January 2012 - Pharo-dev - lists.pharo.org

[Pharo-project] [update 1.4] #14296

by Marcus Denker

14296 ----- - actually add the code missed in the last update -- Marcus Denker -- http://marcusdenker.de

Jan. 24, 2012

Re: [Pharo-project] [Vm-dev] Split sound package -> Cog not loading in last 1.4

by Henrik Johansen

On Jan 24, 2012, at 8:12 14PM, Mariano Martinez Peck wrote: > Hi guys. Cog is not loading anymore in 1.4 because of some classes depending on Project from the 'Sound' package found in PharoSound repo in SS. At the beginning I thought to create a separate package "SoundVMSupport" and put there the needed classes, but I discover that there are a lot which are needed. And moreover, all of those needed classes are in the category Sound-Synthesis. The classes problematic (regarding Project and morphs) are in the other package Sound-Scores. I have just tried and split the Sound package in two: SoundScores and SoundSynthesis. Then I changed ConfigurationOfCog to load SoundSynthesis rather than Sound. That works and I can now build the VM without problems. > > I think that the split makes sense even without of the need of Cog. Do you agree? If true, I can commit everything. > One of the things I am not sure is how to make sure Sound package is not used anymore. All I can do is: > > a) Put a comment in squeaksource page saying that the packages are now SoundScores and SoundSynthesis. > b) Create a new baseline and version of ConfigurationOfPharoSound which reflecs this change > c) Commit a version of everything empty (no single class) to the package Sound. > > Which ones of these should I do ? My 2c: Just split of Sound-Scores into a separate package, leaving Sound containing Sound-Synthesis. Other than that, it sounds like a good plan. Why? Because it: 1. Makes it easy to still synch with the version in Squeak. 2. Avoids having to create a new baseline. 3. The extra work of either a), b), or c) is unnecessary. Cheers, Henry

Jan. 24, 2012

[Pharo-project] [update 1.4] #14295

by Marcus Denker

14295 ----- Issue 5215: FailingTest: Uncategorized Method http://code.google.com/p/pharo/issues/detail?id=5215 Issue 5216: Failing Test: Obsolete classes http://code.google.com/p/pharo/issues/detail?id=5216 Issue 5212: Fix CompiledMethod>>#browse http://code.google.com/p/pharo/issues/detail?id=5212 - do #cleanUpForRelease -- Marcus Denker -- http://marcusdenker.de

Jan. 24, 2012

Re: [Pharo-project] [Event related] Character vs Keys

by Johan Fabry

I am all for cleaning this up, but I'm sorry that I cannot be very constructive on this. I'd just like to confirm that keyboard handling with regard to special keypresses does not behave correctly. Try this in a workspace or transcript: hit ctrl-1 This prints 1. Same for 2 till 8. Guess what ctrl-9 does. No, it does not print 9. Try it ... In my opinion, ctrl-whatever should not put anything at all in the transcript window as it is not a normal keypress. Same holds for all ctrl, option, alt, command, et cetera modifier key presses, except for shift of course. But even so, the strange behavior of ctrl-9 does not really have a clear justification. On 24 Jan 2012, at 16:20, Guillermo Polito wrote: > Hi! > > While playing (again) to add Function keys support, reading documentation of how keyboard handling works on our vm and other platforms: > > .Net: http://msdn.microsoft.com/en-us/library/system.windows.forms.control.keydow… > Javascript: http://www.quirksmode.org/js/keys.html > > And a quick explanation of Igor of how our vm works: http://lists.squeakfoundation.org/pipermail/vm-dev/2011-November/009661.html > > > What puzzles me is some of the following: > > - Keypress event should only be raised to send keystrokes having a character representation. > - but some keystrokes with no character representation are taken as keypresses! (arrowup, arrowdown, esc...) (and so, used to handle events instead of doing it on keydown) > - the strokes above should be only raised in keydowns IMHO. > - more on, we have in Character class representations for those non characters :P which is funny because they are not characters. > > And everything works, and fits in a kind of strange way :). > > Now, rearranging this to follow those conventions means: > - reify Keys (vs Characters). So we can model modifier keys, function keys, arrow keys, independent from Characters. > - Keys (non-characters) should only be raised as single events on keydown. > - and the worse part is to fix all the hardcoded shortcuts messed everywhere in the image. > > Any thoughts on this? > > Thanks! > Guille ---> Save our in-boxes! http://emailcharter.org <--- Johan Fabry jfabry(a)dcc.uchile.cl - http://dcc.uchile.cl/~jfabry PLEIAD Lab - Computer Science Department (DCC) - University of Chile

Jan. 24, 2012

[Pharo-project] [Event related] Character vs Keys

by Guillermo Polito

Hi! While playing (again) to add Function keys support, reading documentation of how keyboard handling works on our vm and other platforms: .Net: http://msdn.microsoft.com/en-us/library/system.windows.forms.control.keydow… Javascript: http://www.quirksmode.org/js/keys.html And a quick explanation of Igor of how our vm works: http://lists.squeakfoundation.org/pipermail/vm-dev/2011-November/009661.html What puzzles me is some of the following: - Keypress event should only be raised to send keystrokes having a character representation. - but some keystrokes with no character representation are taken as keypresses! (arrowup, arrowdown, esc...) (and so, used to handle events instead of doing it on keydown) - the strokes above should be only raised in keydowns IMHO. - more on, we have in Character class representations for those *non characters* :P which is funny because they are not characters. And everything works, and fits in a kind of strange way :). Now, rearranging this to follow those conventions means: - reify Keys (vs Characters). So we can model modifier keys, function keys, arrow keys, independent from Characters. - Keys (non-characters) should only be raised as single events on keydown. - and the worse part is to fix all the hardcoded shortcuts messed everywhere in the image. Any thoughts on this? Thanks! Guille

Jan. 24, 2012

[Pharo-project] Split sound package -> Cog not loading in last 1.4

by Mariano Martinez Peck

Hi guys. Cog is not loading anymore in 1.4 because of some classes depending on Project from the 'Sound' package found in PharoSound repo in SS. At the beginning I thought to create a separate package "SoundVMSupport" and put there the needed classes, but I discover that there are a lot which are needed. And moreover, all of those needed classes are in the category Sound-Synthesis. The classes problematic (regarding Project and morphs) are in the other package Sound-Scores. I have just tried and split the Sound package in two: SoundScores and SoundSynthesis. Then I changed ConfigurationOfCog to load SoundSynthesis rather than Sound. That works and I can now build the VM without problems. I think that the split makes sense even without of the need of Cog. Do you agree? If true, I can commit everything. One of the things I am not sure is how to make sure Sound package is not used anymore. All I can do is: a) Put a comment in squeaksource page saying that the packages are now SoundScores and SoundSynthesis. b) Create a new baseline and version of ConfigurationOfPharoSound which reflecs this change c) Commit a version of everything empty (no single class) to the package Sound. Which ones of these should I do ? Thanks -- Mariano http://marianopeck.wordpress.com

Jan. 24, 2012

Re: [Pharo-project] Exception chapter PBE2

by StÃ©phane Ducasse

On Jan 23, 2012, at 10:54 PM, Norbert Hartl wrote: > I've read the exception chapter of the new PBE. And I really enjoyed reading it. I find it easy to understand and well-balanced between how-to-use and how-it-works. Great work! Thanks. Pharo by example volume two is slow to get out but it will contain part of my memory and what I would have loved to read to learn somethingâ¦ > > Norbert > > P.S.: I thought I've found a typo where a message sentTo: is used. But then I discovered there is indeed a message called sentTo: Unfortunately I didn't get the difference to sendTo: :) I forgot ;) this is why I wrote chapters.

Jan. 24, 2012

Re: [Pharo-project] [Seaside] FOSDEM2012 smalltalk devroom 5 feb, sprint 4 feb

by Tudor Girba

I wish I was there :) Have fun, Doru On Tue, Jan 24, 2012 at 10:02 AM, Stephan Eggermont <stephan(a)stack.nl> wrote: > Feb 4-5 2012 Brussels hosts FOSDEM2012, > the free and open software developers european meeting. > > Sunday there will be a smalltalk devroom (AW 1.126) with > a number of interesting presentations. > > 09:30 Norbert Hartl, Take a small REST. Simple approaches for REST in smalltalk > 10:00 Stephane Ducasse, Marcus Denker, The next steps for the Pharo Vision > 11:00 Laurent Laffont, John Thornton, Amber, the Smalltalk for web developers > 12:00 Nick Ager, An introduction to jQuery Mobile > 12:30 Stefan Marr, RoarVM, Sly > 13:00 David Chisnall, Compiling Smalltalk to fast native Code > 13:30 Craig Latta, Spoon > 14:00 Stephan Eggermont, Willem van den Ende, Diego Lont, Back to the future, (re)learn smalltalk > (till 16:30). > > http://fosdem.org/2012/schedule/track/smalltalk_devroom > > As there is no devroom on saturday, the VUB has kindly offered to organise > a sprint. There will be a room available at walking distance from FOSDEM. > > Stephan > > > _______________________________________________ > seaside mailing list > seaside(a)lists.squeakfoundation.org > http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside -- www.tudorgirba.com "Every thing has its own flow"

Jan. 24, 2012

[Pharo-project] ZipWriteStream error when closing it twice. Normal?

by Mariano Martinez Peck

Hi guys. I found that GZipWriteStream throws an error when closing it twice. Example: | aStream zipStream | aStream := ByteArray new writeStream binary. zipStream := GZipWriteStream on: aStream. zipStream close. zipStream close. All the streams I know do not throw an error in that case. And moreoever, it gets things complicated when using #on:do:. For example: FileStream forceNewFileNamed: testFileName do: [ :stream | GZipWriteStream on: stream binary do: [ :gzipStream | FLSerializer newDefault serialize: anObject on: gzipStream ] ]. In thise case, FLSerializer closes the stream at the end. No matter whether Fuel should do that or not, the thing is that in this I have the mentioned problem. So...my question is, is this a bug? if it is, do you have a better idea that changing #close to this: ZipWriteStream >> close literals isNil ifTrue: [ "#initialize sets this instVar to a specific value. #release puts it in nil, and #release is called from #flushBlock: which is called from #close:. Therefore, if it is nil, it means the stream has already been closed. If we try to close it twice, #flushBlock will throw an error because its instance variables are with nil" ^ self] self deflateBlock. self flushBlock: true. encoder close. -- Mariano http://marianopeck.wordpress.com

Jan. 24, 2012

[Pharo-project] FOSDEM2012 smalltalk devroom 5 feb, sprint 4 feb

by Stephan Eggermont

Feb 4-5 2012 Brussels hosts FOSDEM2012, the free and open software developers european meeting. Sunday there will be a smalltalk devroom (AW 1.126) with a number of interesting presentations. 09:30 Norbert Hartl, Take a small REST. Simple approaches for REST in smalltalk 10:00 Stephane Ducasse, Marcus Denker, The next steps for the Pharo Vision 11:00 Laurent Laffont, John Thornton, Amber, the Smalltalk for web developers 12:00 Nick Ager, An introduction to jQuery Mobile 12:30 Stefan Marr, RoarVM, Sly 13:00 David Chisnall, Compiling Smalltalk to fast native Code 13:30 Craig Latta, Spoon 14:00 Stephan Eggermont, Willem van den Ende, Diego Lont, Back to the future, (re)learn smalltalk (till 16:30). http://fosdem.org/2012/schedule/track/smalltalk_devroom As there is no devroom on saturday, the VUB has kindly offered to organise a sprint. There will be a room available at walking distance from FOSDEM. Stephan

Jan. 24, 2012