May 2015 - Pharo-dev - lists.pharo.org

[pharo-project/pharo-core] 947163: 50060

by GitHub

Branch: refs/heads/5.0 Home: https://github.com/pharo-project/pharo-core Commit: 947163790b1294d8502473a0c56b1a0bddbdac8e https://github.com/pharo-project/pharo-core/commit/947163790b1294d8502473a0… Author: Jenkins Build Server <board(a)pharo-project.org> Date: 2015-05-22 (Fri, 22 May 2015) Changed paths: R Graphics-Fonts.package/FontSet.class/class/as yet unclassified/convertFontsNamed_.st R Graphics-Fonts.package/FontSet.class/class/compiling/compileFont_.st R Graphics-Fonts.package/FontSet.class/class/converting/convertFontsNamed_inDirectoryNamed_.st R Graphics-Fonts.package/FontSet.class/class/converting/convertTextStyleNamed_.st R Graphics-Fonts.package/StrikeFont.class/class/examples/convertFontsNamed_.st R Graphics-Fonts.package/StrikeFont.class/instance/file in%2Fout/writeAsStrike2On_.st R Graphics-Fonts.package/StrikeFont.class/instance/file in%2Fout/writeAsStrike2named_.st A Reflectivity-Tests.package/HookGeneratorTest.class/instance/tests - hooks/testConditionWithArguments.st M Reflectivity.package/HookGenerator.class/instance/ast manipulation/wrapCondition_.st A Reflectivity.package/ManifestReflectivity.class/README.md A Reflectivity.package/ManifestReflectivity.class/definition.st A Reflectivity.package/ManifestReflectivity.class/instance/code-critics/ruleRBBadMessageRuleV1FalsePositive.st A Reflectivity.package/RFCondition.class/instance/accessing/valueSelector.st A Reflectivity.package/extension/BlockClosure/instance/valueSelector.st A Reflectivity.package/extension/Boolean/instance/valueSelector.st R ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50059.st A ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50060.st R ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50059.st A ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50060.st M ScriptLoader50.package/ScriptLoader.class/instance/public/commentForCurrentUpdate.st R SmartSuggestions.package/SugsAbstractContext.class/instance/tools/tools.st M SmartSuggestions.package/SugsAccessesToInstVar.class/instance/as yet unclassified/execute.st M SmartSuggestions.package/SugsBrowseClassReferences.class/instance/as yet unclassified/execute.st M SmartSuggestions.package/SugsFindImplementors.class/instance/as yet unclassified/execute.st M SmartSuggestions.package/SugsFindSenders.class/instance/as yet unclassified/execute.st M SmartSuggestions.package/SugsStoresIntoInstVar.class/instance/as yet unclassified/execute.st R SmartSuggestions.package/extension/AbstractTool/instance/browseAllAccessesTo_from_.st R SmartSuggestions.package/extension/AbstractTool/instance/browseSendersOf_name_autoSelect_.st Log Message: ----------- 50060 10120 StrikeFont default writeAsStrike2On: strikefont.bin fails - DNU https://pharo.fogbugz.com/f/cases/10120 15587 clean up SugAbstractContext>>#tools https://pharo.fogbugz.com/f/cases/15587 15591 Generate Hooks for links with conditions that get arguments passed in https://pharo.fogbugz.com/f/cases/15591 http://files.pharo.org/image/50/50060.zip

May 22, 2015

Re: [Pharo-dev] Opportunity to steal good ideas from dad ;)

by Esteban A. Maringolo

2015-05-22 6:43 GMT-03:00 Sven Van Caekenberghe <sven(a)stfx.eu>: >> On 21 May 2015, at 16:50, Esteban A. Maringolo <emaringolo(a)gmail.com> wrote: >>> About the spaceship operator (<=>), is this something we'd like to have >>> around? (I like it). http://en.wikipedia.org/wiki/Three-way_comparison >> http://www.smalltalkhub.com/#!/~emaringolo/SortFunctions >> I use them extensively. > I had a look. > > Apart from the fact that for some reason, the extensions on MAElement are in the Core package, hence loading just Core gave errors, things looked OK. > > This is generally useful, small and a very good example of super cool OO design. > > I would vote to integrate this (<=> and all). For some reason the extensions of MAElementDescription where both in SortFunctions-Core and SortFunctions-Magritte. Also the dependency to Magritte in the Metacello config was wrong. Now it works as expected (please test :)). And a new stable version was uploaded to the smalltalkhub repo. Please notice my port includes the option to collate the undefined elements at the beginning or the end, in a similar fashion to ORDER BY [column] NULLS [LAST|FIRST] in SQL (Oracle/PostgreSQL at least). I would also vote to have this in the base image. It is my base image at least :) Regards! -- Esteban A. Maringolo

May 22, 2015

Re: [Pharo-dev] can someone explain me this?

by Alejandro Infante

If you remove the Traits you get exactly the double (at least in my moose image). As far as I remember, Traits are not subclasses of Objects and they do not have a metaclass for each Trait. Cheers, Alejandro > On May 22, 2015, at 12:39 PM, Eliot Miranda <eliot.miranda(a)gmail.com> wrote: > > Oops, my math was wrong and so I got it wrong way round. But you know what I mean... > > Eliot (phone) > > On May 22, 2015, at 8:37 AM, Eliot Miranda <eliot.miranda(a)gmail.com> wrote: > >> Hi All, >> >> On May 21, 2015, at 1:14 PM, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: >> >>> Smalltalk allClassesAndTraits size. "5057" >>> Object withAllSubclasses size. â9900" >> >> As others have said, meta classes account for many of these, but I would have expect something closer to twice as many. What does this tell you? >> >> (IdentitySet withAll: (Object withAllSubclasses reject: [:b| b isMeta]) >> removeAll: Smalltalk allClassesAndTraits; >> yourself >> >> Are there perhaps lots of obsolete classes that need cleaning up? >> >> >>> I really do not understand :) >>> >>> Esteban >

May 22, 2015

Re: [Pharo-dev] can someone explain me this?

by Eliot Miranda

Oops, my math was wrong and so I got it wrong way round. But you know what I mean... Eliot (phone) On May 22, 2015, at 8:37 AM, Eliot Miranda <eliot.miranda(a)gmail.com> wrote: > Hi All, > > On May 21, 2015, at 1:14 PM, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > >> Smalltalk allClassesAndTraits size. "5057" >> Object withAllSubclasses size. â9900" > > As others have said, meta classes account for many of these, but I would have expect something closer to twice as many. What does this tell you? > > (IdentitySet withAll: (Object withAllSubclasses reject: [:b| b isMeta]) > removeAll: Smalltalk allClassesAndTraits; > yourself > > Are there perhaps lots of obsolete classes that need cleaning up? > > >> I really do not understand :) >> >> Esteban

May 22, 2015

Re: [Pharo-dev] can someone explain me this?

by Eliot Miranda

Hi All, On May 21, 2015, at 1:14 PM, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > Smalltalk allClassesAndTraits size. "5057" > Object withAllSubclasses size. â9900" As others have said, meta classes account for many of these, but I would have expect something closer to twice as many. What does this tell you? (IdentitySet withAll: (Object withAllSubclasses reject: [:b| b isMeta]) removeAll: Smalltalk allClassesAndTraits; yourself Are there perhaps lots of obsolete classes that need cleaning up? > I really do not understand :) > > Esteban

May 22, 2015

Cross platform key modifiers

by Ben Coman

Not sure what past history we inherited, but I am questioning the semantics of our key modifiers. My understanding is that the following is industry standard... OSX Windows Linux option alt alt cmd ctrl ctrl with Copy being the simplest example of this... cmd-c ctrl-c ctrl-c Supporting this view: * "option" and "alternative" are synonyms * "command" and "control" have similar meaning (e.g. in the army if you command someone, you control them) * Microsoft say so (https://support.microsoft.com/en-us/kb/970299) * Apple say so (https://support.apple.com/kb/PH18812?locale=en_US) and indeed marks its keyboards with "alt" and "option" on the same key ( http://www.apple.com/au/keyboard/) So examining this with KMLog setDebug.Transcript open. On OSX Pharo 50044 I get... option+arrowLeft --> [keystroke '<Opt-left>'] cmd+arrowLeft --> [keystroke '<Cmd-left>'] ctrl+arrowLeft --> OS changes screens but I am told that Windows gives... alt+arrowLeft --> [keystroke '<Cmd-left>'] ctrl+arrowLeft --> [keystroke '<Ctrl-left>'] when I believe Windows should be... alt+arrowLeft --> [keystroke '<Opt-left>'] ctrl+arrowLeft --> [keystroke '<Cmd-left>'] Part of the issue I think is that our keystroke mapping should be more abstract. When I was using Pharo on Windows, sometimes it felt like a second class citizen wrt key mappings, mapping ctrl-c to an apple-named <cmd-c>. Perhaps what we should see is: OSX... option+arrowLeft --> [keystroke '<Meta-left>'] cmd+arrowLeft --> [keystroke '<Hotkey-left>'] Windows... alt+arrowLeft --> [keystroke '<Meta-left>'] ctrl+arrowLeft --> [keystroke '<Hotkey-left>'] Otherwise there will always be the tension on windows that ctrl-c should generate [<keystroke '<Ctrl-c>] and we end up with definitions repetitively formed like... PharoShortcuts>>acceptShortcut ^ $s ctrl win | $s ctrl unix | $s command mac which opens the door for inconsistencies. We should only need to do something like... PharoShortcuts>>acceptShortcut ^ $s hotkey and leave it to the platform default or user preference the define what the hotkey actually is. A status line in the Keymap Browser could display which actual real keys map to <Hotkey> and <Meta>. What are you thoughts? cheers -ben

May 22, 2015

Re: [Pharo-dev] UI Design

by Dimitris Chloupis

I am looking into implementing a futuristic GUI for Pharo but will be a very long term project since I am still quite unfamiliar with Morphic. The effort has started with my Blue theme but will expand the theme to make it customise Pharo a lot deeper. My goal is to also make the theme as dynamic as I can, currently this already possible with the small wizard I offer for customizing the colors. The project is called Nireas and is available through the Configuration Browser. Its very basic right now but will slowly grow larger and more sophisticated. On Fri, May 22, 2015 at 4:23 PM Torsten Bergmann <astares(a)gmx.de> wrote: > Is there something new planned for Pharo 5 regarding the UI look? > New theme or integration of an animation framework to support better > looking Pharo apps? > > Looks like material design is the next fashionable UI hype: > > https://www.youtube.com/watch?v=7xH26aGqHDM > https://www.youtube.com/watch?v=6XSyBV0h2n4 > https://www.youtube.com/watch?v=6XSyBV0h2n4 > > Anyone working on something for rich client apps in Pharo? > > Thanks > T. > >

May 22, 2015

UI Design

by Torsten Bergmann

Is there something new planned for Pharo 5 regarding the UI look? New theme or integration of an animation framework to support better looking Pharo apps? Looks like material design is the next fashionable UI hype: https://www.youtube.com/watch?v=7xH26aGqHDM https://www.youtube.com/watch?v=6XSyBV0h2n4 https://www.youtube.com/watch?v=6XSyBV0h2n4 Anyone working on something for rich client apps in Pharo? Thanks T.

May 22, 2015

Re: [Pharo-dev] [Reflectivity] Progress for Breakpoints

by Clara Allende

2015-05-22 8:53 GMT-03:00 Marcus Denker <marcus.denker(a)inria.fr>: > > https://pharo.fogbugz.com/f/cases/15516/add-suggestion-for-breakpoint > > this is of course completely untestedâ¦ it works only when there is no > unaccepted > text in the editor and maybe only within Nautilus. > > And of course existing breakpoints are not yet in any form visualised in > the editor. > > > This is in 50041. Please use with careâ¦ still needs work to be really > usable. > > > I now finished to add everything from Claraâs original SmartBreakpoints > project. > > There are still lots of things that need to be done > > -> Something goes wrong when re-adding a one-shot breakpoint again. > Yep, I'm aware of that bug, I couldn't figure out why it does it happen :/ At one point I got it work and randomly fail, so it was difficult to debug :( If right now it fails consistently, I would like to take a look. > -> Adding should be disabled when the editor has non-submitted code. > -> The experience in the debugger is not that nice > (need to think about using the <debuggerCompleteToSender> trick and > explore the filtering capabilities of the debugger) > I'm aware of this one too. I did add a filter for the Reflectivity/breakpoints contexts, but in my experience, the top most of those contexts remained on the debugger stack (because we need it in case we want to go back to that point after a lot of step into/out etc), making it a little bit confusing... I mean, when I saw it, I put myself in the place of the user, and could imagine him/her wondering wtf does that mean. I agree we need to re-think it :/ > > I am now continuing to flesh out the Reflectivity features: > > -> hook AST generation is now done by HookGenerator > -> correct hooks are already generated for the case that a link passes > arguments to the meta object (see the RFReification hierarchy, for now > we have only #object *and* actual compiling does not yet work due to > missing > semantic analysis of the AST snippets for the hooks) > -> We have support for âvirtualâ meta objects. Any valid reification can > be the metaObject > (this is fun for #node). > > TODO > - add all reifications for all nodes (#arguments, #node, #context, #value > #name, #arg1â¦) > - do semantic analysis correctly and add test for links with arguments > - add conditions with parameters (e.g. for object specific links: [:object > | object == myObject]) > - handle primitives: when putting a link on a primitive method, wrap the > compiledMethod instead > of AST manipulation. > - â¦. > > I will continue to flesh it out and send examples to the list as these > things start to work for real. > > > Marcus > >

May 22, 2015

Re: [Pharo-dev] Best practice for overwriting methods in other classes

by Ben Coman

On Fri, May 22, 2015 at 1:08 PM, Kasper Osterbye <kasper(a)itu.dk> wrote: > I have now twice in two different situations felt a need to "steal" a > method > in Nautilus and make that method an extension method in my own package. > > In situation A, I wanted to steal the #buildCommentPane to use a different > Morph for class comments, > I've not encountered this myself, but I guess the approach would be to get a change integrated into Pharo that allows an external package to configure it to use a different Morph -- though I'm not sure of the best approach to do that. > In situation B, Sergio and I needed to redefine #getComments and > #addComments:notifying: to fill out behaviour for package comments. > > In this case, I thought the aim of your package-comments work was to integrate it into Pharo, so actually you should be redefining the built-in methods as needed (??) cheers -ben What is the kosher way to work and contribute in such situations? I guess it > will also pop up in situations outside of Nautilus. > >

May 22, 2015