Squeak-dev April 2010

squeak-dev@lists.squeakfoundation.org

122 participants
595 discussions

Re: Cobalt VR Meetup follow-up
by Chris Cunnington 02 Apr '10

02 Apr '10

That would be great. I'll be posting the announcement about it 6 May. If the day comes and you have any questions about how to proceed, please feel free to email me directly. Cool Cobalt World? I don't have one. I have a default space that looks like the holodeck at night on the Enterprise. Cobalt also comes with a very beautiful world. I got ribbed once for not interior decorating. But then I go and put a floating triangle in my holodeck, and people can't join my space. Seems you need to use 3D Studio Max or Sketchup or something to add stuff to the world. You don't need a cool world. Windows 7 or Ubuntu? Which is more stable? That's a great question and one I'm not best suited to ask. I'm on a Mac. But really, that's what the Cobalt VR Meetup is all about. We use Skype IRC and that's solid as a rock. Then we can try stuff in Cobalt to see what's working and what isn't. Get Skype and OpenCobalt and join the dialog. It's pretty cool stuff. Your students will love it. It's Primo, as in the beer. Chris

1 0

OpenGL into morphic ala lumiere?
by Lawson English 02 Apr '10

02 Apr '10

More than likely, this is way too advanced for moi, but I spent about an hour rummaging through the Lumiere code for Pharo and just couldn't grok how the OpenGL rendering is directed to a morph instead of to raw coordinates on the main window. Is there some special property of Pharo that allows this kind of thing, or is there some relatively simple way in Squeak to direct OGL drawing to a morphic surface/canvas/thingie so it renders inside the boundaries of the morph when moved? Or does my question even make sense? Thanks. Lawson

2 3

Daily Commit Log
by commits＠source.squeak.org 02 Apr '10

02 Apr '10

Changes to Trunk (http://source.squeak.org/trunk.html) in the last 24 hours: http://lists.squeakfoundation.org/pipermail/packages/2010-April/003185.html Name: Collections-ar.353 Ancestors: Collections-nice.352 Fix do:displayingProgress: to suppress changing the label to nil if the display block returns nil (accidentally or deliberately). ============================================= http://lists.squeakfoundation.org/pipermail/packages/2010-April/003186.html Name: Kernel-ar.437 Ancestors: Kernel-ar.436 Fixes needed for new condensing method: * Allow class comments to be moved to the sources file * Make CompiledMethod>>setSourcePointer: less expensive by avoiding #become: if at all possible ============================================= http://lists.squeakfoundation.org/pipermail/packages/2010-April/003187.html Name: System-ar.307 Ancestors: System-ul.306 Adds a new condensing method. Smalltalk appendChangesTo: 'myfile.sources' will copy the original sources file and append a condensed version of the changes. These later versions are linked to the earlier ones so that the 'versions' of some method show its ancestry in the sources file. This gives us the ability to preserve the 'shipped' ancestry of methods while avoiding all the intermediate versions during development. Also bullet-proofs ChangeSet>>scanVersionsOf: in the case of broken preambles. =============================================

1 0

Why do the buttons in the code browser change size along with the rest of the panes?
by Ken Causey 01 Apr '10

01 Apr '10

I've been ignoring this issue for a while now thinking someone else would get around to fixing it, and frankly it has been a while since I've used an updated trunk image. I finally updated today and it's still there. Am I the only one that resizes the code browsers? Who likes the fact the the button panes resize along with the other panes? Why would that pane be defined to be a percentage of the parent pane and not based on the size of the tallest button? I'm not sure where to look to fix this, pointers would be appreciated. Ken

3 2

Caching MultiByteFileStream bug?
by Andreas Raab 01 Apr '10

01 Apr '10

Hi Levente - Could you check the failure of FileStreamTest>>testNextChunkOutOfBounds? This looks like it might be a problem with caching file stream behavior. If it's not (i.e., the infinite recursion is the 'expected' behavior) can we add something that addresses the issue? Thanks, - Andreas

2 6

Copying the sources file
by Andreas Raab 01 Apr '10

01 Apr '10

Hi - I was just trying to implement the code for condensing the changes to the end of the sources file and ran into some 'interesting' issues. I would expect this to work: fullName := 'test1.sources'. "Copy sources file; change file name accordingly" FileStream forceNewFileNamed: fullName do:[:newFile| sourcesFile := SourceFiles at: 1. sourcesFile position: 0. FileDirectory default copyFile: sourcesFile toFile: newFile. newFile position = sourcesFile size ifFalse:[self error: 'File copy failed']. ]. But it fails. Then I tried something else: fullName := 'test2.sources'. "Copy sources file; change file name accordingly" FileStream forceNewFileNamed: fullName do:[:newFile| sourcesFile := SourceFiles at: 1. sourcesFile position: 0. [sourcesFile atEnd] whileFalse:[newFile nextChunkPut: sourcesFile nextChunk]. newFile position = sourcesFile size ifFalse:[self error: 'File copy failed']. ]. That fails to. Then I tried: fullName := 'test3.sources'. "Copy sources file; change file name accordingly" FileStream forceNewFileNamed: fullName do:[:newFile| sourcesFile := SourceFiles at: 1. sourcesFile position: 0. [sourcesFile atEnd] whileFalse:[newFile nextChunkPutWithStyle: sourcesFile nextChunkText]. newFile position = sourcesFile size ifFalse:[self error: 'File copy failed']. ]. That fails as well. Help! What am I doing wrong? How does one copy the sources file properly and why do the above methods not work? Cheers, - Andreas

4 8

Re: dlopen() crashing
by Schwab,Wilhelm K 01 Apr '10

01 Apr '10

Bert, What do you mean by "declare the mode parameter" - did I not do that via the argument named int? Bill

1 0

dlopen() crashing
by Schwab,Wilhelm K 01 Apr '10

01 Apr '10

Hello all, I am trying to use dlsym() which means I have to call dlopen(), and it is crashing with messages such as Inconsistency detected by ld.so: dl-open.c: 643: _dl_open: Assertion `_dl_debug_initialize (0, args.nsid)->r_state == RT_CONSISTENT' failed! I am calling it as follows: dlOpen:aString mode:int < cdecl: void* 'dlopen' ( char* long ) > ^ self externalCallFailed. With zero for the mode parameter, it runs but returns an null handle; with 2 (now) or 10 (now|global), it crashes. Any ideas or working examples? Bill

2 1

Cobalt VR Meetup tonight
by Chris Cunnington 01 Apr '10

01 Apr '10

I'm starting a Cobalt VR Meetup on the first Thursday of each month at 8pm Toronto time ( - 4 GMT). All this really means is that I'll be on the Cobalt Skype IRC at that time to wave a flag and connect with whomever wants to show up. Then we can link our Cobalt worlds. http://www.opencobalt.org/community/connect Get Skype, go to the link above, and find "Click here to access the live chat." To see how to connect your downloaded Cobalt app watch this: http://www.youtube.com/watch?v=EqE0WnZCn5M Chris

1 0

Re: immutibility
by Chris Cunnington 01 Apr '10

01 Apr '10

Hi Randal, I'm going to disagree with you pretty hard about this. I think what you're proposing sounds tedious beyond words. And I don't think it's what Squeak's about. Databases are tedious. Operating systems are tedious. Cincom and Gemstone obsess about databases and tedious minuta, because that's their job. I spent several days reading CentOS archives and I thought I would pass out, it was so dull. Squeak is not dull. It's about multimedia, 3D worlds, web development, desktop applications, etc. Squeak should always be first in space doing wild things. Whatever VPRI has up its sleeve we should look at and play with. Squeak is about programming for its own sake not somebody else's. One of the things that Pharo will have to contend with is sounding reasonable to other people. If they want to be appealing to business, they will have to listen to lots of stupidity from people who have databases and operating systems for brains. I think that's going to be a pretty unpleasant part of their adventure. That's not what Squeak's about. Jesse has some crazy idea about object memory? I say great. Fulmer wants to bring Cobalt to the Trunk? I say that's what we're all about. VPRI has some craziness called Ometa? Bring it on. Squeak is not now, nor should it ever be, about boring, reasonable plans. This is where the cool stuff happens. Dynabooks to the third world? Right on. If it weren't for the oddness of continuations, Seaside would be dull as dishwater. Andreas wants to throw generators into the Trunk? Are they nifty? Yes they are; so let's do it. Databases are for peasants. We are unique and occupy a unique position. Let's embrace that. Stop trying to turn us into shadows of Gemstone, Cimcom, or other merchants of tedium. If Kent Beck isn't happy with Squeak, I don't see it as our job to cater. There is nothing else on earth like us and what we do. Squeak is a never ending game of capture the flag. Chris

16 47

Jump to page: