Squeak-dev January 2011

squeak-dev@lists.squeakfoundation.org

94 participants
379 discussions

A Bootstrap Compiler
by Yoshiki Ohshima 17 Feb '11

17 Feb '11

Hi, We've been playing with John's MicroSqueak and it occured to me that having a bytecode compiler that is implemented outside of Squeak opens some possibilities, such as generate a growable image file from all text files, or make deep changes to the system without shooting yourself. I wrote a longer explanation so if you are interested, please go to: https://github.com/yoshikiohshima/SqueakBootstrapper and check it out. Thank you! -- Yoshiki

14 30

Alternative icons for Squeak menus
by Michael Davies 06 Feb '11

06 Feb '11

Hi All, I've put together an alternative icon set for Squeak menus, based on the Tango public domain icons from http://tango.freedesktop.org/Tango_Desktop_Project . They're a combination of untouched Tango icons, some tweaked ones, and some that I've created specifically for Squeak uses, so the polish may vary. They're in the inbox as Morphic-MAD.505.mcz - have a look and let me know if there's any interest in bringing them into the trunk. Cheers and Happy New Year, Michael

5 10

Error: Instances of StrikeFont are not indexable
by Bernhard Pieber 04 Feb '11

04 Feb '11

Dear Squeakers, I did the following: 1. I opened the latest trunk image Squeak4.2-10779-alpha.image with the Mac OS X VM Squeak 5.8b12. 2. Update Squeak. 3. Run all tests. After all the tests I get a debugger with "Error: Instances of StrikeFont are not indexable" --- The full stack --- StrikeFont(Object)>>error: StrikeFont(Object)>>errorNotIndexable StrikeFont(Object)>>size StrikeFont(Object)>>basicAt:put: StrikeFont(Object)>>instVarAt:put: StrikeFont(Object)>>readDataFrom:size: ReferenceStream(DataStream)>>readInstance ReferenceStream(DataStream)>>next ReferenceStream>>next ReferenceStream(DataStream)>>readArray ReferenceStream(DataStream)>>next ReferenceStream>>next [] in StrikeFontSet class>>installExternalFontOn:encoding:encodingName:textStyleName: [] in [] in MultiByteBinaryOrTextStream(PositionableStream)>>untilEndWithFork:displayingProgress: BlockClosure>>ensure: [] in MultiByteBinaryOrTextStream(PositionableStream)>>untilEndWithFork:displayingProgress: [] in BlockClosure>>newProcess Is there a way find out where this process is created? Tracing senders of StrikeFontSet class>>installExternalFontOn:encoding:encodingName:textStyleName: leads ultimately to no senders. In a Process Browser I do not see anything unusual. Cheers, Bernhard

8 20

final 4.2 image available
by Chris Muller 04 Feb '11

04 Feb '11

Squeak4.2-10950.zip is now available at http://ftp.squeak.org/4.2, which is intended to be the actual-released 4.2 image (unless we find some problem, which we won't). The trunk will be reopened shortly, thanks for your patience. - Chris

16 22

Accessing external resources in a cross-platform way
by Hernán Morales Durand 03 Feb '11

03 Feb '11

Let's say you want to resolve paths in a cross platform way Binaries @ 'MyExec' Libraries @ 'MyLib' SystemLibraries @ 'SystemLib' etc. then Binaries is In Windows: #( 'C:\Program Files\' 'C:\Archivos de Programa\' ...) " and all standard locations including translations " In UNIX: #('/usr/bin/' '/usr/local/bin' ....) " and all standard locations " etc Now you may say it is necessary to scan all directories to find MyExec, yes but you may then save that location myBinaryLocation := Binaries @ 'MyExec'. so you scan for it just once. What do you think? it is already implemented anywhere? If not what package do you think is more suitable or prepared for implementing it? Cheers, -- Hernán Morales Information Technology Manager, Institute of Veterinary Genetics. National Scientific and Technical Research Council (CONICET). La Plata (1900), Buenos Aires, Argentina. Telephone: +54 (0221) 421-1799. Internal: 422 Fax: 425-7980 or 421-1799.

4 5

Re: [squeak-dev] About Squeak 4.1 Language Support
by K. K. Subramaniam 02 Feb '11

02 Feb '11

On Friday 28 Jan 2011 10:14:43 am Eric wrote: > And it still doesn't work, does it mean i need to import some fonts for it > or other ways? Were you able to inspect the decoded character values? If the decoding is right then the issue is with fonts. See isFontAvailable and installFont methods in LanguageEnvironment to see how to load language-specific fonts on startup. GreekEnvironment and JapaneseEnvironment classes depend on such external font files. > Also another question is in Squeak4.1 what could i do to make sure the the > language environment I use is chinese or some others? What is your locale setting? Inspect: Locale current localeID SimplifiedChineseEnvironment>>supportedLanguages contains 'zh' > My understanding now in Squeak4.1 that the default language is UTF8 or > Unicode, where could i do the switch? Unfortunately not, because of the need for backward compability across multiple OSes and platforms. To select the right converter, use class side methods in SimplifiedChineseEnvironment. See JapaneseEnvironment for help. > I am just a beginner on Squeak Smalltalk, please correct me if I make any > mistakes on understanding Generally, requests to this mailing list receive brief responses. If you need more detail, you could try the beginners mailing list. Squeak needs people like to you to increase support for Chinese locales. Subbu

2 7

Perform and context frame size primitive tests
by Nicolas Cellier 01 Feb '11

01 Feb '11

Here is an interesting bug solved in Pharo affecting Squeak too with regular (non cog) vm http://code.google.com/p/pharo/issues/detail?id=3468 A patched CompiledMethod>>needsFrameSize: is in Pharo.

3 2

Uri vs. Url
by DeNigris Sean 01 Feb '11

01 Feb '11

It confuses me that the URI and Url class hierarchies are separate, given that a URL is a subset of URI (http://en.wikipedia.org/wiki/Uniform_Resource_Identifier#Relationship_to_UR…). Does it make sense to unify these? Thanks. Sean

2 1

Daily Commit Log
by commits＠source.squeak.org 01 Feb '11

01 Feb '11

Changes to Trunk (http://source.squeak.org/trunk.html) in the last 24 hours: http://lists.squeakfoundation.org/pipermail/packages/2011-January/004200.ht… Name: MultilingualTests-ul.13 Ancestors: MultilingualTests-ar.12 Added a test to demonstrate http://bugs.squeak.org/view.php?id=7603 . ============================================= http://lists.squeakfoundation.org/pipermail/packages/2011-January/004201.ht… Name: MultilingualTests-ul.14 Ancestors: MultilingualTests-ul.13 Added a widestring to MultiByteFileStreamTest >> #testNextPutAllStartingAt, otherwise it passes. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2011-January/004202.ht… Name: Collections-ul.422 Ancestors: Collections-ul.421 Introduced Stream >> #basicNext:putAll:startingAt: which will be used by MultiByteFileStream to fix http://bugs.squeak.org/view.php?id=7603 . ============================================= http://lists.squeakfoundation.org/pipermail/packages/2011-January/004203.ht… Name: Multilingual-ul.135 Ancestors: Multilingual-ul.134 Fixed an indexing mistake. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2011-January/004204.ht… Name: Multilingual-ul.134 Ancestors: Multilingual-ul.133 A fix for http://bugs.squeak.org/view.php?id=7603 . Requires Collections-ul.422. Added TextConverter >> #next:putAll:startingAt:toStream: which is used by MultiByteFileStream and MultiByteBinaryOrTextStream to properly implement #next:putAll:startingAt:. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2011-January/004205.ht… Name: Multilingual-bf.136 Ancestors: Multilingual-ul.135 - make MacUnicodeInputInterpreter use utf32 keycode unconditionally, since VMs older than 4.x are not supported anymore anyway. =============================================

1 0

FakeClassPool adopt: Pouah !
by Nicolas Cellier 31 Jan '11

31 Jan '11

FakeClassPool is a class copying temporarily classPool / sharedPool of a target class. The FakeClassPool is a way to pass states to the Compiler via a global variable (Pouah !). As such, it deserves a -1 raised to the power of an odd integer. It is supposed to let the Compiler access the class / shared variables of a class. OK, but the compiler already has access to the class/shared variables via the doitReceiver. If the doitReceiver is set correctly, then the FakeClassPool classPool/sharedPool are useless. Eventually, a FakeClassPool could be used to obfuscate a real receiver and replace it with a fake... But where is it of any use ? Can any grey beard remember ? Can a billion of young neurons extricate a valuable usage by comparing implementors of #selectedClass and of #doitReceiver ? Anyway, If I would want my doitReceiver to be a Fake, I would implement this feature in #doitReceiver. Wouldn't you? Any other usage would be intentionnally obscure. Indeed, it seems that a new FakeClassPool is used only when the selectedClass is nil... ... which means that the only case when the class and shared pool of the FakeClassPool are used is the case when they are nil! I would thus tend to nuke the #adopt: message. It would be a nice thing to kick out the whole class too. But measuring the consequences are out of my reach for the few minutes of Smalltalk left this evening. It seems that simple to remove this freak, there must be a trap somewhere... Any help appreciated. Nicolas

1 0

Jump to page: