Thanks

These are now all on

http://box1.squeakfoundation.org/files

or

ftp://box1.squeakfoundation.org

though I'm note sure that the ftp server is quite ok yet.

cheers

bruce
Michael Rueger <michael at impara.de> wrote:
> Date: Wed, 23 Feb 2005 15:12:28 +0100
> From: Michael Rueger <michael at impara.de>
> Subject: [Updates] 19 updates posted
> To: Squeak <squeak-dev at lists.squeakfoundation.org>
> organization: impara.de
> reply-to: The general-purpose Squeak developers list <squeak-dev at lists.squeakfoundation.org>
> content-length: 9618
> 
> Hi all,
> 
> just posted 19 updates to 3.8 and 3.9.
> 
> All these updates are from the squeakland update stream.
> 
> Michael
> 
> 
> --------------------
> 
> 
> "Change Set:		FontFixupSpeedup-nk
> Date:			8 November 2004
> Author:			Ned Konz
> 
> This CS makes loading of projects only pay the price for font fixup if 
> they actually had missing fonts.
> 
> On my system, a particularly complicated project with 278 
> ScriptEditorMorphs took the following time to load and enter:
> 
>          load   enter
> before  32       14
> after    29       3
> 
> This is important because on entry there is no indication that anything 
> is actually taking place, and so it looks dead for the 14 seconds.
> 
> (shouldn't we add some kind of indication that it's still working?)
> "
> 
> "Change Set:		FontProgress
> Date:			1 December 2004
> Author:			Andreas Raab
> 
> The 'lost' displayFontProgress method."
> 
> "Change Set:		removedMethFix-sw
> Date:			1 December 2004
> Author:			Scott Wallace
> 
> Provides the missing method PseudoClass.stampAt: that was needed to make 
> the improved removed-method reporting in a file-contents-browser in 
> update 0368tidbits-sw correct."
> 
> "Change Set:		allScriptsButtons-sw
> Date:			8 December 2004
> Author:			Scott Wallace
> 
> Restores the look of the dismiss, expand, and help buttons on the All 
> Scripts Tool to their appearance last year.  And the corresponding 
> buttons in the 'gallery of players' are made to conform to this 
> appearance as well."
> 
> "Change Set:		TTCFontSetReg
> Date:			2 December 2004
> Author:			Yoshiki Ohshima
> 
> Instead of using #allInstances in #familyName:pointSize: it now sets up 
> a registry and search fonts from there.  Also fixes some loose ends."
> 
> "Change Set:		ProjectLoad-tk
> Date:			2 December 2004
> Author:			Ted Kaehler
> 
> Clean up the code that load a Project from disk or a server.
> Improved class comments in Project and ImageSegment.  Speed of
> loading is only modestly better.
> 
> Provides a new way for objects in a Project to adjust themselves when
> they enter a different version of the system.  After a project is
> loaded, every object is sent fixUponLoad: aProject refStream:
> smartRefStrm.
> 
> Just to review:
> 	comeFullyUpOnReload: smartRefStream		Used to
> re-discover an object that already exists in this image, such as a
> resource, global variable, Character, or Symbol.  (sent to objects in
> outPointers)
> 	convertToCurrentVersion: varDict refStream: smartRefStrm
> 		fill in fields that have been added to a class since
> the object was stored.  Used to set the extra inst var to a default
> value.  Or, return a new object of a different class.  (sent to
> objects that changed instance variables)
> 	fixUponLoad: aProject refStream: smartRefStrm
> 	change the object due to conventions that have changed on the
> project level.  (sent to all objects in the incoming project)
> "
> 
> "Change Set:		corruptScriptFix-sw
> Date:			15 December 2004
> Author:			Scott Wallace
> 
> Fixes up certain damaged etoy projects -- request 'attempt misc repairs' 
> from the authoring-tools menu to invoke.
> *  uniclassScripts that have nil scriptEditors are expunged.
> *  uniclassScripts that have nil scriptedClasses are fixed up.
> *  scriptEditors whose scriptNames conflict with the selector of the 
> uniclassScript that holds on to them are fixed up."
> 
> "Change Set:		EnableKeystrokeEvent-nk
> Date:			13 December 2004
> Author:			Ned Konz
> 
> CS 0359keyboardEvents-nk did not register the keystroke event trigger 
> properly (or rather, 0367specificEvents-nk didn't change 
> PasteUpMorph>>initialize as it should).
> 
> This change set restores the 'keyStroke' event for PasteUpMorphs 
> including the World.
> 
> "
> 
> "Change Set:		dismissFix-sw
> Date:			22 December 2004
> Author:			Scott Wallace
> 
> If the user removes a phrase from an etoy scriptor by using the dismiss 
> handle on the halo of a PhraseTileMorph, the needed script-recompilation 
> notification will now take place."
> 
> "Change Set:		balloonCenter-wiz
> Date:			31 December 2004
> Author:			Jerome Peace
> 
> Fix for Mantis bug 783: Balloon Morph text goes outside of balloons when 
> lots of text.
> 
> Large balloons such as that for the trash can are not centered in the 
> balloon.
> adjustedCenter tried to fudge it.
> This fix returns the original center of the textMorph within the balloon 
> by selecting the vertices that represent the morphs corners and finding 
> their center. A perfect fit every time."
> 
> "Change Set:		visibleThumbnail-sw
> Date:			29 December 2004
> Author:			Scott Wallace
> 
> Assures that the thumbnails of objects with extreme aspect ratios do not 
> vanish.."
> 
> "Change Set:		sibHalo-sw
> Date:			2 January 2005
> Author:			Scott Wallace
> 
> *  The greenish top-right-halo handle, which may be either the olive 
> 'make-sibling' or the bright green 'make-deep copy (duplicate)' handle, 
> will now, when depressed while the *shift* key is down, trigger the 
> 'other' action, i.e. shift click on olive handle yields make-deep-copy, 
> and shift-click on bright green handle yields 'make-sibling.'
> *  A further visual cue is provided when operating a green handle:  the 
> color of the handle while dragging the resulting sibling or deep-copy 
> will reflect the actual kind of copy made, which may be different from 
> the original color of the handle if the shift key were found to be down.
> *  Whether or not a scripted player will show the olive handle in the 
> halo is now governed by a new preference, 
> #oliveHandleForScriptedObjects, which is automatically set to false when 
> the eToyFriendly preference is set to true.
> *  Also adds, to the morph halo's 'siblings' submenu, for scripted 
> objects, a new item, 'flash siblings', which will momentarily flash an 
> object and all of its siblings.
> This version of this fileout intended for Squeakland3.8"
> 
> "Change Set:		literalChg-sw
> Date:			5 January 2005
> Author:			Scott Wallace
> 
> Recovery from the recent language change that made '#(nil) first' parse 
> to nil rather than to #nil, and similarly for #(true) and #(false). 
> This update restores the pre-existing semantics of several methods 
> affected by that change."
> 
> "Change Set:		smartTileMorph5-tak
> Date:			4 January 2005
> Author:			Takashi Yamamiya
> 
> Cleanup of mouse=handling in TileMorph and its subclasses, including:
> Click on current-value for any type with enumerated textual values (and 
> also on numeric assignment tiles) uniformly brings up a pop-up of the 
> list of alternatives.
> - Fix for http://bugs.impara.de/view.php?id=189 (floatPrecision was not 
> saved when extension arrow was hit).
> - Fix for bug that you could type in to the readout portion of a boolean 
> readout tile.
> - Some redundancies in SymbolListTile removed.
> This version is for Squeakland 3.8.
> 
> A couple of small fixes made and some changes to the preamble wording 
> done by sw 04Jan05"
> 
> "Change Set:		customizeFix-yo
> Date:			6 January 2005
> Author:			Yoshiki Oshima
> 
> Fixes a bad typo in StandardScriptingSystem method #customizeForEToyUsers:"
> 
> "Change Set:		muchBetterWatchers-sw
> Date:			6 January 2005
> Author:			Scott Wallace
> 
> Enhancements to etoy watchers:
> * Readouts for all watchers are now type-appropriate.  Up until now, all 
> readouts had been textual; now for a color type you'll get a watcher 
> showing a color swatch, and so forth down through all the available types.
> *  So-called 'fancy watchers' are now allowed on variables of any type, 
> not just numbers, and, as for the simple watchers, the readout supplied 
> is type-appropriate.
> *  The *name* of a watcher now reveals its provenance.  Formerly, it was 
> extremely difficult to figure out what a simple watcher was watching. 
> And the watcher is dynamically updated to track the name of the object 
> it is watching.
> *  When you drop a slot-retriever phrase (such as car's x) on the 
> desktop, if the new preference #dropProducesWatcher is true, a watcher 
> will be produced.  Try it, you may like it!
> *  A watcher can be dropped into a type-compatible tile pad in a script.
> *  Some adjustments made to which data-types can be ascribed to 
> user-defined variables."
> 
> "Change Set:		bounceSoundFix-sw
> Date:			10 January 2005
> Author:			Scott Wallace
> 
> Fixes the bug that made execution of an etoy #bounce: raise a 
> deprecation warning.  This bug had resulted from update 
> 5747soundSystemCleanup-gk, dated 23Feb2004"
> 
> "Change Set:		watcherDefaults-sw
> Date:			12 January 2005
> Author:			Scott Wallace
> 
> Restores former watcher feel:.
> * Wordings 'simple watcher' and 'detailed watcher' (as opposed to 
> 'unlabeled watcher' and 'labeled watcher') restored.
> *  The dark blue color restored.
> *  The preference that governs whether the drop of a getter produces a 
> watcher or not, dropProducesWatcher, is set to false by the postscript 
> of the update.
> *  Bug fix (Mantis Squeakland bug 824):  a wacher can again be dropped 
> in a 'test' area, whereupon it will expand into the appropriate relation 
> as usual.
> *  Also fixes long-standing bug (dating back to at least 2001) that 
> mishandled the coercion of typed getters for variables of any SymbolList 
> type)
> *  Also fixes bug that arrived with update 0383smartTileMorph5-tak, in 
> which the code for a Scriptor was not getting recompiled when a 
> relational tile (such as = or <=) had its relation changed by direct 
> manipulation, either by operating one of its carets or by choosing a new 
> value from a pop-up obtained by clicking on the tile."
> 
> "Change Set:		suppressSounds-sw
> Date:			29 January 2005
> Author:			Scott Wallace
> 
> Allows certain of the standard 'built-in' sounds to be suppressed from 
> sound-choice popups except when they are the current choice.  This 
> affects both SoundTiles and SoundReadoutTiles.
> 
> The list of formerly-standard sounds that should be suppressed is 
> hard-coded in method StandardScriptingSystem.soundNamesToSuppress; for 
> now, these are 'scrape' and 'peaks'."