- Pharo-dev - lists.pharo.org

[Pharo-project] Alien FFI
by Torsten Bergmann Nov. 24, 2008

Nov. 24, 2008

There is now a squeaksource project for Alien - the new FFI. See http://www.squeaksource.com/Alien.html -- Sensationsangebot nur bis 30.11: GMX FreeDSL - Telefonanschluss + DSL fÃ¼r nur 16,37 Euro/mtl.!* http://dsl.gmx.de/?ac=OM.AD.PD003K11308T4569a

5 6

[Pharo-project] Host window events
by Igor Stasenko Nov. 24, 2008

Nov. 24, 2008

Hello list, i managed to build a new host windows plugin which currently including - window management - event handling - display/draw support The difference between old plugin, that for each host window an image always keeping own instance of HostWindow and provides an easy and fast way to translate a window hanlde (reported in events) to a corresponding HostWindow instance. A plugin, instead of operating with a raw OS window handles, keeping own list of open windows and lets language side operate with index in that list. Now, since all window handles is ordered and start counting from 0 , all we need is to keep a WeakArray and that array at: windowHandle can return us the corresponding HostWindow instance. This will allow to direct event handling to a proper HostWindow instance. That's where the fun begins. But! A host window is not a Form or DisplayScreen. It having a simple primitive which , like in old plugin, allows to blit a portion of form to a host window surface. Also, a plugin generates events to notify image about different things like window dimensions has changed or portion of screen need to be redrawn. What i would like to ask, is how to wire Form (or some DisplayScreen subclass) with HostWindow. Obviously, a HostWindow acts as backend , and don't needs to know anything about who wants to draw on it or listen for events. I need an advice , how to bind couple of events, which HostWindow should handle, and report them further to something which controls what to draw in it. Current 'interesting' number of events is quite small: - a window size has changed (and Form/DisplayScreen need to resync with new size, obviously) - a portion of window should be updated/redrawn (and Form/DisplayScreen needs to repaint the damaged region, obviously) there, of course will be more, but first things first :) What you think is the best way for reporting such events? -- Best regards, Igor Stasenko AKA sig.

4 13

[Pharo-project] [ANN] 10166
by StÃ©phane Ducasse Nov. 24, 2008

Nov. 24, 2008

Fixing author prompting in SUnit. So now you can run tests with the test runner and see the status of your tests in the browser. Thanks guys. Next I would love to see SUnit+ of Keith been integrated/used/getting feedback from the pharo community. Stef PS: Here we are under a snow storm - I was not aware that this could happen at lille :) Lot of snow falling and lot of wind.... but not enough snow for kids at the end

2 3

[Pharo-project] Fwd: Issue 319 in pharo: SparseLargeTable not sparse
by StÃ©phane Ducasse Nov. 23, 2008

Nov. 23, 2008

Begin forwarded message: > From: codesite-noreply(a)google.com > Date: November 23, 2008 9:54:45 PM CEST > To: stephane.ducasse(a)gmail.com > Subject: Issue 319 in pharo: SparseLargeTable not sparse > > > Comment #1 on issue 319 by ncell...@ifrance.com: SparseLargeTable > not sparse > http://code.google.com/p/pharo/issues/detail?id=319 > > I'd like to integrate these changes. > However, Eliot's interpretation of base is somehow uncompatible with > current > Multilingual usage: > > see senders of #base doing: > StrikeFont>>bonk:with: > StrikeFontSet>>bonk:with:at: > x base to: x size - 1 do: [:i | ...]. > > sent by #makeBoldGlyphs and #makeItalicGlyphs > > Other usage of SparseLargeTable in Pharo are in: > EFontBDFFontReaderForRanges via StrikeFont>> > - readEFontBDFFromFile:name:rangeFrom:to: > - readEFontBDFFromFile:name:ranges: > - readEFontBDFForJapaneseFromFile:name:overrideWith: > - readEFontBDFForKoreanFromFile:name:overrideWith: > Unicode class>>parseUnicodeDataFrom: > > Clearly Multilingual stuff from Yoshiki... > > -- > You received this message because you are listed in the owner > or CC fields of this issue, or because you starred this issue. > You may adjust your issue notification preferences at: > http://code.google.com/hosting/settings

1 0

[Pharo-project] New pharo-dev based on 10166
by Damien Cassou Nov. 23, 2008

Nov. 23, 2008

http://gforge.inria.fr/frs/download.php/11507/pharo0.1-10166dev08.11.6.zip http://code.google.com/p/pharo/wiki/Downloads?tm=2 -- Damien Cassou http://damiencassou.seasidehosting.st

1 0

[Pharo-project] Pharo crashes with 64 bit VM on Ubuntu
by paul provoost Nov. 23, 2008

Nov. 23, 2008

Hi, I compiled and installed the latest 64bit Squeak VM for Linux using the source tarball of Ians Piumarta's website. Compilation was succesfull, and Damien Cassous Squeak dev image runs fine. Pharo however crashes after the first try to claim more memory. The output on the commandline is: squeak: /test/squeak/Squeak-3.10-4/platforms/unix/vm/sqUnixMemory.c:172: uxGrowMemoryBy: Assertion `newDelta >= 0' failed. This problem can easily be reproduced by starting up the base image Pharo0.1Core-10164.image and opening the system browser, or doing something that requires extra memory like 10000 factorial in a workspace. Damiens Pharo dev image behaves exactly in the same way. Starting Pharo with command line argument -memory 4096m only delays the problem: after a few mouseclicks Pharo crashes. Any suggestions? Cheers, Paul Provoost The Netherlands _________________________________________________________________ Express yourself instantly with MSN Messenger! Download today it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

2 1

[Pharo-project] Rio File/Directory
by Keith Hodges Nov. 23, 2008

Nov. 23, 2008

Just to let you guys know, A while back I did the work of moving Rio over to use classes File and Directory, and String>>asFile and String>>AsDirectory. I am tempted to leave it that way for now (i dont have time to do otherwise for a bit) since I do want to propose it as a standard thing that could be in Squeak/Pharo/Whatever. Code size stats: self compiledSize: 'File-Kernel'. 6906 (for Kernel Images) self compiledSize: 'File-Base'. 8215 (includes some ftp remote directory support) self compiledSize: 'Files-Directories'. 11471 self compiledSize: 'FileMan-Core'. 7098 self compiledSize: 'Network-RemoteDirectory'. 21481 regards Keith

1 0

Nov. 22, 2008

Hi all I would like to stress two points: - I would like to see the code we are harvesting more tested. - but I would like that in alpha phase we continue to move on. So this is clearly difficult. In french we say that we cannot do an omelet without breaking eggs. Still I would like to stress two points: - I would like to see the code we are harvesting more tested. - but I would like that in not alpha phase we continue to move on. You know what I mean and we are trying to do our best. Now it means that people should also react fast to their submission feedback, which is the case since now. But we should continue but keeping in mind the two points above are important. In the beta phase we will not go at that pace and really stabilize the system. Then also we should favor improvement at the package level: for example the SUnit changes should in the future be really considered from the perspective of a external package improvement process. So a group of people can work on building a better SUnit (integrating/merging... keith improvement for example) and that we can build pharo-dev. Stef

1 0

Nov. 22, 2008

Hi I would like to have Nile in the image because I want to get rid of the old stream implementations slowly migrating the missing parts slowly to nile and I'm afraid that we lose the great work on damienC. I tried to harvest yesterday the item related to the back broken behavior in stream hierarchy but too much changes already happened. I could not find my way and lost time. I asked damien to help me but this is a sign. The plan could be to have Nile in Pharo-Dev for a couple of milestone and at a point switch (remove the old stream hierarchy and replace it by Nile). Also with DamienP we wrote more tests on collections and traits are really working great for that. Really! I imagine that we will have a really good test basis for the collections soon and that with Nile in the image I can improve the test coverage of Streams. I should say that I really fed up with some of the code aspects of Squeak (fileout of classes for example is a huge mess, file/directory....). Stef

1 0

Nov. 22, 2008

Hi guys I added two type entries Type-Defect = Report of a software defect Type-Enhancement = Request for enhancement Type-Task = Work item that doesn't change the code or docs Type-Patch = Source code patch for review Type-Mantis = Ticket on bugs.squeak.org to integrate Type-Squeak = Squeak version to integrate Type-Other = Some other kind of issue Type-ToDo = To list some todo for the coming versions Type-NextMilestone = To remember some actions for the next milestones I would like that we used them to sort well the items Stef

3 6