Squeak-dev July 1999

squeak-dev@lists.squeakfoundation.org

139 participants
367 discussions

Start a nNew thread

Ad: instance Variables...
by Peter Novak 28 Jul '99

28 Jul '99

3 2

[UPDATES] More on the way to 2.5
by Dan Ingalls 28 Jul '99

28 Jul '99

Folks - I've just released a number of new updates for test pilots. These include tidbits like instant access to WordNet (world menu/help.../definition of...), numerous fixes, refactoring (project -> worldState in PasteUpMorph), and a new ClassBuilder facility that is capable of a number of formerly impossible changes to the class hierarchy (such as moving subclasses from ClassDefinition down to Class). The summaries follow below. - Dan --------------------------- 1324DrawFixes-ar -- Andreas Raab -- 15 July 1999 Fixes FormCanvas to honor shadowDrawing in certain cases." 1325misc-tkKU -- Ted Kaehler -- 16 July 1999 Fixes bug when you Cancel after choosing 'save morph at URL'. Refrains from putting 'recompiling...' in the Transcript if the class has no methods. Another minor fix." 1326scriptingFixes-sw -- Scott Wallace -- 15 July 1999 1. Fixes up the step methods of half a dozen morphs which did not call 'super step'; with this fix, those kinds of morphs can both function in the way they expect to and also obey any scripting commands. 2. Fixes the bug that made the penColor readout in a viewer for a rotated object not be hooked up properly. 3. Bulletproofs TransformationMorph.printOn: against the situation where the transformation has no submorphs (it happens -- indeed it happened when debugging the previous item.) 4. When updating a flap-based viewer after some external change, gets its header right. 5. Fixes the bug that when a morph got flexed the external name of the original got set to the string 'nil'. 1327emptyScripts-sw -- Scott Wallace -- 16 July 1999 Offers some recourse against the proliferation of empty scripts caused by the awkward drag & drop action of tile scripting. Three avenues of improvement are supplied: (1) When the drop of a tile phrase results in the establishment of a new scriptor, the opportunity is taken to delete all now-empty and never-named tile scripts belonging to the object. This deals with the primary offending cases. (2) When you dismiss an empty unnamed scriptor, it is also automatically removed from the player and its viewers. (3) There is a new command in the viewer menu allowing you proactively to expunge all the empty scripts of the player." 1328BSDAccept-ar -- Andreas Raab -- 16 July 1999 This change set adds an optional BSD-style accept() mechanim for sockets. It requires two additional functions in the support code: * sqSocketAcceptFromRecvBytesSendBytesSemaID(...) is used to create a new socket by accept()ing it from the given server socket. The function may fail by calling 'success(false)' if not supported. * sqSocketListenOnPortBacklogSize(...) initializes a socket for future accept() operations. If this primitive fails accept() should not be used (see ConnectionQueue>>listenLoop for how determine the right way of accepting connections). The function may fail by calling 'success(false)' if not supported. A note on state changes: Whenever a socket has been initialized by the second method its state will be unconnected. When a connection can be established by the accept primitve its state changes to connected. After a connection has been accepted the state of the server socket will change back to unconnected *only* when there are no more connections pending, e.g. a loop like: [server isConnected] whileTrue:[ client _ server accept. ]. will extract all pending connections." 1329PrettyPrinting-sma -- Stefan Matthias Aust -- 18 July 1999 Makes browsing with prettyPrint work in the File Contents Browser. 1331EngLang-tkKV -- Ted Kaehler -- 16 July 1999 Gives programatic access to the WordNet lexicon at Princeton Univ. At http://www.cogsci.princeton.edu/cgi-bin/webwn/ Keep the parts of speech entries as separate streams. Query them for information. 1332definition-sw -- Scott Wallace -- 18 July 1999 Adds a UI in the help... branch of the screen menus allowing you to look up the definition for any English word, using the WordNet lexicon at Princeton Univ." 1333InfFormFix-ar -- Andreas Raab -- 19 July 1999 Fixes two more problem with InfiniteForms" 1334ClassFixes-ar -- Andreas Raab -- 19 July 1999 This change set is only a doIt for fixing several problems in the class hierarchy. It does not fix all them though (this is to dangerous) but it fixes and reports a bunch of them." 1335ClassBuilder-ar -- Your Name -- 9 July 1999 A complete new class for creating or modifying classes. Also does a number of things like: * Fixing the meta class structure for subclasses of nil (e.g., ObjectOut class was a subclass of Object class and not a subclass of Class as it should be) * Changing the instVar Class>>subclasses to use an Array instead of a Set (for compactness; but mainly for allowing #become: on the classes ;) * Of course, using #become: to fix up references to the classes getting recompiled (but very, very carefully ;-) * Changing Encoder>>bindTemp: to only print a warning when non-interactive (very annoying otherwise) * Adding the method #keysAndValuesDo: to the appropriate subclasses of Collection (Dictionary and SequenceableCollection). * Adding a morphic-style progress indication when recompiling classes (which may be turned off such as for the end-user system)" 1336Alice2Anims-jsp -- Jeff Pierce -- 20 July 1999 A first pass at moving animations over to version 2 of Alice." 1337rotCenterFix-sw -- Scott Wallace -- 20 July 1999 Purports to fix the bug that change-costume did not properly take rotation centers into account when positioning the new costume. Also fixes a bug that could interfere with proper running of tile scripts by rotated objects that were not SketchMorphs." 1338InstVarMoves-ar -- Andreas Raab -- 20 July 1999 This change set provides the ability to move instance variables of a living class to another class while keeping its value." 1339debanking-sw -- Scott Wallace -- 20 July 1999 Deletes lots of code relating to an earlier scripting architecture which involved numbered banks rather than categories with textual titles." 1340CGFix-ar -- Andreas Raab -- 20 July 1999 Fixes the code generator not to prune any exported methods (e.g., primitives that are called by name)." 1341SocketFix-ar -- Andreas Raab -- 20 July 1999 Fixes two problems in Socket" 1342customMenufix-sw -- Scott Wallace -- 20 July 1999 Provides a backstop method in CustomMenu so it won't fail when apprised of balloon text." 1343bankFix-sw -- Scott Wallace -- 20 July 1999 Fixes a bug indroduced in update 1339." 1344FlashCodec 1345shiftClickFix-sw -- Scott Wallace -- 21 July 1999 Makes it such that a shift-click in a list pane does NOT give keyboard focus to the item clicked on. Among other things, this allows shift-click on a message name to result in the display of decompiled code in a Browser in Morphic, as it always has done in mvc Browsers." 1346VMStuff-ar -- Andreas Raab -- 16 July 1999 Modifies image reading to use an optional image offset. This is a preparation for embedded images (e.g., images stored somewhere in another file) such as one-click executables. The method #readImageFromFile:HeapSize: has been modified to #readImageFromFile:HeapSize:StartingAt: but for the VM support code a #define is provided in sq.h for backward compatibility. Also makes SystemDictionary>>#snapshot:andQuit: return the resumingFlag so clients can check whether we're just starting up." 1347paintFixes-sw -- Scott Wallace -- 22 July 1999 (1) When an object that is NOT wearing a sketch costume is told to look like another object that IS wearing a sketch costume, the rotation style of the latter is now correctly applied to the new sketch costume of the former. (2) If the sketchMorph currently being painted has been removed from the world (by dismissing its enclosing window, for example) before painting is completed, a fix is made so that the paint palette is properly disposed of and the user is not dropped into a debugger." 1348Alice2OutWin-jsp -- Jeff Pierce -- 25 July 1999 Adds the AliceTextOutputWindow that somehow escaped from the earlier Alice v2 updates." 1349WnldDirLight-jsp -- Jeff Pierce -- 25 July 1999 Adds a 'direction' instance variable to WonderlandDirectional Light. I either forgot to add this variable or it disappeared along the way, but in either case it was causing direction to show up in Undeclared. This change set fixes that." 1350FlashDithering-ar 1351SundryFixes -- Dan Ingalls -- 16 July 1999 Fix to WaveEditor showEnvelope suggested by Matt Rosen. Fix to Semaphore smartInspect suggested by Bob Arning. Enhancement to fileList supporting background images, submitted by Ian Trudel. Fix to Parser temporaries submitted by Bob Arning and Leandro Caniglia. 1352ReshapeClasses-di -- Dan Ingalls -- 18 July 1999 Removes the instVar, 'subclasses' from Metaclass -- partly as a cleanup, -- and partly as an acid test of Andreas's new ClassBuilder. 1353WorldState-di -- Dan Ingalls -- 19 July 1999 Refactors MorphicProject into the old simple Project, and a new class WorldState. This makes worlds within worlds work better, since only one project is involved. 1354WorldState2-di -- Dan Ingalls -- 24 July 1999 Fixes a few obsolete references to PasteUpMorph>>project, and removes it. Then renames the field 'project' to be 'worldState', consistent with the new factoring.

2 4

re: Re: re: Re: Moving objects... (serialization)
by Alan Lovejoy 28 Jul '99

28 Jul '99

> ** Original Sender: Marcel Weiher <marcel(a)system.de> > > From: Alan Lovejoy <sourcery(a)pacbell.net> > > > > To say nothing of selector names. What's the Java/C++ translation of a > > Smalltalk message selector such as > > #displayOn:at:clippingBox:colorPalette:? > > And the inverse problem would also have to dealt with: Smalltalk > wants a > > colon-terminated keyword for each argument. > > Well, since this is about moving *objects* not *code*, there is > actually no problem. But in Smalltalk, code can be objects. Classes and methods are objects. Objects can hold classes and methods in their instance variables--and often do! > > And then there's the universal problem of different class > libraries, with > > each language's (or implementation's, or environment's) idiosyncratic > > set of names, capabilities and semantics of entities and operations. > > Although all of this is beyond the scope of what I was talking > about, having a self-describing format helps, because tools such as > XSL could be used to transform between the serialization formats of > superficially incompatible systems. Deeply incompatible systems will > have to be modified to talk to each other, what did you expect? I expected nothing, of course. I was not the one hoping to develop or use a universal object serialization format, remember? > > How would Java/C++ deal with #doesNotUnderstand:? Or with > "thisContext"? > > Or with a metaclass (which would also have to be dealt with, in > addition to > > classes as objects)? What about #become: and > #changeClassToThatOf: (VW)? > > What about adding/removing/redefining classes and/or methods as the > > program runs? > > All of these are runtime issues that have little or nothing to do > with serialization of objects. Geez, who was talking about magically > turning C++ into a dynamic language via a common serialization > format? Get a grip. These things have everything to do with serialization of objects, if the objects in question are Smalltalk objects (see above). It was not clear from the prior discussion that you wanted to exclude classes and methods from consideration, and simply assume that the receiving system already had all the required classes/methods. I agree that it would be both possible and useful to design a class library that would allow exchange of serialized objects between Smalltalk and Java (or perhaps even C++), provided certain constraints were respected (the Smalltalk objects to be serialized and exchanged with the Java/C++ world must not have instances of certain problematical classes anywhere in their object reference graph). --Alan

2 1

Re: Out of the box apps?
by Dwight Hughes 28 Jul '99

28 Jul '99

In the standard Squeak image you can look in the system categories for "Interface-Web Browser", "Interface-Mail Reader", "Interface-IRC Chat" (indeed, most of the "Interface-..." categories are useful to look at); also check out the "Morphic-Demo" and "Morphic-Games" categories. There is also another tutorial available: "Fun with the Morphic Graphics System - Morphic Tutorial 1": http://www.squeak.org/tutorials/morphic-tutorial-1.html -- Dwight Adriano wrote: > > Hello everybody, > > I just finished the "BankAccount" tutorial. > > Well there's not so much to read further, I never programmed before, so > it's not so easy for me to get oriented. > > I wonder if there are any applications ready to use for me to browse source. > > For example, doesn't a mailer application exist? > > I saw some network and POP/SMTP related classes in the browser, so I guess > someone did developed such apps. > > Where could I find them? > > Is there a newsreader? Is there a web browser? Is there anything else? > > Thanks for your attention > Bye > Adriano > > P.S. I suggest to write a sort of "Inside Squeak" reporting about every > class, with examples and so on.

1 0

Re: Moving objects... (serialization)
by Marcel Weiher 28 Jul '99

28 Jul '99

I haven't looked at OBV specificall yet, but let me reiterate: - ASCII, human readable format - self describing - little or no difference between 'serialization format' and 'interchange file format' - in fact, both could be identical in many cases - tools for purely syntactic manipulation (rule transformations etc.) - ... Moving behavior would also be an interesting subject. My approach would be to serialize objects representing the expression tree. Marcel > From: "Peter Smet" <peter.smet(a)flinders.edu.au> > > Marcel, > > If you want a cross platform way of moving object state (but not behaviour) > then doesn't the OMG have a specification of Objects By Value that > standardizes this? How is XML better than OBV?

1 0

Thoughts on moving objects, XML and namespaces etc...
by "Gran" Hultgren 28 Jul '99

28 Jul '99

Hi all! A few thoughts (after going through this list after my vacation): 1. Go, go, Squeakers! The activity on the list seems to be growing and more names pop up by the day. Very nice! 2. A few things that really NEED to be done in Squeak ASAP IMHO: Exception handling and good networking code · la Correspondents. There is a fast growing amount of code out there and these things really are essential for most of the work in Squeak. Exceptions because everyone needs them and networking code because many of the projects that can use Squeak today depend on it. 3. Before throwing in Namespaces i Squeak though... Is it only me that feels that they really "get in the way" when programming Java? Import this and import that, oops, old imports laying about... Gah! Of course namespaces are cool. And useful. But they also break the "modelessness" that Smalltalk has. You can always hack a oneliner in any of Smalltalks textareas (inspectors, debuggers, browsers, workspaces etc.) and just doit. The only context you have to think about is your "current obect" which is natural. But in Java I always have to qualify myself to death and it is truly annoying. I glanced through Allan's PDF-document but I did not see any reference to how he thinks namespaces will affect our lovingly productive and effective Smalltalk environment and how we work with our code/objects, anyone have any thoughts? Allan? 4. XML is very nice. I work with it in Java and we are using it right now both as a more robust way of representing objects in files than serialization. If you do it right and couple XML with some nice "builder" objects you can isolate your testdata better from the evolving class schemas. Another perfect fit for XML is representing UIs as Mozilla are doing in their Xul stuff. VWs windowsSpecs together with builder objects were in retrospect "the right way to go". :-) I hate UI code generation and I do not really believe in UI serialization as Sun seems to do. And hacking the UIs by hand is tedious and results in code bloat. Go XML. But serialization of objects between languages is hard. HARD on the brink of impossibility I would presume. OMG/CORBA are trying aren't they? I say good luck. :-) If you want more inspiration on the subject of moving objects, take a good long look at Voyager from ObjectSpace. Truly cool technology. RMI deluxe with mobile agents on top. Something like that would really be cool in Squeak. I have not looked at the fly-by-wire stuff, is it something like Voyager? And finally how truly wonderful it would be to have some XML package in Squeak. :-) regards, Gran === Gran Hultgren, gohu(a)rocketmail.com, icq#:6136722 GSM: +46 709 472152, http://195.22.65.4 "First they ignore you. Then they laugh at you. Then they fight you. Then you win." -- Gandhi _________________________________________________________ DO YOU YAHOO!? Get your free @yahoo.com address at http://mail.yahoo.com

2 1

RE: Category all in 2.4
by Norton, Chris 28 Jul '99

28 Jul '99

Hi Richard. Thanks! Carl & I appreciate the compliment. Squeak On! ---==> Chris > -----Original Message----- > From: R. A. Harmon [SMTP:harmonra@webname.com] > Sent: Tuesday, July 27, 1999 4:10 PM > To: squeak(a)cs.uiuc.edu > Subject: Category all in 2.4 > > I really appreciate the addition in 2.4 of the all category in the > browser. > I didn't realize what I was missing until I had it. Thanks. > > -- > Richard A. Harmon "The only good zombie is a dead zombie" > harmonra(a)webname.com E. G. McCarthy

1 0

Category all in 2.4
by R. A. Harmon 28 Jul '99

28 Jul '99

I really appreciate the addition in 2.4 of the all category in the browser. I didn't realize what I was missing until I had it. Thanks. -- Richard A. Harmon "The only good zombie is a dead zombie" harmonra(a)webname.com E. G. McCarthy

2 1

Fwd: Quesa 0.0.2
by Lawson English 27 Jul '99

27 Jul '99

[this is from the Quesa mailing list. Not sure if anyone is really interested, but as I pointed out earlier, a collaboration between Squeak and Quesa 3D people would be of benefit to both camps. The Quesa people are professional 3D applications programmers in need of a cross-platform replacement for QD3D, while the Squeak community has a strong multi-platform development environment that would benefit greatly from using a professional-level 3D library with a compiled-speed image database that was available cross-platform] Hi, For those who are interested, Quesa 0.0.2 is now available for download. The status so far is: Header Functions Done %ge -------------------------------------------- QD3D.h 35 35 100% QD3DCamera.h 43 43 100% QD3DController.h 47 0 0% QD3DCustomElements.h 9 0 0% QD3DDrawContext.h 52 32 61% QD3DErrors.h 8 8 100% QD3DExtension.h 16 6 37% QD3DGeometry.h 315 51 16% QD3DGroup.h 38 0 0% QD3DIO.h 103 0 0% QD3DLight.h 45 45 100% QD3DMath.h 133 119 89% QD3DPick.h 34 24 70% QD3DRenderer.h 49 36 73% QD3DSet.h 21 16 76% QD3DShader.h 38 10 26% QD3DStorage.h 26 0 0% QD3DString.h 6 6 100% QD3DStyle.h 46 46 100% QD3DTransform.h 49 49 100% QD3DView.h 48 38 79% QD3DViewer.h 57 0 0% QD3DWinViewer.h 47 0 0% -------------------------------------------- Total: 1265 564 44% You can download this release, and find out how to subscribe to the Quesa mailing list, from the Quesa home page: <http://quesa.designcommunity.com/> A more detailed change list will be posted to the Quesa mailing list. -dair ___________________________________________________________ (Work) mailto:dair@lightwork.com http://www.lightwork.com/ (!Work) mailto:dair@webthing.net http://www.webthing.net/

2 1

Re: Exceptions
by Dan Ingalls 27 Jul '99

27 Jul '99

Benoit St-Jean <lamneth(a)aei.ca> wrote... >Just curious, > >Anybody heard about new developments regarding Exceptions? Will they be >implemented in Squeak 2.5? If so, will they follow the ANSI draft? Can we >use Craig Latta's exception package in the meantime (till exceptions make >it in the image) or another exception package (TFEI, Vassili Bykov, etc). > >What's the status on this long debated issue? Exceptions will not be in 2.5. however, as soon as 2.5 is out (hopefully later this week (I got called out of town both of the last two weeks)) I'm hoping we can organize ourselves around half a dozen major improvements including exception handling. I'll send out a message concerning this today. To quote from my mesage of 7/14, which I'm still following... >1. I am NOT going to deal with the big projects [incl exceptions], some of which I listed in my first message on this topic (but keep reading). Me feeling is that these will take time to sort out appropriately. > >2. I AM going to get out a 2.5 release as soon as I can stop communicating and get to work. > >3. At the time of the 2.5 release, a new version will also be declared on the update servers called, eg, 'Squeak2.6test'. As before, bug-fixes and no-brainer enhancements will be issued for normal users of the 2.5 release. Anyone who wishes to be a test pilot can simply advance their version to Squeak2.6test and fetch updates again. Knowing that only test pilots will receive them, we can cheerfully issue the latest updates weekly (I will still buffer them, since we do occasionally foul up). > >4. Regarding the larger projects that Squeak needs but that we are not controlling, I want to try a new approach. A couple of people have suggested voting. I'm not really in favor of this and, what's more, it doesn't address the real problem. My proposal is that we identify a list of such projects (we have a start already) and then, for each one, identify a group of 3 to 5 competent people to carry it through to completion. I'm going to suggest that one of these people be from Squeak Central, even if they don't actually do any coding. This will ensure that any concerns we have will get addressed, that we can track the progress, and that whoever is doing the hard work will know there is buy-in at Squeak Central from the start. As used above, "to completion" means that it's ready to be issued as an update to the test pilots.

1 0

← Newer
1
...
4
5
6
7
8
9
10
...
37
Older →

Jump to page: