Squeak-dev October 2000

squeak-dev@lists.squeakfoundation.org

112 participants
424 discussions

[SSS] Re: Squeak on Amazon!
by Lex Spoon 02 Oct '00

02 Oct '00

Henrik Gedenryd <Henrik.Gedenryd(a)lucs.lu.se> wrote: > >> much-anticipated "The Battle of Bubble and Squeak" books... > > Well, unless Squeak becomes syntax-compatible with Python, gains an > industrial-strength CORBA implementation, native widgets for BeOS, and a > Comanche version faster than Apache, then Bubble will surely win and Squeak > will DIE!!! Heheh, forgot the [SSS] tag. -Lex

1 0

Re: SqueakOnLinuxFramebuffer getStatus.
by Jecel Assumpcao Jr 02 Oct '00

02 Oct '00

On Sun, 01 Oct 2000, Aaron J Reichow wrote: > On Sun, 1 Oct 2000, Karl Ramberg wrote: > > A link to the SqueakNOS, a no os Squeak project. Seems feasible and > > more ram efficient than putting Squeak on top of the linux kernel. > > It'd also be a lot more work, re-writing drivers for all that Linux > already supports. Using Squeak on top of the Linux fb (not X), you can > put a Squeak OS on any architecture Linux already runs on, whereas > SqueakNOS would require writing drivers/porting for the various > hardware framebuffers, video cards, processors, network cards, keyboards > and mice (USB, PS/2, &c), sound, serial, so on and so forth... You can borrow drivers from Linux and BSD using the OSKit: http://www.cs.utah.edu/flux/oskit/index.html -- Jecel

3 2

RE: AttachmentMorph design questions
by spair＠advantive.com 02 Oct '00

02 Oct '00

> As I see it, such a Morph would be connected to two (or more?) Morphs > and would maintain some kind of spatial constraint between the points > at which it was connected. A simple version would make sure that the > two attached Morphs were attached at the same point. Is this related to Fabrik? Could the connectors in a Fabrik like system could use AttachmentMorphs for implementing connections. It would be interesting to see if the constraints could be made pluggable such that new and interesting things (things akin to Fabrik for example) could be done with AttachmentMorphs. Would the constraint behavior be best in a different class whose instances are used by AttachmentMorphs? - Stephen

2 1

Re: Squeak on Amazon!
by Henrik Gedenryd 02 Oct '00

02 Oct '00

>> much-anticipated "The Battle of Bubble and Squeak" books... Well, unless Squeak becomes syntax-compatible with Python, gains an industrial-strength CORBA implementation, native widgets for BeOS, and a Comanche version faster than Apache, then Bubble will surely win and Squeak will DIE!!! H

1 0

Re: openInWorld vs activate
by Lex Spoon 02 Oct '00

02 Oct '00

I think the problem is that SystemWindow's think they are a main part of the UI, and so they try to initially position and resize themselves depending on what windows are already open and how big the display is. Arguably, SystemWindow's should *not* do this if their extent has been set explicitly. It's a tricky area. One thing you can try is to give the window a model and override that model's #initialExtent method. This will work as long as you only have one window per model, and as long as your model can come up with a reasonable initial size. :| -Lex David Farber <dfarber(a)numenor.com> wrote: > > Here's a question for all you morphic gurus out there: > > Why does > > aMorph openInWorld > > behave differently from > > World addMorph: aMorph. > aMorph activate. > > (or, more to the point, why does the former not behave like the latter.) > To illustrate further, compare the behavior of these two snippets when > execute at the end of the following: > > | alignMorph win | > alignMorph := AlignmentMorph new. > alignMorph > centering: #center; > hResizing: #shrinkWrap; > vResizing: #shrinkWrap; > orientation: #vertical. > > alignMorph addMorphBack: (BasicButton new label: 'a button'). > alignMorph addMorphBack: (BasicButton new label: 'this is a button'). > alignMorph addMorphBack: (BasicButton new label: '''nother button'). > > alignMorph fullBounds. > > win := SystemWindow labelled: 'fooMail console'. > win bounds: ((alignMorph position - > ((0@win labelHeight) + win borderWidth)) > corner: alignMorph bottomRight + win borderWidth + > (2@2)). > win addMorph: alignMorph frame: (0@0 extent: 1@1). > > "World addMorph: win. > win activate" > > "verses" > > "win openInWorld" > > When I use win activate, i get my buttons-inside-an-alignment-morph all > wrapped in a system window. When I use openInWorld, the system window > reverts to some default size. > > What's the scoop? > > thanks, > david > > -- > David Farber > dfarber(a)numenor.com

1 0

Writing Flash Morphs on a SmartRefStream on a RWBinaryOrTextStream
by Wouter Gazendam 02 Oct '00

02 Oct '00

Hi All, I seem to have a problem writing FlashMorphs to a RWBinaryOrTextStream. Writing to a file works: | outputStream inputStream | outputStream := FileStream newFileNamed: ('test' asFileName). outputStream fileOutClass: nil andObject: MyFlashMorph. inputStream := FileStream fileNamed: ('test' asFileName). inputStream fileInObjectAndCode openInWorld But when I replace the FileStream with a RWBinaryOrTextStream: | stream | stream := RWBinaryOrTextStream on: (ByteArray new). stream fileOutClass: nil andObject: MyFlashMorph. stream reset. stream fileInObjectAndCode openInWorld The thingy results an error: 'Error: Expected start of object, but found 0' I thought the RWBinaryOrTextStream was interchangeable with the FileStream. Any help/suggestions/rants would be welcome. Thanks, Wouter Gazendam

1 1

Re: [updates] 29 more for 2.9alpha
by Bruce ONeel 02 Oct '00

02 Oct '00

Hi, These changes are here: <ftp://st.cs.uiuc.edu//Smalltalk/Squeak/2.9_alpha/Squeak2.9a-2748.zip> cheers bruce Dan Ingalls <Dan.Ingalls(a)disney.com> wrote: > Folks - > > Here is another bunch of SqC updates. Many of these are smallish tweaks as we try to converge on a stable internal for use in a school this fall. I also now have a list of the "approved" Celeste fixes from the summer, which I will try to get out in the next day or so. > > You may notice the numbering skips from 2710 to 2720. Not to worry. We just moved to non-consecutive updates internally, and this caused a wrinkle in my export process. > > Enjoy > - Dan > --------------------- > 2720NoChameleons-di -- DanIngalls -- 27 September 2000 > Enhances the direction handles so as to use colors that are not too similar to the underlying morph. > > 2721ColorPickerTweaks-di -- Dan Ingalls -- 27 September 2000 > Why have a preference for modal vs modeless (pop-up vs enduring) color pickers? > This changeSet allows you to anchor any pop-up color picker analogously to anchoring a pop-up menu. Just click or drag on the little drag-dot, and you have an enduring color picker. > But wait. There's more... > Introduces balloon help for the different control areas (even during modal operation). > A color selection that begins in the color area can only affect color; one that begins in the translucency control can only affect translucency. > Oh, and... > Also includes a change that bundles a restore-cursor into the restore-display choice in the world menu. Simpler than exiting the project. > > 2722textFix-raa -- Bob Arning -- 27 September 2000 > - initialize a couple of BitBlt vars so that the string display primitive will not fail" > > 2723betterThumbGIF-raa -- Bob Arning -- 27 September 2000 > - fix project publish failures when thumbnail has more than 256 colors" > > 2724geeGolly-raa -- Bob Arning -- 27 September 2000 > - fix red menu so that sketches in transformtions will show the GeeMail submenu when appropriate > - insure project renamer not visible before making thumbnail for project save directory dialog > - removed some obsolete methods/classes/variables > - fixed up parsing of old project names which contained periods" > > 2725PerfectCircles -- Andreas Raab -- 27 September 2000 > Fixes a problem when removing some flexing wrapper. The wrapper did not align the flexed object by its reference position leading to tiny offsets depending on width and height of the wrappee. Also improves #referencePosition: a bit to prevent the generation of float points when unnecessary." > > 2726RemoveFractionalPosition -- Andreas Raab -- 27 September 2000 > Oh well - I had removed the fractional position stuff already before posting the last CS... so here it comes." > > 2727clipCanvas-raa -- Bob Arning -- 27 September 2000 > - ensure that the canvas used by the World is clipped to the bounds of Display so that time is not wasted drawing things that are off-screen > - nudge ProjectNavigatorMorph a little further off-screen to minimize cycles consumed redrawing" > > 2728moreProj-raa -- Bob Arning -- 27 September 2000 > - start cleaning up different ways of enumerating Projects > - cleanup some old refs in ProjectHistory" > > 2729WorldHaloTweaks-di -- Dan Ingalls -- 27 September 2000 > Two little changes with big impact for halos in the world. The first ensures dismissal of all halos (including world halos) when you click in the world. > The second prevents world halos from noting the entire screen as damaged. For this reason, there is much less delay when dismissing the world halo. > > 2730TranslucencySlider-di -- Dan Ingalls -- 28 September 2000 > Adds a slider effect to the translucency control of the color picker. Fixes a couple of glitches in moving from modal to modeless operation. > > 2731SmallerArrow -- Andreas Raab -- 28 September 2000 > Per Kims request - makes the direction arrow just a little smaller." > > 2732MessageSend-value -- Michael Rueger -- 28 September 2000 > Enables MessageSend>>value to send messages with arguments." > > 2733flash-fixes -- Michael Rueger -- 28 September 2000 > Fixes a bug in storeColor:on: and forces the computation of the pixel ramp at read time. The lazy computation of pixel ramps interfers with animations." > > 2734StreamFileIn -- Michael Rueger -- 26 July 2000 > Moves some methods up the stream hierarchy so fileIn may be performed on internal streams." > > 2735Sound -- John Maloney, Michael Rueger -- 27 July 2000 > This change set adds > - messages to sound readers enabling them to read from any stream, not just a file. > - two new sounds marimba and pluckedElecBass > - messages for settings in the ScorePlayer." > > 2736http-url -- Michael Rueger -- 26 July 2000 > Cleans up some minor inconsistencies. > Adds methods to specify the mimeType in the getResponse calls. > Setting a proxy in Squeak usually keeps you from connecting to local hosts within your domain. This change set adds a list of exempt host or domain names to HTTPSocket, which should not be contacted through the proxy. > Adds a method to build a list of URL args without encoding the arguments. Usefull if you have args already containing special characters. > Add also a class URLArgumentList which does not change the order of arguments. > > 2737FileStream-exceptions -- Michael Rueger -- 26 July 2000 > This change set introduces exceptions for the handling of file exists/does not exist exceptions so applications can handle these their own way. The default action still pops up the well known menus, now moved to StandardFileStream error-handling. > See also FileDoesNotExistException>>example." > > 2738http-fix -- Michael Rueger -- 28 September 2000 > The check for a String as the return value of a getDocument call was done too late." > > 2739HeadingTweak-ar -- Andreas Raab -- 28 September 2000 > A little tweak that prevents flexing of sketches immediately after being created." > > 2740FlapsAndNavigator -- Andreas Raab -- 28 September 2000 > Makes flaps and the project navigator stay at the same layer in projects, e.g., in front of a lot of other things but keeping the same relation to each other." > > 2741GIFfix-ar -- Andreas Raab -- 28 September 2000 > Make the GIFReadWriter accept > 8 bit forms and convert it silently. If somebody is trying to store a 16 or 32 bit form it should work even if it costs some colors." > > 2742KidNavPlus-raa -- Bob Arning -- 28 September 2000 > - adds balloon help to the kid navigator to help understand the pop out/in behavior > - adds option to warn user before quitting Squeak if there are unsaved projects > implementation is a bit fuzzy at the moment. > Preferences - checkForUnsavedProjects in the general category" > > 2743EHReturnFix-ar -- Andreas Raab -- 28 September 2000 > Fixes a problem with return values from default actions." > > 2744PDATweaks2 -- Dan Ingalls -- 29 September 2000 > Fixes a postscript problem and better handling of CRs in the bottom pane. > > 2745SqR-HashFix -- Andres Valloud -- 18 September 2000 > Fix the new hash for SequenceableCollections so that it never overflows. It rarely occurs, but it should never happen" > > 2746restartLoop-raa -- Bob Arning -- 29 September 2000 > Reiterate the removal of <exitFlag>, then ensure world loop restart (in the postscript)." > > 2747misc-tk -- Ted Kaehler -- 26 September 2000 > 1) SequenceableCollection.hasEqualElements: caused an error if the argument was not a Collection. Called by RunArray.= > 2) DamageRecorder.recordInvalidRect: corrects a spelling error in a comment. > 3) ServerDirectory.checkServers fixed an error when it is detected that the updates.list file does not match on the two servers. (It happened for the first time recently!) > 4) Upon fileOut, there is a dialog box that asks if you want to see the methods with slips in them. Reversed the order of the choises, and changed the wording. > 5) Provided the reverse conversion method so that Projects written with no 'exitFlag' variable can be read into older BJ images." > > 2748http-fixes -- Michael Rueger -- 29 September 2000 > Fixes a recently introduced problem with http get where in the get the hostname was passed and not just the full path."

1 0

RE: How to make Morphs that only appear when the mouse is overthe m?
by Pennell, David 02 Oct '00

02 Oct '00

Why not make all AttachmentMorphs that are attached to a morph become visible when the mouse is over the attached morph? Other graphic programs that I've used seem to use this approach. -david > -----Original Message----- > From: Dan Ingalls [mailto:Dan.Ingalls@disney.com] > Sent: Sunday, October 01, 2000 1:07 PM > To: Ned Konz > Subject: Re: How to make Morphs that only appear when the mouse is > overthem? > > > >I'm working on my AttachmentMorphs, and am trying to add a mode where > >they won't be visible unless the mouse is over them. > > > >But using mouseEnter: and mouseLeave: to control visibility doesn't > >work, since the mouseOver logic in the Hand ignores invisible Morphs. > > > >I guess I could use step and query the cursor position every time, > >but is there a cleaner way? > > > >(Actually, using step would let me make them visible when the mouse > >was merely near them. This might be better). > > Ned - > > You might want a global command (like to any pasteUp or the > world) to make all attachers visible. I only mention this > beacuse of, eg, the make-visible command in many text editors > that lets you see all non-printing characters. This means > you needn't be bothered by the attachers when you don't want > them, and you don't have to fish for them when you do. > > Mouse-over and proximity are certainly equally reasonable > preferences to play with, though. > > - Dan > >

2 2

Re: How to make Morphs that only appear when the mouse is overthem?
by Ned Konz 02 Oct '00

02 Oct '00

"Andrew P. Black" wrote: > > At 10:40 -0700 2000.10.1, Ned Konz wrote: > > > > > Mouse-over and proximity are certainly equally reasonable > >preferences to play with, though. > > > >Mouse-over turns out to be easier than proximity (which requires computation > >every step time). > > Couldn't you get the "proximity" effect using mouse-over, by just > adding an invisible border around the morph, i.e., making it bigger > thanit appears on the screen? Probably. I think that highlighting the attachment points on the target of a connection should be sufficient for now. -- Ned Konz currently: Stanwood, WA email: ned(a)bike-nomad.com homepage: http://bike-nomad.com

1 0

Squeak on Amazon!
by Mark Guzdial 02 Oct '00

02 Oct '00

Thanks to Jeff Pierce for sending me this! (This is almost as exciting as actually having the book come out :-) http://www.amazon.com/exec/obidos/ASIN/0130280283/o/qid=970306184/ What's funny is to do a search on the word "Squeak." Yes, my book comes up -- along with about a dozen *Disney* books with "Squeak" in the title. My book is on the "To come out list" along with the much-anticipated "The Battle of Bubble and Squeak" books... Mark -------------------------- Mark Guzdial : Georgia Tech : College of Computing : Atlanta, GA 30332-0280 Associate Professor - Learning Sciences & Technologies. Collaborative Software Lab - http://coweb.cc.gatech.edu/csl/ (404) 894-5618 : Fax (404) 894-0673 : guzdial(a)cc.gatech.edu http://www.cc.gatech.edu/gvu/people/Faculty/Mark.Guzdial.html

2 1

Jump to page: