Squeak-dev March 2004

squeak-dev@lists.squeakfoundation.org

179 participants
629 discussions

[ENH] Accufonts changeset candidate for 3.7alpha
by Doug Way 15 Apr '04

15 Apr '04

With the help of a bit of handy code from Ned, I've converted the Accufonts.sar file on SqueakMap with all of its bundled .sf2 font files into a single changeset which could be added to the update stream. The entire changeset is just over 100K, not too bad, although a little too large to post to this list, so I put it here: http://www.mindspring.com/~dway/smalltalk/InstallAccuFonts.8.cs.gz This is the candidate for inclusion in the 3.7alpha update stream. Note that the main point of the changeset is to remove the original Apple fonts (NewYork) from the image, so that the font clause in Squeak-L becomes irrelevant, allowing the possibility of making progress with sublicensing (or even re-licensing) Squeak. The replacement fonts look very similar so there shouldn't be much that's noticably changed. As a bonus, several extra bitmap fonts are thrown in. This also removes the ComicPlain/Bold bitmap fonts, which are not free font and were removed so that potential business users of Squeak have a "clean" image to work from font-wise. On the other hand, apparently there is some legal precedent recently that bitmap fonts are not copyrightable, so this may or may not really be an issue. Anyway, "Accujen" is a reasonable looking sans-serif replacement for the Comic fonts. This changeset does not remove the ComicSansMS TrueType font (yet), which probably should be removed, because outline fonts definitely are copyrightable, and this is not a free font either as far as I know. However, I was going to wait and possibly remove this one as part of incorporating the BitStream Vera TrueType fonts, so we have a bunch of replacement TT fonts before getting rid of this one. The "Atlanta" bitmap font, which someone donated to Squeak awhile back, remains. The Accufonts come with their own (very simple) license, included in the changeset preamble. This seems fine, fonts are a very different entity from source code, and it is not uncommon for software to provide a separate license for included fonts. There are still a couple of smallish issues with these fonts, but I wanted to get these in soon so that people can try them out, and we can fix/tweak things based on feedback from that. The problems I know of are: The character widths for some punctuation characters could use some adjusting... e.g. the Accuny period character is quite wide and looks funny IMHO. Also, some fonts such as Accuny09 and Accuny10 have the same size of 12, which can cause problems for some tools, such as Boris Gaertner's useful FontEditor tool. One odd last thing... I notice that the "Smalltalk garbageCollect" at the end of the postscript does not remove the old fonts from "StrikeFont allInstances", but doing a manual garbageCollect after filing in the changeset does. Maybe the ChangeSet object is somehow holding onto the fonts up 'til then, or something. No biggie I guess, we'll make sure garbage is collected sometime before 3.7 goes final. ;-) - Doug

4 7

[FIX][BUG] Refactoring Browser - extract to component ( [er][sm][et] does as advertised )
by frank＠crystal-objects.com 13 Apr '04

13 Apr '04

(and i found a new menu in RB to boot)

3 2

[UPDATES] 57 for Squeak3.7alpha
by Doug Way 11 Apr '04

11 Apr '04

Hm, I posted this Sunday and the list seems to be back up now, but I don't see it, so here it is again... -------- Original Message -------- Subject: [UPDATES] 57 for Squeak3.7alpha Date: Sun, 29 Feb 2004 20:07:21 -0500 From: dway(a)mailcan.com To: <squeak-dev(a)lists.squeakfoundation.org>, <dway(a)mailcan.com> The latest batch of updates... whew. (This one seemed to take longer than usual for various reasons, probably 5-6 hours.) Lots of goodies here, including the SM2 update, PNG writing, snapshot cleanup, Genie removal, etc. We're still shooting for moving to beta this Friday, March 5th. I'll send out a separate email making more of an announcement about that. - Doug ------------------------------ 5708SM1Removal-gk -- Göran Krampe -- 12 February 2004 This update removes old SM1 from the image and stores the installedPackagesDictionary in a global for subsequent retrieval in the next update that installs SM2." "Begin by toasting the old SMLoader if present in image. We also close any open loaders." (Smalltalk hasClassNamed: #SMLoader) ifTrue: [ (Smalltalk at: #SMLoader) allInstances do: [:e | e myDependents: #(). e delete]. SystemOrganization removeSystemCategory: 'SM-Loader']. "Remember installed packages if we have them and toast old SM." (Smalltalk hasClassNamed: #SMSqueakMap) ifTrue: [ Smalltalk at: #SM1InstalledPackagesDictionary put: (Smalltalk at: #SMSqueakMap) default installedPackagesDictionary. "This is a brutal hack, need to figure out why it is needed..." (Smalltalk at: #SMSqueakMap) setName: #SMSqueakMap. SystemOrganization removeSystemCategory: 'SM-domain'] 5709SM2Install-gk -- Göran Krampe -- 12 February 2004 This changeset installs SM2 into the image if it has not been installed already, by contacting the SM2 server at map1.squeakfoundation.org and asking it for the URLs to the packages to download and install. It first installs VersionNumber and MCInstaller if needed." | repDir server stream installedMC installedVN map | (Smalltalk hasClassNamed: #SMPackage) ifFalse: [ installedVN _ false. "Install first versions of VersionNumber for SM2 if it is missing, we can't rely on using old SM because it is now removed." (Smalltalk hasClassNamed: #VersionNumber) ifFalse: [ ChangeSorter newChangesFromStream: ((('http://map1.squeakfoundation.org/sm/package/35e02e3a-48e2-4843-94ee- e86d651aeafe/autoversion/1/downloadurl') asUrl retrieveContents content) asUrl retrieveContents content unzipped readStream) named: (ChangeSet uniqueNameLike: 'VersionNumber'). installedVN _ true]. "Load the new SM base, using available Monticello or MCInstaller, and installing MCInstaller otherwise" stream _ HTTPClient httpGet: ('http://map1.squeakfoundation.org/sm/package/c4c13ea3-e376-42c7-8d9e-d c23b09f9f29/autoversion/6/downloadurl' asUrl retrieveContents content). installedMC _ false. (Smalltalk hasClassNamed: #MCMczReader) ifTrue: [(MCMczReader versionFromStream: stream) load "Using Monticello"] ifFalse: [ "No Monticello installed, will use MCInstaller and install it first if needed" (Smalltalk hasClassNamed: #MczInstaller) ifFalse: [ ChangeSorter newChangesFromStream: ((('http://map1.squeakfoundation.org/sm/package/af9d090d-2896-4a4e-82d0- c61cf2fdf40e/autoversion/2/downloadurl') asUrl retrieveContents content) asUrl retrieveContents content unzipped readStream) named: (ChangeSet uniqueNameLike: 'MCInstaller'). installedMC _ true]. MczInstaller installStream: stream]. "This clears the default map, but keeps info on installed packages if there where any" (Smalltalk hasClassNamed: #SMPackage) ifTrue: [ map _ (Smalltalk at: #SMSqueakMap) default. map installedPackagesDictionary: (Smalltalk at: #SM1InstalledPackagesDictionary ifAbsent: [nil]). Smalltalk removeKey: #SM1InstalledPackagesDictionary ifAbsent: [nil]. "Remove old SM Package Loader, if present" map clearInstalledPackageWithId: '047a3b12-7e52-4c5d-be8b-d06635fc4f1c'. "Remove old SqueakMap Base, if present" map clearInstalledPackageWithId: 'fffa45d3-2459-4b7d-b594-9cfae17c864d'. "Remove old SqueakMap loadscript, if present" map clearInstalledPackageWithId: '4f0b9db6-8add-43aa-8d6b-53e6a0ea8442'. "Add that we now have versions of SqueakMap2 base, VersionNumber and MCInstaller installed" map noteInstalledPackageWithId: 'c4c13ea3-e376-42c7-8d9e-dc23b09f9f29' autoVersion: '6'. installedVN ifTrue: [map noteInstalledPackageWithId: '35e02e3a-48e2-4843-94ee-e86d651aeafe' autoVersion: '1']. installedMC ifTrue: [map noteInstalledPackageWithId: 'af9d090d-2896-4a4e-82d0-c61cf2fdf40e' autoVersion: '2']. "Make sure map is updated" map loadUpdates] ifFalse: [self error: 'Installation of "SqueakMap base" failed']. "Install new SMLoader using SM2 itself." map installPackageWithId: '941c0108-4039-4071-9863-a8d7d2b3d4a3' autoVersion: '2'] 5710snapshotSavingCleanups -- tim(a)sumeru.stanford.edu -- 15 February 2004 Some important cleanups in the saving of snapshots and initialising of FileDirectory. Makes it possible to save the snapshots in a different directory than the original snapshot file came from (ie you can start up /var/squeak/squeak.image and save ~joesixpack/joesworking.image). This version built for post-update 5707, compensating for the effects of recent KCP changes in the area of image saving and some changes in the 5707 updates" 5711AttrStreamFix-gk -- Göran Krampe -- 12 February 2004 After having looked at Boris Gaertner's fix for Scamper I agree with the override AttributeTextStream class>>new. No need to make any other changes to that class at this time. Just added a comment etc." 5712pathPartsFixes-tpr -- tim(a)sumeru.stanford.edu -- 9 February 2004 Some mild improvements to users of FileDirectory>pathParts. Revised per Frank Shearer's discovery of stupidity in respect to ServerDirectory>pathParts" 5713FileOutCategory-jf -- Julian Fitzell -- 7 February 2004 #fileOutCategory:asHtml: would try to create files without first validating the file names. This was a problem if you had a system category with a / in it for example. Fix this by adding a call to #asFileName on the file name string." 5714UTF8String-JO -- Jonas Öhrn -- 16 September 2002 Two small methods added to String for conversion between ISO-8859-1 and UTF-8. 5715CopyObjectMethod-tpr -- tim(a)sumeru.stanford.edu -- 14 February 2004 Revised Object>copyFrom: to use numbered optional prim 168 to copy objects" 5716GifReadWriterDecode-klc -- Ken Causey -- 26 January 2004 It seems that someone started some changes to handle comment blocks in GIF files but never finshed it. The code was completely unused and buggy (the block terminator was not being processed) and there was no way to tell what the goal of it was, so I removed it. This fixes a problem with reading GIF files with comments." 5717TextPropertiesMorph-fc -- Frank Caggiano -- 12 February 2004 Ned Konz added revert for extent as well. TextMorph 'text color...' interface 'cancel' button doesn't work Not just a problem with the cancel button. The TextMorph created inside the TextPropertiesMorph (activeTextMorph) and the TextMorph being edited (myTarget) were sharing the same attributes. Certain changes (color, text style) to the activeTextMorph were immeditely set in myTagret. This could be seen by forcing myTagret to redraw itself even while the textPropertiesMorph was still up (for example clicking the wrap button or covering and uncovering the target text morph). There still might be an issue with the wrap button. I'm not really sure how it is suppose to work." 5718MessageListDragFix-nk -- Ned Konz -- 15 February 2004 Fixes message dragging from MessageLists. 5719GIFReadingFixes-nk -- Ned Konz -- 15 February 2004 A few small fixes and enhancements for the GIF readers. * Add an 'offset' instVar to get rid of Undeclared refs. * Open either an ImageMorph or AnimatedImageMorph as appropriate from a FileList. * Allow opening GIFs from remote servers in the FileList. " 5720UrlTestFixes-gk -- Göran Krampe -- 12 February 2004 A bunch of extra tests for FileUrl given the FileUrlRewrite." 5721indentClipText-bf -- Bert Freudenberg -- 9 February 2004 When copying the text of an indented list via the red halo menu, the individual items where not indented, making the result illegible. Fixed." 5722FixObsoleteChronlgy-nk -- Ned Konz -- 15 February 2004 The Chronology changes weren't done correctly, so there are a number of bindings that refer to the old ZTime, etc. classes. This CS fixes those. | baddies | baddies _ ClassBuilder reallyObsoleteClasses asArray select: [ :ea | ea name first = $Z ]. baddies do: [ :bad | | newName | newName _ bad name allButFirst asSymbol. Smalltalk at: newName ifPresent: [ :newClass | ClassBuilder new recompile: false from: bad to: newClass mutate: false ] ]. Behavior flushObsoleteSubclasses. 5723ChangedWith-nk -- Ned Konz -- 17 February 2004 This adds #changed:with: and #update:with: to Object, as discussed on 2 and 3 December 2003. This is also VisualWorks compatible, for those packages that are ported from VW. 5724tests-reorganize-md -- Marcus Denker -- 16 February 2004 The unit-test from the Tests package have been added to the image. This changeset now renames all Tests-* categories to be *-Tests. Rational: When breaking stuff out (e.g. make a MC package Network-Url) then we sure want the tests be part of that, not continue to be part of the image. Some Tests* categories remain, e.g. Tests-KCP and Tests-Bugz. 5725timeSecondsPrintOn-brp -- Brent Pinkney -- 16 February 2004 Time printString should only show second precision." 5726SCN-compileInobtrsv-avi -- Avi Bryant -- 17 February 2004 Just a simple fix that adds an overlooked send to #uniqueInstance when referencing SystemChangeNotification in ClassDescription>>compileInobtrusively:classified:." 5727SmaCC-RuntimeSM-md -- Marcus Denker -- 16 February 2004 upate the SmaCC Runtime from SM" 5728fixPreferencesCat-md -- Marcus Denker -- 15 February 2004 Three methods in Preferences were categorized to be *Refactory. This changeset moves them to their correct catefory" 5729TestInteger-bg -- Boris Gaertner -- 2 February 2004 Peter Wiliam Lount found a problem when he constructed integers from four bytes and Bert Freudenberg published a fix. This test case checks the correct construction of value around SmallInteger maxVal." 5730DecorateDebuggerBtns-nk -- Ned Konz -- 18 February 2004 Makes the debugger's inheritance button colored like in the Browsers. 5731acceptCodeFix-bg -- Boris Gaertner -- 3 February 2004 KLC - I removed the commented out send in Browser>>selectOriginalCategoryForCurrentMethod otherwise identical. this change set fixes a bug that was reported on the list at Feb 03, 2004 by arj . A day later arj reported two additional ways to demonstrate undesired behavior of the brower. " 5732AccunyEdit-nk -- Ned Konz -- 19 February 2004 Edits the smallest two Accuny fonts to distinguish between the lower-case L and the number 1 glyphs. (commented out the demonstration display of the before & after glyphs. -dew) "((TextMorph new contentsAsIs: ((Text fromString: 'l1l1l1l1l1 Before') addAttribute: (TextFontReference toFont: (TextStyle named: 'Accuny') defaultFont); addAttribute: TextFontChange font1 from: 1 to: 2; addAttribute: TextFontChange font2 from: 3 to: 4; addAttribute: TextFontChange font3 from: 5 to: 6; addAttribute: TextFontChange font4 from: 7 to: 8; addAttribute: (TextFontChange fontNumber: 5) from: 9 to: 10; yourself)) imageForm magnifyBy: 4) asMorph position: ActiveHand position; openInWorld." 5733FileUrlRewrite-gk -- Göran Krampe -- 12 February 2004 Note: This changeset may have side effects, we simply need to test drive it (Filelist with FTP etc). A problem that cropped up with Scamper was a fix Markus made in update 5414. Markus made FileUrl>>pathForFile and #pathForDirectory to prepend the pathName delimiter if the FileUrl is absolute. The refinement is to avoid prepending if the file url has a path that starts with a DOS style drive letter, since such paths are absolute per definition. This triggered a BIG make-over of FileUrl, where it was made much more proper by using 'file://' as the prefix instead of 'file:'. According to the RFCs a file URL can not have relative file paths, the isAbsolute etc still works but the result sending toText (which I didn't bother to change even though it has a strange name) will be proper and does not reflect isAbsolute. I have taken care of adding 'smarts' on how FileUrl creates itself given a non valid String representation - it tries to be quite helpful. This is partly based on what Mozilla does in this regard. Now it is IMHO much improved, seems to work, has a bunch of new tests (posted separately), and Scamper works fine from filelist on DOS drives. 5734DebuggerHighlightFix-nk -- Ned Konz -- 20 February 2004 This fixes the misalignment between the displayed text in the Debugger and the highlighted PC range, when you have the browseWithPrettyPrint Preference enabled. 5735FixServerDirectories-nk -- Ned Konz -- 12 December 2003 Change Set 5605CleanupUpdateServers-dew did not clean up the ServerDirectory servers dictionary properly. This quickie will remove the bad servers. | badOnes | badOnes _ Dictionary new. ServerDirectory servers keysAndValuesDo: [ :k :v | (v isKindOf: ServerDirectory) ifFalse: [ badOnes at: k put: v ] ]. badOnes keysDo: [ :k | ServerDirectory servers removeKey: k ]. 5736nilTrueFalseLiteralArr-avi -- Avi Bryant -- 16 February 2004 Modifies Scanner>>scanLitVec to be ANSI-compliant and treat #(nil true false) as a collection of an UndefinedObject and two Booleans, rather than as three Symbols. Includes a test case (which also ensures that #(#nil #true #false) and #(#'nil' #'true' #'false') still contain symbols)." 5737hour12fix-avi -- Avi Bryant -- 21 February 2004 #hour12 in Time and DateAndTime was using a strange algorithm that gave incorrect results - for example, (Time fromString: '3:00 pm') hour12 gave 4 instead of 3. This fixes the problem. 5738SeparateCommentPane-nk -- Ned Konz -- 15 February 2004 This adds (for Morphic only, currently) a separate class comment pane instead of the combined class comment and class definition introduced in 5681browseWithComments-ls The 15 Feb. version avoids the search through the dependents, and avoids re-creating the comment pane morph. (merged Browser>>messageCategoryListIndex: with 5731acceptCodeFix-bg. -dew) 5739TextEmpahsis-fc -- Ned Konz -- 15 February 2004 Every time we change the Preferences class, we hold onto the old versions in individual preferences. This CS repairs existing Preferences to point to the current Preferences class. | obsoletePrefs | Smalltalk garbageCollect. obsoletePrefs _ Preference allSubInstances select: [ :ea | | ci | ci _ (ea instVarNamed: 'changeInformee'). ci isBehavior and: [ ci isObsolete ] ]. obsoletePrefs do: [ :pref | | ci | ci _ (pref instVarNamed: 'changeInformee'). pref instVarNamed: 'changeInformee' put: (Smalltalk at: ci name ifAbsent: []) ]. Behavior flushObsoleteSubclasses. 5740StartDragFix-nk -- Ned Konz -- 17 February 2004 Ghanges the start-drag logic to send the mouse-down event, instead of the event that is detected as the start of drag. This avoids the 10 pixel offset that can sometimes lead to incorrect behavior when dragging from PluggableListMorphs. 5741IntervalNewFromTests -- stephane ducasse -- 21 February 2004 another test for interval creation" 5742SyntaxMorphAddNoiseStr-fbs 5743MorphFileInFix-bf -- Bert Freudenberg -- 13 February 2004 Loading a morph from file opened the file in read/write mode, resulting in the inability to load a read-only morph file. Fixed." 5744FileContentsBrowser2-nk -- Ned Konz -- 18 February 2004 Changes isMetaClass to isMeta in PseudoClass and PseudoMetaclass to avoid problems with browsing code files that include classes that aren't already defined. 5745PosStream-upToEnd -- Boris Gaertner -- 19 February 2004 When reading past the end of a stream, our implementations of #next answer the value nil, which is certainly questionable. The method PositionableStream>>upToEnd checks for nil to detect the end of a stream which is certainly wrong. It has to use atEnd. " 5746FixObsoletePrefs-nk -- Ned Konz -- 15 February 2004 Every time we change the Preferences class, we hold onto the old versions in individual preferences. This CS repairs existing Preferences to point to the current Preferences class. | obsoletePrefs | Smalltalk garbageCollect. obsoletePrefs _ Preference allSubInstances select: [ :ea | | ci | ci _ (ea instVarNamed: 'changeInformee'). ci isBehavior and: [ ci isObsolete ] ]. obsoletePrefs do: [ :pref | | ci | ci _ (pref instVarNamed: 'changeInformee'). pref instVarNamed: 'changeInformee' put: (Smalltalk at: ci name ifAbsent: []) ]. Behavior flushObsoleteSubclasses. 5747SoundSystemCleanup-gk -- Adam Spitz, Göran Krampe -- 23 February 2004 This is the first changeset intended for the update stream that comes out of Adam's monster work Cleanup.sar. I am working through it alphabetically. This update is comprised of Adam's first two changesets: - AAAMultipleSoundSystems.3.cs - DSASoundCleanup.1.cs ...rewritten quite a lot: - Renamed SoundSystem to SoundService - Renamed BadSoundSystem to DummySoundSystem. - Renamed GoodSoundSystem to BaseSoundSystem. - Move DummySoundSystem and AbstractSoundSystem to System-Support - Kept references to the recently introduced Beeper as is, Beeper is now integrated on top of SoundService. - Tweaked Beeper with simplifications and more comments etc. - Moved general checks of Preferences soundsEnabled into Dummy/BaseSoundSystem, seems smart! - Added a playSoundNamedOrBeep: method - Added a base class AbstractSoundSystem - Added class comments for all new classes. - Threw in the DSASoundCleanup into this changeset since it also has to do with sound, preamble below. - Plus probably a bunch more little things. Lets get this one in because cleaning code rot isn't fun. :) /Göran Here follows original preambles: Preamble from AAAMultipleSoundSystems:: There are lots of places in the image that play sounds but shouldn't really depend on the sound system; if sound support isn't present, they should still be able to do their jobs (just a little more quietly :). This changeset adds a class called SoundSystem, so that we can say things like: SoundSystem default playSoundNamed: 'croak'. Preamble from AAAMultipleSoundSystems:: The DigitalSignatureAlgorithm class uses the sound system for some sort of randomness-generating thingy. This changeset is a half-assed attempt to remove this dependency, so that we can eventually create images without sound support. 5748TTCfontNameWPointSize-nk -- Ned Konz -- 12 February 2004 Fix for missing method. 5749setClassSelPackagePane-avi -- Avi Bryant -- 21 February 2004 Browser>>setClass:selector: now uses #selectCategoryForClass: so that it works properly from the PackagePaneBrowser as well. This fixes the bug with out of bounds errors when defining new classes in the PackagePaneBrowser that was caused by update 5487." 5750EventRecPlayback-los -- Lothar Schenk -- 26 February 2004 Image: 3.7a Clicking on a tape in a Filelist that was recorded with the EventRecorderMorph (e.g. EventRecorder.tape) and then choosing the 'open for playback' button results in a walkback: 'MessageNotUnderstood: EventRecorderMorph>>rewind'. The reason is this call in EventRecorderMorph class>>openTapeFromFile: openTapeFromFile: fullName (EventRecorderMorph new readTape: fullName) rewind openInWorld EventRecorderMorph has no 'rewind' method or anything resembling it. Leaving out the call to 'rewind' seems to work fine (also plausible, considering that the tape has just been read in from the file). The attached change set contains a corresponding patch. Regards, Lothar 5751CategoryDefaultName-dgd -- Diego Gomez Deck <DiegoGomezDeck(a)ConsultAr.com> -- 25 February 2004 better default for category name (the selector is used instead of #none). useful for etoys translation using Babel. 5752MPEGeToyVocab-dgd -- Diego Gomez Deck <DiegoGomezDeck(a)ConsultAr.com> -- 16 February 2004 Export the current funcionality in MPEGMorph to eToys environment. The motivation is to improve the 'Active Essay' features of squeak from eToys level. Now you can play, stop and rewind; read/modify the volume and the position; change the repeat option and ask is the move/sound is running. See an example of use at: http://swiki.agro.uba.ar/small_land/uploads/1/video_from_etoys.jpeg 5753SkipListClassComment-klc 5754BrowseComprsdCodeFiles-nk -- Ned Konz -- 17 February 2004 Adds the ability to open a FileContentsBrowser on the contents of a compressed stream, both from the FileList and from the BFAV. 5755beginsEndsWith-bp -- Bernhard Pieber -- 21 February 2004 Implements beginsWith: and endsWith: for all SequenceableCollections, not only for Strings." 5756testBeginsEndsWith-fbs 5757MorphicTextEditorRegistry -- Adam Spitz -- 11 February 2004 Creates a kind of AppRegistry called MorphicTextEditor. Installs PluggableTextMorph as the default one. (The SVI package, for example, could register itself as another kind of MorphicTextEditor. This would allow programs to be SVI-compliant simply by writing 'MorphicTextEditor default' instead of 'PluggableTextMorph'.)" 5758PNGReadWriterV7 -- Andreas Raab & Ned Konz -- 10 February 2004 Provide PNGReadWriter with the ability to write files. In addition, fix some issues with transparency in PNGs which was not handled correctly. v2: Vastly improved version. Includes a test suite which ensures that we can use PNG consistently to save and load Squeak forms. v3: 2/12/2004 - Fixed CRC storage - Added CRC validation for reading - Provide logging for test images - Attempt to fix byte-sex problem v4: Some fixes for the PNG reading and writing: * don't crash Squeak when performing tests (fixed in Zlib writer). * generate Adler32 checksum in compressed image chunks so other programs can read Squeak's PNG files. * verify Adler32 checksum in compressed image chunks to detect corruption. * modify test suite so that if you have the PngSuite.zip installed we also check for detection of corrupted files. v5: * Fixed big endian color problems (thanks to Avi) * Added more tests v6: * Fixed final swizzle problems for big endian machines v7 (NK): * Fixed gzip CRC generation. 5759KCP185SpaceUsedFix -- stephane ducasse -- 1 February 2004 clema spaceUsed implementation so that we can know it on Behavior instances too and avoid the hardcoded isKindO: Metaclass so that other extensions can use it too." 5760KCP190MoveUIInteraction -- stephane ducasse -- 1 February 2004 move the methods letUserReclassify: and categoryFromUserWithPrompt: aPrompt from ClassDescription to CodeHolder the only client of this method. This way we can have a better layered system...let us hope slowly. 5761KCP195MoveUpZapAll -- stephane ducasse -- 1 February 2004 Behavior is not an abstract class. So redefine and move zapAllMethods from ClassDescription to Behavior. (removed testZapMethods for now because there is no BehaviorTest class. -dew)" 5762chronology-tests1 -- Brent Pinkney and Tom Koenig -- 16 February 2004 These are the Chronology Tests, done by Brent Pinkney and Tom Koenig. The changeset adds some good utility methods to ClassTestCase for ensuring full test coverage. md: this is a fileout of Chronology-b-Tests-SqueakAndSmalltalk80Compatibility Chronology-h-Tests-NewGeneration Chronology-k-Tests-Cleanup Tom did more tests, these will follow in an additional changeset" 5763GenieRenameFor37-nk -- Ned Konz -- 20 February 2004 This CS adds all the genie- method categories to the Genie package so we can apply Monticello changes. This preamble does the basic category renaming needed. The rest of the CS changes a few other methods, primarily ones that shouldn't be in Genie. 20 Feb: removed inspector. | pi bad | SystemOrganization renameCategory: 'Tests-Genie-Engine' toBe: 'Genie-Tests'. pi _ PackageInfo named: 'Genie'. bad _ OrderedCollection new. (ClassDescription allSubInstances reject: [ :ea | ea isObsolete ]) do: [ :cls | | org | org _ cls organization. org categories do: [ :cat | (pi includesClass: cls) ifFalse: [ ( cat asLowercase beginsWith: 'genie') ifTrue: [ (org listAtCategoryNamed: cat) do: [ :sel | org classify: sel under: ('*', cat). ChangeSet current adoptSelector: sel forClass: cls. bad add: (cls -> sel) ]. org removeEmptyCategories ]] ifTrue: [ ( cat asLowercase beginsWith: '*genie-') ifTrue: [ | newName | newName _ cat copyFrom: 8 to: cat size. newName isEmpty ifTrue: [ newName _ cat copyFrom: 2 to: cat size ]. (org listAtCategoryNamed: cat) do: [ :sel | org classify: sel under: newName. ChangeSet current adoptSelector: sel forClass: cls. bad add: (cls -> sel) ]. org removeEmptyCategories ]. ]. ]]. "(bad explore) dependents first setLabel: 'moved methods'" 5764GenieRemoval-nk -- Ned Konz -- 14 February 2004 Removes Genie so that it can be loaded later as a package. Tested with 3.7a-5707. HandMorph allSubInstancesDo: [ :ea | ea disableGenie ]. Morph allSubInstancesDo: [ :ea | ea gestureDictionaryOrName: nil ]. FileList unregisterFileReader: CRRecognizer. <This automated message brought to you by the UpdateIncorporationTool.> <See the UpdateIncorporationTool package on SqueakMap for details.>

8 16

[FIX] AddAllFix-ac
by andrew＠beta4.com 07 Apr '04

07 Apr '04

from preamble: "Change Set: AddAllFix-ac Date: 6 August 2003 Author: Andrew Catton Fixed bug where: someCollection addAll: someCollection causes an infinite loop. - #addAll*: methods now use a copy of the parameter if it is the same as the receiver. - The old behavior was moved to #basicAddAll*: variants. - #basicAddAll: is used by SortedCollection>>copy to avoid an #addAll:->#copy->#addAll: infinite loop when adding a collection to itself. - Removed OrderedCollection>>addAll: as it didn't actually change the semantics inherited from Collection>>addAll: and would have required a further override on SortedCollection had it remained."!

4 5

[BUG][PackageInfo] PackageInfo>>extensionMethodsForClass: does not guard nil
by Chris Muller 07 Apr '04

07 Apr '04

Please disregard the attachment in my prior submission and use this one. It fixes the implementor rather than requiring all the senders to perform a nil guard. - Chris

2 1

[ANN] RMT (Remote Messaging Toolkit)
by Masashi Umezawa 06 Apr '04

06 Apr '04

Hi, I've put this on SM. http://map1.squeakfoundation.org/sm/package/94bba4b0-3b56-4af7-8a6e-5c4bb6a… --- RMT (Remote Messaging Toolkit) is a toolkit for developing network applications. - Support sync, async sends, RPC-style invocations. - Connection pooling. - Auto clients clean-up (by heartbeats). Toolkit itself can be used for doing simple RPC. You can subclass connector(client) and acceptor(server) for customized behavior. Prerequisites: KomServices --- Cheers, --- [:masashi | ^umezawa]

3 12

Report from a novice VM h4x0r.
by Alan Grimes 05 Apr '04

05 Apr '04

om, I've been obsessed with hacking the VM for the last 5 days or so, often staying up till 3AM watching builds compile.... I had hoped to be able to provide you with a snapshot of my progress but, unfortunately, I have hit some snags. I will describe these presently as I review the things I have learned from this excercise. My optimization plan was roughly as follows: 1. remove all unused or redundant variables and methods. 2. tune out all redundant computations. 3. use C library functions where possible (or make better use of C constructs). 4. Go through the compiler to cancel out any negative side effects from #1. 5. go through the code and re-insert variables only where necessary to reverse any regressions that couldn't be fixed by #4. I was doing okay, I made it all the way to steps 2-3 when I reached the point where my image was segfaulting or recursive erroring (with errionious instruction pointer values!) and there was no way to find what mistake I had made from the versions of the interpreter I had keept as backup for comparison. The 3.7 alpha image had too many problems to beginwith anyway. I started from scratch with a 3.6 image as a more stable base. I then inserted functions from my changes into the new image 1-by-1, comparing it to a "clean" interp.c (which I beleive to be identical to the one that I have used as my workhorse for several months now.) I went through several of these re-inclusions, running the SU test runner each time. I added a batch which caused errors, I thought "Horray! I found the bug". I was wrong. I backed out those changes and tried again, the error re-occoured. I backed out further changes from earlier versions and still got errors! Beginning to suspect foul play, I ran the test runner on the same image with my workhorse VM and was able to produce the same errors! -- I have come to the conclusion that SU test runner is very sloppy about its cleanup and exacerbates a problem deeper in the system, somewhere.... For this reason I am unable to continue untill there is some way to do an integrity check on any given image and a way to repair any corruption that may occour. =\ In the meantime, I have made a number of findings that may be of interest: 1. Many of the computations in the VM involve simple integer pointer arithmetic. I propose that instead of being addressed by only "char * memory", there be several pointers, char[] = int[] = &x which will simplify the VM code greatly... Furthermore, making all integer values word-aligned (are they already?) will tremendously improve memory access times on many architectures. 2. While the construct: | foo | foo _ self bar: baz foo = bat ifTrue: [Mars invade]. will help the inliner when bar is a non-trivial method significantly, it becomes counterproductive when bar happens to be "integerAt:" (which it is in several places) which is actually a compiler macro. 3. Many class, and a few instance variables are treated by the compiler as constants. Apparently this is a compiler hack which makes the C code more readable without affecting the binary. Unfortunately, this badly complicates the compiler class definition. Also, the C compiler does not follow the C idiom of making all constants ALLCAPS. My proposal is to edit the compiler so that it will detect degenerate methods (methods which only produce a constant value) and to treat these as constants. I beleive that this is an important change because it will allow class variables to be used by a future multithreaded interpreter which will probably require class variables such as "ActiveSmalltalkInterpreters". 4. A great many computations involve extracting bit-fields. In many cases these bit fields include important scalar information such as "sizeBits". A useful improvment would be to make these either bytes or short-ints and then let the C compiler figure out how best to extract them from the words... 5. I started experamenting with replacing long-coded block moves with calls to C's " memmove" and "memcpy" where appropriate. The motovation is that the 386 has an ungodly fast "rep movsb " compound instruction which does the operation at the speed of the FSB. It would be even better to use an inline assembly command for this but to remain portable I started trying to insert the C library calls mentioned. There are actually a number of DIFFERENT implementations of block memory copy in the VM source. One example in the Interpreter is called by a function which takes a word count then divides it by 4 to make a word coust for the method call, When it is processed by the method it is multiplied again to make the byte count used to determine the stop pointer for the For loop! 6. At many places in the VM, a whileTrue loop is used in place of a x to: y by: z do: [] loop. (it makes prettier C code.) 7. The compiler emits many unnecessary gotos. 8. The compiler will compile: foo _ foo + 1. as foo += 1, Which isn't bad code but on a non-optimizing C compiler this may turn into an add[immediate] opcode instead of the shorter inc x, opcode. 9. Both the optimizer and the interpreter have a number of implicit subclasses which are probably coded inline for performance or other reasons. These include many tables, headers, and other constructs which might be more maintainable as seperate classes. I would propose that a C calling convention be made so that message invocation such as: foo _ Bar new. foo baz. would compile as Bar_baz(); (when not inlined). 10. I came across an opcode in the interpreter whith a comment that stated it was supposed to be obsolete after 2.6!!! 11. It might be profitable to use parts of the internal smalltalk compiler in the cCode generator (if it doesn't already.) 12. One of my orrigional motives for this was the observation that certain bit-wise logic primitives on small ints will fail unnecessarily if they are stored as negative... A more general method of accessing any 32-bit int would seem to be appropriate. 13. The add opcode seems to attempt the add without changing the stack pointer and change it only after succeding. The logic operations (and some others) will change the stack pointer 3 times on success or failure! ( doing two pops followed by a push) 14. Many of the opcodes that access the stack without pushing will use the method "stackValue: 0" instead of the cleaner "stackTop". This probably won't affect the binary but it adds alot of constant arithmetic to the generated C code. It also indicates that the cCompiler relies on the native compiler too much to optimize out constant arithmetic...) 15. C99 adds a number of interesting features. Among which are improved libraries for dealing with integer types on a variety of platforms as well as new native boolean types. Squeak obviously uses sperate classes for true and false. It is not clear to me wheather there is any way to profit from this enhancement to the underlying language. C99 also adds complex math to the language. It would seem that adding primatives to take advantage of these enhancements would be of benefit. 16. C99 adds block contexts. The current cCompiler will take all inlined methods as well as loop variables and put them at the beginning of the generated function. Since any current C and even older C++ compilers will support for (int i; i < x ; i ++ ), it would be much better form to use this instead of heaping all variables at the beginning of the block. 17. The last major revision to the interpreter was 5 years ago. =\ om I am still fairly moronic with regards to matters of the interpreter but I hope that the people who have bothered to read these will find them to be of some use. Thanks for squeak!

9 23

Squeak listening to specific IP Addresses?
by Peter William Lount 03 Apr '04

03 Apr '04

Hi, Is it possible to configure Squeak to listen to specific IP Addresses and Ports, i.e. 1.2.3.4:8000, on a machine that supports multiple ip addresses? If so how? If not, how can we add this feature? A further example is running a Comanche squeak server on 1.2.3.4:80 and on 5.6.7.8:80. Two basic configurations interest me. 1. Restrict a squeak image to a specific ip address (or more than one specific address) from the set of addresses that are being homed on the server. 2. Enable squeak to listen to specific ipa:port pairs from the set of addresses that are being homed on the server. All the best, Peter William Lount, Smalltalk.org Senior Editor peter(a)smalltalk.org peter(a)activeinfo.ca

12 32

Re: How to bring money in the community
by Anthony Hannan 01 Apr '04

01 Apr '04

stephane ducasse <ducasse(a)iam.unibe.ch> wrote: > For my side I would pay for: > - a better compiler framework Do you like my closure compiler (I redesigned to whole compiler trying to make it more modular and extensible)?. If not, what don't you like about it? If you do, I'll be happy to take some money ;) Cheers, Anthony

2 1

How to bring money in the community
by stéphane ducasse 01 Apr '04

01 Apr '04

Hi all Often I start to code something and I bump into some uglyness (file systems) or some parts that are not finished (Projects) and I got slow down. Saying that part of X or Z is bad does not help, because you have the "participate" or do it yourself syndrome that I understand perfectly. Now I would be ready to ***PAY*** to have some better packages. I think that we should invent a market for better Squeak components. Even if the components get released in the image. For now we cannot pay even if we would like. So does anybody an idea how to proceed? For example, if one of you can evaluate the amount of effort to fix some broken/strange part and start with a price and participation people could then allocate money so that this part gets improved. For example: - we say the file system improvement should cost 500 $ - and that the parts are 25$ -> we need to find 20 people/companies/organization to pay or less. - once the people committed to pay then the provider do it. I would really like that we give a try because I'm ready to pay to get better software in Squeak. But right now we cannot even pay!!! Stef

12 21

← Newer
1
2
3
4
5
6
7
8
9
...
63
Older →

Jump to page: