^self new first: x second: y! ! !SequenceableCollection methodsFor: 'converting' stamp: 'raok 9/4/2001 13:49'! with: anotherCollection "answer a new PairedCollection for iteration" ^PairedCollection first: self second: anotherCollection! ! !PairedCollection methodsFor: 'private' stamp: 'raok 9/4/2001 13:44'! species "Answer the preferred class for reconstructing the receiver." ^Array! ! !PairedCollection methodsFor: 'accessing' stamp: 'raok 9/4/2001 13:36'! at: anIndex ^Pair first: (first at: anIndex) second: (second at: anIndex)! ! !PairedCollection methodsFor: 'accessing' stamp: 'raok 9/4/2001 13:36'! at: anIndex put: anObject self error: 'a PairdCollection cannot be changed' ! ! !PairedCollection methodsFor: 'accessing' stamp: 'raok 9/4/2001 13:49'! first: aCollection second: anotherCollection aCollection size = anotherCollection size ifFalse: [self error: 'collection sizes must match']. first _ aCollection. second _ anotherCollection. size _ aCollection size. ^self! ! !PairedCollection methodsFor: 'accessing' stamp: 'raok 9/4/2001 13:35'! size ^size! ! !PairedCollection class methodsFor: 'instance creation' stamp: 'raok 9/4/2001 13:46'! first: aCollection second: anotherCollection ^self new first: aCollection second: anotherCollection! !

2 1

[updates] 21 more for 3.1 alpha
by Dan Ingalls 04 Sep '01

04 Sep '01

See summaries below. I know I have said this before, but this week we should begin to see some update contributions from this list appearing on the servers. More to come... - Dan --------------------------- -- 4283soundInNav-raa -- Bob Arning -- 23 August 2001 Add an option sound control to the project navigator (Mac only)" Preferences addPreference: #includeSoundControlInNavigator category: 'media' default: false balloonHelp: 'If true, add a system sound control (Mac only) to the project navigator.' 4284InterruptFix2-ar -- Andreas Raab -- 23 August 2001 A post-fix for the last fix which could result in 'debugging' the idle process (not good...). So only use the preempted process if its priority is >= the project's process in which case the preempted process may be running wild." 4285FixCmdShiftW -- Andreas Raab -- 23 August 2001 Remove CMD-Shift-W dependency from honorDeskTopCmdKeys - in an eToy-friendly environment we really want to be able to get to it." 4286TextCRAction-ar -- Andreas Raab -- 23 August 2001 Add a #crAction in text fields." 4287RenameOnCR-ar -- Andreas Raab -- 23 August 2001 Allow users to accept the name of a project on hitting enter." 4288RenameOnCRAgain-ar -- Andreas Raab -- 23 August 2001 Allow users to accept the name of a project on hitting enter." 4289EtoyLoginMorph-ar -- Andreas Raab -- 23 August 2001 Transfers the simple eToy login from the BJSqueak image." 4290SoundSliderTweaks-ar -- Andreas Raab -- 23 August 2001 Make the volume control slider reflect the currently set volume. BTW, the volume control is part of the common sound support primitives (and not Mac only) so rename it from setMacVolume to setSoundVolume (setting the Mac volume could have interesting implications if you get my meaning ... ;-)" 4291pluginFixes -- Andreas Raab, Michael Rueger -- 23 August 2001 Fixes a number of bugs relating to project loading and storing. - added reset inital preference settings to the squeakland release process - fixes saving project to local disc when external resource are present - reload the cached resources when the resource dictionary gets installed the first time." 4292pluginMods -- Michael Rueger -- 24 August 2001 Adds the widget flap to the etoy configuration. Adds the ObjectsTool to the widget flap. Plays the system beep on releasing the volume control." 4293etoyUsers -- Andreas Raab, Michael Rueger -- 24 August 2001 Enable a simple login mechanism for schools. *NOT* intended to reliably identify people; only intended to provide basic author information." Preferences addPreference: #eToyLoginEnabled categories: #(security scripting) default: false balloonHelp: 'If true, eToy logins are requested if one of the known servers provides a user list'. 4294flapTweaks -- Michael Rueger -- 24 August 2001 Enables flaps in standalone mode. Fixes a bug that prevented the widget flap from being enabled." 4295macDblClk -- Michael Rueger -- 24 August 2001 Adds double clickable project support for the mac. Fixes a problem with FileUrls when storing to local disk." 4296etoyFlapsFix -- Michael Rueger -- 24 August 2001 Makes the creation of default flaps dependend on the eToyFriendly preference." 4297flapFix2-mir -- Michael Rueger -- 24 August 2001 Yet another stab at the etoy flap problem." 4298StackInstVar-tk -- Ted Kaehler -- 26 August 2001 When an object in a Stack such as a field is made to store separate data for each card, it left the field in Undeclared. The inst var in the CardPlayer was created, so the Undeclared variable is just an annoyance. This fix declared the inst var without putting the name unto Undeclared." 4299SymbolicBorderFix-ar -- Andreas Raab -- 26 August 2001 Fixes a problem with viewer readouts for symbolic borders (which really expected a color to work). Also removes a few (now obsolete) #drawOn: methods in favour of the newly introduced #insetColor protocol (which is the equivalent to #raisedColor; both being sent to figure out on what color the bevel should be based)." 4300etoyDo-sw -- Scott Wallace -- 2 September 2001 Adds a #do command to the #misc category in etoy viewers, which performs a given script once, on the next tick." 4301collapseBalloon-apb -- Andrew P. Black -- 16 August 2001 Updates the balloon text on collapse and expand of the window, so that it says the right thing for a collapsed window. Tweaked by sw to make it robust in the face of a window without the title-bar buttons." 4302fileListPanes-sbw -- Stephan B. Wessels -- 29 August 2001 The button bar pane and pattern pane now have fixed heights in FileList. This is a revision to an earlier change set. The changes here now account for those published in update #4266" 4303debugIt-vbDew -- Vassili Bykov and Doug Way -- 13 August 2001 Adds a 'debug it' command to the text editor pop-up menu. Tweaked by sw so that the correct shifted menu gets invoked."

1 0

really silly question
by John Hinsley 04 Sep '01

04 Sep '01

Given two arrays of the same size filled with numbers, can someone give me a code snippet that will show me how to add the numbers at the same index to produce a third array? So that firstArray:= #(1 2 3 4) secondArray:= #(5 6 1 2) will result in a thirdArray (6 8 4 6) Cheers John (hangs head in shame....) -- Can't cope anymore? Desperate for help? Join the 12 step program for those who yearn to give up Microsoft: http://home.earthlink.net/~penguinrox/index.html

10 12

Re: About Celeste
by raymondasselin＠sympatico.ca 04 Sep '01

04 Sep '01

Felix wrote: > take a look at the method MailMessage(class)>>omittedHeaderFields. > You can add 'delivered-to' 'reply-to' etc. to omit those lines. > > Cheers, > > felix > > > raymondasselin(a)sympatico.ca wrote: > > Is there a way to not display, in the bottom pane of Celeste email > > browser, all the stuff like > > > > 'delivered-to: squeak-dev(a)lists.squeakfoundation.org > > reply-to: squeak-dev(a)lists.squeakfoundation.org > > list-help: > > <mailto:squeak-dev-request@lists.squeakfoundation.org?subject=help> > > list-post: <mailto:squeak-dev@lists.squeakfoundation.org> > > list-subscribe: <http://lists.squeakfoundation.org/listinfo/squeak-dev>, > > <mailto:squeak-dev-request@lists.squeakfoundation.org?subject=subscribe> > > list-id: The general-purpose Squeak developers list > > <squeak-dev.lists.squeakfoundation.org> > > list-unsubscribe: > > <http://lists.squeakfoundation.org/listinfo/squeak-dev>, > > <mailto:squeak-dev-request@lists.squeakfoundation.org?subject=unsubscrib > > e> > > list-archive: <http://lists.squeakfoundation.org/pipermail/squeak-dev/>' > > > > for each message and keeping this optional? > > > > I thought somebody did something about it but can't find it on > > 'http://macos.tuwien.ac.at:9009/Server.home' > > > > ...may be this was a dream? Great!! Thank you very much. Ray

1 0

Re: Modularity vs monolithic images
by Richard A. O'Keefe 04 Sep '01

04 Sep '01

cg(a)cdegroot.com (Cees de Groot) wrote: To me, sounds like a perfect model for the Squeak VM. Dynamically loaded objects aren't any slower than statically linked librarys on the OSes I know, and on the Smalltalk level the issue is non-existant anyway. SPARC Solaris, Alpha OSF/1: statically linked libraries are faster on those machines. PPC MacOs 8: the procedure calling convention is set up to favour dynamically linked code, a calling convention for statically linked code _could_ be quite a bit faster but I don't know whether any compilers exploit that. The SmallEiffel compiler does whole-program analysis and gets very worthwhile speedups by doing so. The Stalin Scheme compiler does whole-program analysis and really agressive inlining which wouldn't be feasible in a dynamically linked approach. I've been _amazed_ at the things the SPARCompiler C compiler will do when it is given a whole program in one file. Make no mistake: with today's machines and today's top end compiler technqiues, dynamically linked code can be somewhat to much slower than statically built code. Someone was doing type inference for Squeak. Whatever happened to that? Of course code can be modular without being dynamically linked, and it can be dynamically linked without being modular.

1 0

Reminder: Please compress all changesets posted to the list
by Doug Way 03 Sep '01

03 Sep '01

This is just a reminder to everyone posting [FIX]es and [ENH]anhcements to the list: *Please* compress your changesets as .gz files before attaching them to your message. This ensures that the changeset (or .st file) will be decodable and properly formatted, which makes it easier for the Fixes Review Group to review the submission for inclusion in the base image. Plain text .cs/.st files can have CR/LF issues, tabs converted to spaces, and other problems. To compress a changeset in gzip (.gz) format, simply use the "compress" menu item in the File List window in Squeak. For more details on how to post submissions to the list, see the Swiki page: http://minnow.cc.gatech.edu/squeak/398 (Actually, it can sometimes happen that even a gzip compressed changeset won't make it over to SQFIXES, if you use an oddball attachment encoding such as appledouble. This isn't quite as big a deal since they usually still appear in the other (Yahoo, etc.) archives. But it's still good practice to check the SQFIXES site (http://swiki.gsug.org:8080/sqfixes) to make sure your messages and attachments are appearing intact.) - Doug Way speaking as a Squeak Fixes Review Group guy :)

1 0

Squeak beginner
by Jaime Silvela 03 Sep '01

03 Sep '01

Hi all, I haven't found a list for newbies, so I'm writing to this one. Sorry! I've been using Squeak for a couple of weeks and I like it a lot. It's my first Smalltalk experience. I've just finished writing a version of the classic Snake game in Morphic, and I thought some other beginners might benefit from my code, and enjoy the game. Is there a repository for this type of stuff? Also, I'm plannning to rewrite in Morphic an Image processing system I wrote in C++. This would include a TIFF reader/writer (I haven't found one in the 3.0 image) and some morphological image processing operations. From browsing classes and reading the source code for the Demos, I think that for displaying images I should use ImageMorphs. Correct? What is slowing me is actual pixel manipulation. I see three classes relevant for this: Bitmap, Form and BitBlt. I'm learning them form reading source code, but I'd work faster with some documentation. Is there some documentation on these classes? Maybe in some of the books, say Goldberg and Robson? Thanks, Jaime Silvela from Spain

5 4

Design question
by danielv＠netvision.net.il 03 Sep '01

03 Sep '01

I've been looking at and modifying the RefactoringBrowser and the Squeak Parser, and I've been puzzled by something. Look at this bit of code - address _ NetNameResolver addressForName: serverName timeout: 15. address = nil ifTrue: [ -> self error: 'Could not find host address']. "connect the socket" self connectTo: address port: 110. (self waitForConnectionUntil: POPSocket standardDeadline) ifFalse: [ self close. -> self reportToObservers: 'failed to connect to server'. -> ^false ]. Marked by arrows are three different mechanism used by a functional object (a low level object) to request that it's client handle a situation. error: raises an exception to be handled somewhere higher in the call hierarchy. reportToObservers: basically know it has observers (the pattern abstracts away who they are) ^false tells it's immidiate caller that the results of an operation that is finished is false. I've noticed that many doer classes (like the Parser) use some mixture of these techniques. The Parser specifically, is used by quite a few different types of clients, uses both of the first techniques. Note that neither method is used exclusively for terminating a process. (see Parser>>notify:at: and also in Parser>>parse: sourceStream class: class noPattern: noPattern context: ctxt notifying: req ifFail: aBlock see "RequestAlternateSyntaxSetting signal and: [(sourceStream isKindOf: FileStream) not]") The Refactoring classes use all those and more, also. It's not clear to me why each is chosen in each case... it simply looks arbitrary and messy. So - What are the Pros and Cons for exceptions and other callback mechanisms? Daniel

2 1

Request for Combo Box - Drop Down List
by Phil Weichert 03 Sep '01

03 Sep '01

I still have not located a class that supports a read only combo box. This is what VisualWorks call the widget. The widget display a single line of text with a botton on the right. When the button is press a list with the possible selections appear just below the box. The list stays down until a selection is made or the user clicks outside of the list. Crude picture below. Would need a MVC and Morphic version! Anyone know where I can find the widgets? Thanks, Phil __________________ | CAT | V | In normal state: ------------------- __________________ | CAT | V | after button (v) pressed ------------------- | CAT | | DOG | | FISH | | FOX | | ZEBRA | -------------------

2 2

Jump to page: