May 2011 - Pharo-dev - lists.pharo.org

Re: [Pharo-project] [Pharo-users] [COTDC] 68#2 MultiTextComposer

by StÃ©phane Ducasse

+1 On May 27, 2011, at 8:31 AM, Marcus Denker wrote: > > On May 26, 2011, at 10:34 PM, laurent laffont wrote: > >> Alain, you know this too :) ? >> >> Laurent >> >> On Wed, May 25, 2011 at 10:51 PM, laurent laffont <laurent.laffont(a)gmail.com> wrote: >> Today: MultiTextComposer > > One of the problems of the text rendering stuff is that the internationalization was just hacked > on top but never integrated. > > e.g. there is > > NewParagraph / MultiNewParagraph > TextComposer / MultiTextComposer > > and two CharacterScanner hierarchies (MultiCharacterScanner , CharacterScanner) > that seem to be copies yet randomly are used both... > > Just cleaning up would help a lot. And I would not be astonished if it would get even > faster just by cleaning up. (e.g. I removed an isKindOf: MultiNewParagraph from the main > text display routine yesterday...) > > Marcus > > -- > Marcus Denker -- http://www.marcusdenker.de > INRIA Lille -- Nord Europe. Team RMoD. > >

May 27, 2011

[Pharo-project] [update 1.3] #13221

by Marcus Denker

13221 ----- Issue 4293: ObjectScanner is unused http://code.google.com/p/pharo/issues/detail?id=4293 Issue 3167: Workspace loading from a file, file chooser ignores path http://code.google.com/p/pharo/issues/detail?id=3167 Issue 4298: Fix in CommentReference http://code.google.com/p/pharo/issues/detail?id=4298 Issue 4297: RecentSubmissions update http://code.google.com/p/pharo/issues/detail?id=4297 Issue 4299: problem with Dictionary and KeyNowFound http://code.google.com/p/pharo/issues/detail?id=4299 -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

May 27, 2011

Re: [Pharo-project] tab and OCompletion

by Marcus Denker

On May 27, 2011, at 8:05 PM, Alexandre Bergel wrote: > The problem is in: > > Pharo1.3a > Latest update: #13217 > Dev Can you add a bug report? > > Alexandre > > On 27 May 2011, at 07:57, StÃ©phane Ducasse wrote: > >> which version? >> >> Alex do you realize that we are working on 4 different systems? >> Pharo1.2 >> Pharo1.2 dev >> Pharo1.3 >> Pharo1.3 dev >> >> Stef >> >> On May 26, 2011, at 11:59 PM, Alexandre Bergel wrote: >> >>> Hi! >>> >>> When I press tab to complete a word, the cursor caret goes down of one line. >>> This almost makes it unusable. Any rational why it changed recently? >>> >>> Cheers, >>> Alexandre >>> -- >>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >>> Alexandre Bergel http://www.bergel.eu >>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >>> >>> >>> >>> >>> >>> >> >> > > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > > -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

May 27, 2011

Re: [Pharo-project] senders and implementors got distracted when arguments are symbol

by Alexandre Bergel

> on: #currentChangeSetIndex: send: #changed to: self. does not find anything while > on: 33 send: 44 to: self find implementors and senders. > Does anybody confirm? I confirm. Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

May 27, 2011

Re: [Pharo-project] tab and OCompletion

by Alexandre Bergel

The problem is in: Pharo1.3a Latest update: #13217 Dev Alexandre On 27 May 2011, at 07:57, StÃ©phane Ducasse wrote: > which version? > > Alex do you realize that we are working on 4 different systems? > Pharo1.2 > Pharo1.2 dev > Pharo1.3 > Pharo1.3 dev > > Stef > > On May 26, 2011, at 11:59 PM, Alexandre Bergel wrote: > >> Hi! >> >> When I press tab to complete a word, the cursor caret goes down of one line. >> This almost makes it unusable. Any rational why it changed recently? >> >> Cheers, >> Alexandre >> -- >> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >> Alexandre Bergel http://www.bergel.eu >> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >> >> >> >> >> >> > > -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

May 27, 2011

Re: [Pharo-project] Tiling windows

by GastÃ³n Dall' Oglio

version 11: Added screenshot of bounding of visible windows test comming soon, give some time :) 2011/5/27 GastÃ³n Dall' Oglio <gaston.dalloglio(a)gmail.com> > Ohhh look good and simple! I would suggest you use these renovated, but > it's too late, and also do not know if there may be some licensing issue: > http://www.famfamfam.com/lab/icons/silk/previews/index_abc.png > > A question, the correct way to load external icons in the image is used > ThemeIcons>>createIconMethodsFromDirectory: directory? that particular class > should be loaded? PharoUIThemeIcons or TWMBar, for example? in class o > instance side? > > 2011/5/27 laurent laffont <laurent.laffont(a)gmail.com> > >> Now with icons, and thanks Patrick for the icon generation idea :) >> >> Laurent. >> >> >> 2011/5/27 laurent laffont <laurent.laffont(a)gmail.com> >> >>> Now added top window placement Ã la OptimizeLayout. >>> >>> Need to find nice icons. Someone ? >>> >>> Laurent. >>> >>> >>> >>> 2011/5/26 GastÃ³n Dall' Oglio <gaston.dalloglio(a)gmail.com> >>> >>>> :) >>>> >>>> >>>> 2011/5/26 laurent laffont <laurent.laffont(a)gmail.com> >>>> >>>>> - Now loads in PharoCore >>>>> - more tests >>>>> - fill all space even if number of windows is not square. >>>>> >>>>> Laurent Laffont - @lolgzs <http://twitter.com/#!/lolgzs> >>>>> >>>>> Pharo Smalltalk Screencasts: http://www.pharocasts.com/ >>>>> Blog: http://magaloma.blogspot.com/ >>>>> Developer group: http://cara74.seasidehosting.st >>>>> >>>>> >>>>> >>>>> 2011/5/26 GastÃ³n Dall' Oglio <gaston.dalloglio(a)gmail.com> >>>>> >>>>>> >>>>>> 2011/5/26 laurent laffont <laurent.laffont(a)gmail.com> >>>>>> >>>>>>> 2011/5/26 GastÃ³n Dall' Oglio <gaston.dalloglio(a)gmail.com> >>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> 2011/5/26 laurent laffont <laurent.laffont(a)gmail.com> >>>>>>>> >>>>>>>>> GastÃ²n, >>>>>>>>> >>>>>>>>> I've downloaded the new version - thanks a lot for contributing, it >>>>>>>>> brings fun :) >>>>>>>>> >>>>>>>>> Some comment: >>>>>>>>> >>>>>>>>> - tiling browsers don't work any more with >>>>>>>>> OB, WMWithAnOBBrowserAndDockingBarsTest does not create an OBBrowser anymore >>>>>>>>> but a Browser >>>>>>>>> >>>>>>>> >>>>>>>> In 1.3 core there's not OBBrowser, then, when trying to modify a >>>>>>>> method with reference to OBBrowser obtained an error. So I remove their >>>>>>>> references. >>>>>>>> >>>>>>> >>>>>>> >>>>>>> I see. I will refactor this so it will be compatible with Core and >>>>>>> still work with OB (quite easy). >>>>>>> >>>>>> >>>>>> >>>>>> This would be great, because now I think, in a core are never the most >>>>>> useful tools (Seaside control panel, memory monitor), and TWM should know >>>>>> about them in some way (you apply a pattern to solve this problem?). >>>>>> >>>>>> >>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>>> - you've changed TilingWM>>visibleWindows to TilingWM>>windows. The >>>>>>>>> "visible" was not innocent, because tileVisibleWindows cares about windows >>>>>>>>> that are not minimized. On the other side, tileWorkspaces tile all >>>>>>>>> workspaces, minimized or not. So I think in final we may have: >>>>>>>>> >>>>>>>>> #tileVisibleWindows >>>>>>>>> #tileWindows (maybe #tileAllWindows is a better name) >>>>>>>>> #tileVisibleWorkspaces >>>>>>>>> #tileWorkspaces / #tileAllWorkspaces >>>>>>>>> ...... >>>>>>>>> >>>>>>>>> >>>>>>>> Please, see you the comment in de versiÃ³n 5 of de package, "Now all >>>>>>>> botton tile visibles windows.". I think it's better to just get organized >>>>>>>> visible windows, even when done on a particular kind (such as workspaces). >>>>>>>> For this is that the word was redundant now visible. >>>>>>>> >>>>>>> >>>>>>> >>>>>>> I will think about this. >>>>>>> >>>>>> >>>>>> When you have many windows, minimizing them is the way to let them >>>>>> "suspended", therefore I think it should be organized only visibles. >>>>>> >>>>>> >>>>>> >>>>>>> >>>>>>> >>>>>>> - I will add tests to TWMBar finally as it seems to grow >>>>>>>>> >>>>>>>>> - I want TilingWM to be 100% test coverage >>>>>>>>> >>>>>>>> >>>>>>>> Sorry, give me some time for learn about this... >>>>>>>> >>>>>>> >>>>>>> >>>>>>> Don't hesitate to ask. >>>>>>> >>>>>>> >>>>>> ok. >>>>>> >>>>>> >>>>>>> >>>>>>> - Balloon and spaces for TWMBar is a good idea - we need to make this >>>>>>>>> work effectively :) >>>>>>>>> >>>>>>>>> >>>>>>>> yes >>>>>>>> >>>>>>>> >>>>>>>>> - Great to have all the buttons per type of applications ! >>>>>>>>> >>>>>>>> >>>>>>>> it is still very rudimentary way to filter the windows, but it was >>>>>>>> helpful to learn about the hierarchy of classes 1.3, in regard to the Models >>>>>>>> for example. >>>>>>>> >>>>>>> >>>>>>> >>>>>>> Indeed I've disable TaskBarMorph in my image as it's useless now :) >>>>>>> >>>>>> >>>>>> >>>>>> Great!!! >>>>>> >>>>>> I'll be honest, this is my first serious development in Squeak / >>>>>> Pharo, so it gives me much encouragement your comment, there is still much >>>>>> to learn and improve, but it is very encouraging to devote time to something >>>>>> that might be useful to others. >>>>>> >>>>>> >>>>>> >>>>>> >>>>>>> >>>>>>> >>>>>>> Laurent. >>>>>>> >>>>>>> >>>>>>>> >>>>>>>> >>>>>>>>> >>>>>>>>> I've merge some of your changes in TWM-LaurentLaffont.5 >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> :) >>>>>>>> >>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> Laurent Laffont - @lolgzs <http://twitter.com/#!/lolgzs> >>>>>>>>> >>>>>>>>> Pharo Smalltalk Screencasts: http://www.pharocasts.com/ >>>>>>>>> Blog: http://magaloma.blogspot.com/ >>>>>>>>> Developer group: http://cara74.seasidehosting.st >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> 2011/5/26 GastÃ³n Dall' Oglio <gaston.dalloglio(a)gmail.com> >>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> 2011/5/25 laurent laffont <laurent.laffont(a)gmail.com> >>>>>>>>>> >>>>>>>>>>> 2011/5/25 GastÃ³n Dall' Oglio <gaston.dalloglio(a)gmail.com> >>>>>>>>>>> >>>>>>>>>>>> Excelent! >>>>>>>>>>>> >>>>>>>>>>>> IMHO, it's better that's not integrated to TaskBarMorph, lets it >>>>>>>>>>>> free :) >>>>>>>>>>>> For example, click in unused space of the TWMBar and you can >>>>>>>>>>>> move it around you wish in the screen, similar to use >>>>>>>>>>>> DockingBarMorph>>beFloating. I attach screenshot. >>>>>>>>>>>> >>>>>>>>>>>> Laurent, I improved (or not? :)) your work, if you think that is >>>>>>>>>>>> bad for TWM simply undo it ;) >>>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> Feel free to play. Just try to write tests for TilingWM please :) >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>> ok. I load a new version to ss. Now teh button tile visible >>>>>>>>>> windows only. I fixed the test... >>>>>>>>>> >>>>>>>>>> >>>>>>>>>>> For TWMBar I'm not sure this will be a final GUI. Ideally all >>>>>>>>>>> will be keyboard driven, I have to play with KeyMapping. >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>> yes, sure. >>>>>>>>>> >>>>>>>>>> >>>>>>>>>>> I also think the UI theme Patrick has cleaned fits better with >>>>>>>>>>> TilingWM: >>>>>>>>>>> >>>>>>>>>>> To try it: >>>>>>>>>>> >>>>>>>>>>> Gofer it >>>>>>>>>>> squeaksource: 'PBASandbox'; >>>>>>>>>>> package: 'PBSandbox'; >>>>>>>>>>> load. >>>>>>>>>>> StandardUITheme beCurrent. >>>>>>>>>>> >>>>>>>>>>> See screenshot. >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>> I try, yes. >>>>>>>>>> >>>>>>>>>> >>>>>>>>>>> Patrick, could you push this with a nice package name ? >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> Laurent. >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>>> 2011/5/25 laurent laffont <laurent.laffont(a)gmail.com> >>>>>>>>>>>> >>>>>>>>>>>>> Now there's a button to open the world menu. >>>>>>>>>>>>> >>>>>>>>>>>>> Laurent Laffont - @lolgzs <http://twitter.com/#!/lolgzs> >>>>>>>>>>>>> >>>>>>>>>>>>> Pharo Smalltalk Screencasts: http://www.pharocasts.com/ >>>>>>>>>>>>> Blog: http://magaloma.blogspot.com/ >>>>>>>>>>>>> Developer group: http://cara74.seasidehosting.st >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> On Wed, May 25, 2011 at 11:20 AM, laurent laffont < >>>>>>>>>>>>> laurent.laffont(a)gmail.com> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>>> On Wed, May 25, 2011 at 11:06 AM, StÃ©phane Ducasse < >>>>>>>>>>>>>> stephane.ducasse(a)inria.fr> wrote: >>>>>>>>>>>>>> >>>>>>>>>>>>>>> I do not understand. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> how can I get a browser? >>>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> Hey, this is just a start :) I will add a button to open the >>>>>>>>>>>>>> world menu. >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> once this is one is created I get the browser full screen >>>>>>>>>>>>>>> but then how do I get a chance to create a new one if I want. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> dividing the browsers based on the screen size results in >>>>>>>>>>>>>>> rather small browsers :) >>>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> yes, but because browsers waste space IMHO. Maybe we could get >>>>>>>>>>>>>> rid of windows borders when tiled. >>>>>>>>>>>>>> >>>>>>>>>>>>>> Laurent. >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Stef >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> > TWMBar new open. >>>>>>>>>>>>>>> > >>>>>>>>>>>>>>> > The real manager is TilingWM. Try: >>>>>>>>>>>>>>> > >>>>>>>>>>>>>>> > (TilingWM forWorld: World) tileVisibleWindows. >>>>>>>>>>>>>>> > (TilingWM forWorld: World) tileBrowsers. >>>>>>>>>>>>>>> > (TilingWM forWorld: World) tileWorkspaces. >>>>>>>>>>>>>>> > >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>> >>>>>> >>>>> >>>> >>> >> >

May 27, 2011

[Pharo-project] senders and implementors got distracted when arguments are symbol

by stephane ducasse

Hi guys on: #currentChangeSetIndex: send: #changed to: self. does not find anything while on: 33 send: 44 to: self find implementors and senders. Does anybody confirm? Send fix? Stef

May 27, 2011

Re: [Pharo-project] tab and OCompletion

by Mariano Martinez Peck

I saw the same problem :( if someone can fix it please... On Thu, May 26, 2011 at 11:59 PM, Alexandre Bergel <alexandre.bergel(a)me.com>wrote: > Hi! > > When I press tab to complete a word, the cursor caret goes down of one > line. > This almost makes it unusable. Any rational why it changed recently? > > Cheers, > Alexandre > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > > > -- Mariano http://marianopeck.wordpress.com

May 27, 2011

Re: [Pharo-project] 1.3 vs. 1.4

by Alexandre Bergel

Thanks Marcus. Indeed, this is important Alexandre On 27 May 2011, at 09:25, Marcus Denker wrote: > > On May 27, 2011, at 2:34 PM, Alexandre Bergel wrote: > >>> 3341 use #= for integer comparison instead of #== >> >> >> Hi! >> >> I checked the issue http://code.google.com/p/pharo/issues/detail?id=3341 >> but there is no explanation about the rational. Is there a technical reason to compare integer with = is better than with == ? Just wondering > > > This explains it more: > > http://bugs.squeak.org/view.php?id=2788 > > 002788: Bug: Use of == for arithmetic equality > Description From: Dan.Ingalls(a)Post.Harvard.edu > Subject: Bug: Use of == for arithmetic equality > Date: 13. Februar 2006 18:37:42 GMT+01:00 > To: squeak-dev(a)lists.squeakfoundation.org > > I've been playing around with a new VM (heh, heh) which, for a while, happened not to intern (ie force unique instances of) SmallIntegers. In this case the use of == to mean arithmetic equality will not work properly. In my opinion, all such occurrences in the system should be eliminated ASAP; == is not an arithmetic compare in any Smalltalk I know of. While it may work with small constants, it is simply wrong, and an especially bad example for newbies to see. Besides failing in certain interpreters, it will fail in Squeak itself if the integers are not small. > > I regret that I don't have time to fix these right now. However, if there is a well-intentioned soul out there, he or she will perhaps find the method below to be quite useful. It found 165 methods in my system with this pattern. > > > -- > Marcus Denker -- http://www.marcusdenker.de > INRIA Lille -- Nord Europe. Team RMoD. > -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

May 27, 2011

Re: [Pharo-project] 1.3 vs. 1.4

by Marcus Denker

On May 27, 2011, at 2:34 PM, Alexandre Bergel wrote: >> 3341 use #= for integer comparison instead of #== > > > Hi! > > I checked the issue http://code.google.com/p/pharo/issues/detail?id=3341 > but there is no explanation about the rational. Is there a technical reason to compare integer with = is better than with == ? Just wondering This explains it more: http://bugs.squeak.org/view.php?id=2788 002788: Bug: Use of == for arithmetic equality Description From: Dan.Ingalls(a)Post.Harvard.edu Subject: Bug: Use of == for arithmetic equality Date: 13. Februar 2006 18:37:42 GMT+01:00 To: squeak-dev(a)lists.squeakfoundation.org I've been playing around with a new VM (heh, heh) which, for a while, happened not to intern (ie force unique instances of) SmallIntegers. In this case the use of == to mean arithmetic equality will not work properly. In my opinion, all such occurrences in the system should be eliminated ASAP; == is not an arithmetic compare in any Smalltalk I know of. While it may work with small constants, it is simply wrong, and an especially bad example for newbies to see. Besides failing in certain interpreters, it will fail in Squeak itself if the integers are not small. I regret that I don't have time to fix these right now. However, if there is a well-intentioned soul out there, he or she will perhaps find the method below to be quite useful. It found 165 methods in my system with this pattern. -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

May 27, 2011