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

Re: [Pharo-project] What about making the configuration browser more visible?

by StÃ©phane Ducasse

Let me repeat it again (reread the vision document) we will put in place a process that: - check all the configurations of a given repository (for each versions) so for example for MetaRepoForPharo1.4 and MetaRepoForPharo20â¦ - each configuration will be loaded, test runs, rules executed. - we will have inbox for each repository, - if a configuration does not pass, it will move the configurations into an outbox Then people will be able to load configurations that have been loaded and verified. Stef

May 2, 2012

Re: [Pharo-project] Block questions

by StÃ©phane Ducasse

> > > I mean the example is not a useful example because it doesn't show the problem "my" representation solves, which is to make closures independent of their enclosing activations if they outlive their dynamic extent. Yes I'm starting to finish the chapter explanation on explicit return and now I will go in "opening the trunk section" and study for real your implementation because I really want to fully understand it. Stef > > > (define (%apply proc largs) > (cond ((%primitive? proc) (%apply-internal proc largs)) > ((%closure? proc) (%apply-closure proc largs)) > (else (error "Bad ! Un-apply-able object !" proc)))) > > (define (%apply-closure proc largs) > ;; apply a closure: evaluate proc body in > ;; extended the closure environment with > ;; proc arguments and largs > (%eval (%closure-body proc) > (%extend-env (%closure-args proc) largs (%closure-env proc)))) > > (define (%eval-lambda expr env) > ;; lambda in Scheme captures the environment at compile time > ;; expr = ((x) (+ x 2)) > (%make-closure (car expr) (cadr expr) env)) > > > You need to consider how to implement this example: > > > > Object subclass: #AbstractSuperclass > > > > AbstractSuperclass>>computeV > > | v | > > v := 0. > > self maybeCapture: [v := v + 1. v]. > > ^self maybeBlock: v > > > > AbstractSuperclass subclass: #CaptureBlock > > instanceVariableNames: 'block' > > > > CaptureBlock>>maybeCapture: aBlock > > block := aBlock > > > > CaptureBlock>>maybeBlock: value > > ^block > > > > AbstractSubclass subclass: #DontCaptureBlock > > > > DontCaptureBlock>>maybeCapture: aBlock > > > > DontCaptureBlock>>maybeBlock: value > > ^value > > > > > > Then > > | thingOne thingTwo | > > thingOne := CaptureBlock computeV. > > thingTwo := DontCaptureBlock computeV. > > (1 to: 10) collect: [:ignore| { thingOne value. thingTwo value }] > > > > should answer > > #((1 0) (2 0) (3 0) ... (10 0)) > > if 0 value returns 0? yes. (it confused me for a while). > > > right? So where does the l-value for v in computeV live? Iy must be an l-value since it is assigned to in the block in computeV. It must outlive the activation of computeV since CaptureBlock>maybeCapture: captures the block and CaptureBlock>>maybeBlock: answers it and the doit evaluates it. If it lives on the stack of the activation of computeV then there is no problem in a Context VM, but in a VM that maps activations to stack frames something special (and slow) has to happen when returning from computeV. However, if the l-value lives in a separate Array (as it does in my VisualWorks and Squeak closure implementations, and as happens in some Lisp implementations), nothing special has to happen. The block refers to the Array, *not* to the activation of computeV. > > > > Make sense now? > > Yes you allocate on the heap closure binding to be orthogonal to activation context. > > > I will reread your blog now. > > Stef > > > > -- > best, > Eliot >

May 2, 2012

Re: [Pharo-project] Pretty printing on Nautilus

by StÃ©phane Ducasse

Davide could you provide a description and screenshots for the website and success stories. Stef On Apr 30, 2012, at 7:57 AM, Davide Varvello wrote: > Thanks Mariano. > I've zero time now :-) I'm implementing a smalltalk web app for the Italian > market, if you can understand a little bit of Italian you can see a demo > video here: http://bit.ly/JVeWYe > Davide > > -- > View this message in context: http://forum.world.st/Pretty-printing-on-Nautilus-tp4596333p4597485.html > Sent from the Pharo Smalltalk mailing list archive at Nabble.com. >

May 2, 2012

Re: [Pharo-project] Pretty printing on Nautilus

by StÃ©phane Ducasse

Why not using or enhancing the pretty printing of RB? Because lukas invested a lot of time into it. Stef On Apr 29, 2012, at 10:28 PM, Benjamin wrote: > I will implement a pretty print. > But the thing is since I am not the god of pretty print, I definitely want a mechanism to let every user to specify it's own way to pretty print. > Without that, it's useless for me :) > > Ben > > On Apr 29, 2012, at 2:59 PM, Davide Varvello wrote: > >> Hi, >> Today I have a lot of "cosmetic" questions :-) >> How can I enable pretty printing on Nautilus? Moreover, the old cmd-r does >> not work anymore, right? >> Thanks >> Davide >> >> -- >> View this message in context: http://forum.world.st/Pretty-printing-on-Nautilus-tp4596333p4596333.html >> Sent from the Pharo Smalltalk mailing list archive at Nabble.com. >> > >

May 2, 2012

Re: [Pharo-project] Artefact - a PDF framework

by Christian Haider

> Von: pharo-project-bounces(a)lists.gforge.inria.fr [mailto:pharo-project- > bounces(a)lists.gforge.inria.fr] Im Auftrag von Nicolas Cellier > > What are we speaking of exactly ? > Here are two different features of different level : I'll answer for PDF4Smalltalk > 1) the composition : > - 1a) compose a document with a given structure and a style sheet > For the structure, think of section / subsection etc... No (use Report4PDF) > For the style sheet, think of LaTeX article, book or report.sty No > For the composition, think of page layout (when to insert page break, > compose on several columns, wrap text around figures etc...). No (use Report4PDF) > High level composition rules for example might forbid to begin a sub- > section at the bottom of a page. No (use Report4PDF) > - 1b) compose a document from a sequence of lower level graphics (including > composed paragraph). > The composition rule might be very simple, just compose one item over > the other, insert page break if it does not fit on page. No (use Report4PDF) > For composed text, page break can occur at any composed line or only at a > carriage return. No (use Report4PDF) > A good example was VisualWorks Document class. > 2) the fileout of the composition in pdf format > - 2a) including PDF document structuring conventions (DSC) if ever they exist > at Smalltalk level. Fully supported > - 2b) without DSC, but only low level graphics Fully supported > > Composing the paragraph also means having all the font metrics available > plus eventually all the kerning rules, plus rules for inter-word spacing (non > speaking of alignment and tabulation options which should be part of > paragraph spec). No, but Type-1 fonts and OpenType(PS) fonts are implemented so that this information is available. > > Which PDF library is doing what exactly ? > Are the low level Graphics > - entirely PDF library dependent (which would mean no preview via Smalltalk > graphics), Yes > - entirely dialect dependent (using exclusively > VisualPart/Wrapper/GraphicsContext or Morph/Form/Canvas) No > - mixed (transformation of PDF graphics into Smalltalk graphics for preview > and/or translation of Smalltak graphics to PDF graphics) In the last two cases, > porting VW <->Squeak/Pharo might not be immediate ;) No, there is now viewing - just PDF objects Think of PDF4Smalltalk as a backend for graphics programs which do layout and such. With TrueType fonts and bitmap graphics, it should be easy to hook it up as GraphicsContext for VW or as Canvas(?) for Squeak/Pharo. Christian > > What about composition? > > Nicolas > > 2012/5/2 Yanni Chiu <yanni(a)rogers.com>: > > On 02/05/12 10:53 AM, Sean P. DeNigris wrote: > >> > >> > >> Stephan Eggermont wrote > >>> > >>> > >>> Don't you think it would be much less work to port Christian > >>> Haider's work to Pharo? > >> > >> > >> I was thinking the same thing. The usual Smalltalk duplication of > >> effort seems to have been taken to new heights with PDF libs ;-) > > > > > > I've loaded Artefact, browsed the code, tried the demos. > > > > I've not seen the code for PDF4Smalltalk, because I can't figure out > > how to browse it without VW. However, looking at the examples at: > > Â https://gitorious.org/pdf4smalltalk/pages/Examples > > I get the (possibly mistaken) impression that the two frameworks do > > slightly different things, but the final result is a PDF in both cases. > > > > Looking at the PDFDemos of Artefact, it's clear how text and tables > > can be written. However, looking at the PDF4Smalltalk examples, it > > looks like you have to position the text and tables in your own code. > > Of course, I may be completely wrong about PDF4Smalltalk, since I've > > not seen the full code base. > > > > > >

May 2, 2012

Re: [Pharo-project] [Fwd: Re: [Newbies] Remove trailing spaces of aString]

by StÃ©phane Ducasse

Ben the original UIFinder was showing the methods effectively executed using a * and the implementors without. Now when you show the results you do not distinguish any of them. Stef On Apr 29, 2012, at 1:42 PM, Benjamin wrote: > :) > > Good idea ;) > I will do that, it shouldn't be so difficult ^^ > > Ben > > On Apr 29, 2012, at 5:00 AM, Ben Coman wrote: > >> Ah. ha... Very Good. For the whole time I've used Pharo the last 12 months I've had a misconception about Examples. >> I thought they must have been Example coded somewhere like GLMBasicExamples, the MOEaselExamples and these somehow flagged for searching - and that I never got any results because of some lacking implementation. >> >> Here is some usability feedback then... >> >> 1. The help text in the bottom pane says "for further informations, print 'FinderUI methodFinderExplanation' ". >> Rather better would be when [Examples] is selected, that bottom pane is filled with "print 'FinderUI methodFinderExplanation". >> 2. The help text in the bottom pane says "In these four modes, you can also tick the 'Use RegEx' checkbox." >> might say "In the first four modes and sort [Examples] to appear at the bottom of the list. >> Actually that text is a carry over from when Pramgas was not an option and only the first three modes had RegEx. >> >> 3. The help text in the bottom pane says "The 'Select classes'...." and "The 'All classes'...." >> but the actual buttons are <Choose Packages> and <All packages> >> >> 4. Optional extra, while <Regexp> is selected, put some help text for regular expressions with examples in the bottom pane. >> >> I've logged http://code.google.com/p/pharo/issues/detail?id=5733 >> >> cheers -ben >> >> Benjamin wrote: >>> If you use Pharo, you can also do that >>> >>> >>> ------------------------------------------------------------------------ >>> >>> ------------------------------------------------------------------------ >>> >>> >>> >>> Note that you have to change 'Selector' with 'Example', and then you can type a list of arguments separated by a dot and the expected result. >>> (If you want some inner details, it will do something like methodFor: {_WHAT_YOU_HAVE_ENTER_}) >>> >>> Have fun, >>> Ben >>> >>> >>> >>> On Apr 28, 2012, at 4:00 PM, Ben Coman wrote: >>> >>> >>>> I just discovered this new "bit of magic" from the Newbies list. My first impression playing with it is that MethodFinder>>methodFor: should be somehow more prominent in Tools>Finder and also in PharoByExample book. Quite a unique feature and demonstration of Smalltalk power for newcomers. >>>> >>>> cheers -ben >>>> >>>> Am 2012-04-27 um 17:27 schrieb H. Hirzel: >>>> >>>> >>>>> Hello >>>>> I use Squeak 4.3, Update 11860. >>>>> I want to remove the trailing spaces and or tabs and LF of a String >>>>> something like >>>>> 'abc >>>>> ' >>>>> should just be >>>>> 'abc'. >>>>> If I remember that there was a method in class String to something >>>>> like this. Is this correct? >>>>> Is there a package with additional String convenience methods? >>>>> >>>> you probably want >>>> >>>> 'abc ' withoutTrailingBlanks >>>> >>>> You can find that method by evaluating or printing >>>> >>>> MethodFinder methodFor: { {'abc '} . 'abc' } (a list with arguments {'abc '} and an expected result ( 'abc' )) >>>> >>>> HTH >>>> >>>> Best >>>> -Tobias >>>> _______________________________________________ >>>> Beginners mailing list >>>> Beginners(a)lists.squeakfoundation.org >>>> http://lists.squeakfoundation.org/mailman/listinfo/beginners >>>> >>>> >>>> >>>> >>> >>> >> >> > >

May 2, 2012

Re: [Pharo-project] Adding hooks for well known events in morphic

by StÃ©phane Ducasse

Indeed rationalizing the hooks and removing the changed/update mechanism would be good. On Apr 27, 2012, at 8:58 PM, Guillermo Polito wrote: > I've made up a list with all the useful morphic widget I've found in a alpha alpha alpha 2.0 image. > I'd like to put in those widgets empty hooks like: > > Morph>>triggerOnShow > "hook" > > for common events widgets provide usually in MVC frameworks. > That means not changing the whole interface nor refactoring everything. Just adding the hooks. What we were talking with Mariano (I hate you for putting ideas in my head) today in the train was to add those empty hooks and put senders for them, what should be easy. This way we can start to homogenize morphic a bit. > > Most likely, in a not so far future, those empty hooks could be replaced by something like: > > Morph>>triggerOnShow > self announcer announce: (MorphShowAnnouncement show: self) > > Now, here is a list of the most important widgets i've found (I could've missed some) and events they should trigger. I'm not taking into account new multitouch interfaces and stuff. > > For all widgets > onShow > onDelete > onKeyUp > onKeyPress > onKeyDown > onClick > onMouseDown > onMouseUp > onDoubleClick > onDragged > onDropped > onMove? > onResize? > > For all containers > onChildAdded > onChildRemoved > > "Widgets" > > Check/Radio -> CheckboxButtonMorph/SimpleSwitchMorph > onSelectionChanged > > List/Grids -> PluggableListMorph/PluggableMultiColumnListMorph/LazyListMorph/MulticolumnLazyListMorph > onSelectionChanged > > Button -> PluggableButtonMorph/SimpleButtonMorph/MultistateButtonMorph > onClick > > Label -> LabelMorph/FuzzyLabelMorph /MultistateLabelMorph > > Progress Bar-> SystemProgressBarMorph > > Slider -> Slider/PluggableSliderMorph > onSelectionChanged > > Text handling -> TextMorph/PluggableTextMorph > onSelectionChanged > > Field Handling -> PluggableTextFieldMorph/PluggableTextEditorMorph > > Tree -> MorphTreeMorph > onSelectionChanged > onExpandItem > onCollapseItem > > Combo box -> DropListMorph/EditableDropListMorph /SearchMorph/MorphDropListMorph > onSelectionChanged > > Menues -> MenuMorph + friends > onClick (for menu item probably) > > Image -> ImageMorph/ImagePreviewMorph > > Diff -> DiffMorph > > Color Presenter -> ColorPresenterMorph > > Color chooser -> ColorChooserMorph/HSVColorSelectorMorph / SVColorSelectorMorph > onSelectionChanged > > Task var -> TaskbarMorph > > > "Containers" > ScrollPane > Windows -> SystemWindow/PluggableSystemWindow /StandardWindow > Dialog window -> DialogWindow & (lots of) subclasses > Panel -> PanelMorph & subclasses / PluggablePanelMorph > > Tabs -> TabGroupMorph (Other Tab stuff -> PluggableTabBarMorph + PluggableTabButtonMorph ) > onSelectionChanged > > > Guille

May 2, 2012

Re: [Pharo-project] Adding hooks for well known events in morphic

by StÃ©phane Ducasse

On Apr 28, 2012, at 2:39 PM, GastÃ³n Dall' Oglio wrote: > Hi Guillermo. > > When you combine these triggers with ads (as in your example), I see an analogy with the model of qt ui, see: > http://qt-project.org/doc/qt-4.8/signalsandslots.html > > In effect: > "A signal is emitted when a particular event occurs." -> A Announcement is send when a particular event (trigger) occurs. > " Qt's widgets have many predefined signals, but we can always subclass widgets to add our own signals to them" -> Morphic widgets have many predefined triggers, but we can always subclass widgets to add our own triggers to them. > "A slot is a function that is called in response to a particular signal. " -> A AnnouncementSubscription is ? that is used in response to a particular announcement. > > It's a mechanic for decouple between sender and interested in event occurrence. > > > As you surely know Morphic Designer follows basic ideas of the QtDesigner: > https://www.hpi.uni-potsdam.de/hirschfeld/trac/SqueakCommunityProjects/wiki… > > Look at doc here (I never see code of this project), this approach not use announcement framework, but a signal implementation: > https://www.hpi.uni-potsdam.de/hirschfeld/trac/SqueakCommunityProjects/wiki… I do not see the difference between signals and event. And yes we look at the code, it is not a really good idea to use thisContext for every little notifications.

May 2, 2012

[Pharo-project] NativeBoost help

by Luc Fabresse

Hi Igor and other NB Gurus, A group of students is trying to use NativeBoost to wrap the OpenCV C image library. The problem we are facing is that the function cvSize return a struct. typedef struct { int width; int height; } CvSize; CV_INLINE CvSize cvSize( int width, int height ); But the NBExternalStructure seems to not support that if I understood correctly. NBExternalStructure>>coerceReturn: gen " ... should we support handling return of external structures? " self error: 'returning pointer to structure?' Did I miss something? Any idea or workaround? Thanks, Luc

May 2, 2012

Re: [Pharo-project] Artefact - a PDF framework

by Nicolas Cellier

What are we speaking of exactly ? Here are two different features of different level : 1) the composition : - 1a) compose a document with a given structure and a style sheet For the structure, think of section / subsection etc... For the style sheet, think of LaTeX article, book or report.sty For the composition, think of page layout (when to insert page break, compose on several columns, wrap text around figures etc...). High level composition rules for example might forbid to begin a sub-section at the bottom of a page. - 1b) compose a document from a sequence of lower level graphics (including composed paragraph). The composition rule might be very simple, just compose one item over the other, insert page break if it does not fit on page. For composed text, page break can occur at any composed line or only at a carriage return. A good example was VisualWorks Document class. 2) the fileout of the composition in pdf format - 2a) including PDF document structuring conventions (DSC) if ever they exist at Smalltalk level. - 2b) without DSC, but only low level graphics Composing the paragraph also means having all the font metrics available plus eventually all the kerning rules, plus rules for inter-word spacing (non speaking of alignment and tabulation options which should be part of paragraph spec). Which PDF library is doing what exactly ? Are the low level Graphics - entirely PDF library dependent (which would mean no preview via Smalltalk graphics), - entirely dialect dependent (using exclusively VisualPart/Wrapper/GraphicsContext or Morph/Form/Canvas) - mixed (transformation of PDF graphics into Smalltalk graphics for preview and/or translation of Smalltak graphics to PDF graphics) In the last two cases, porting VW <->Squeak/Pharo might not be immediate ;) What about composition? Nicolas 2012/5/2 Yanni Chiu <yanni(a)rogers.com>: > On 02/05/12 10:53 AM, Sean P. DeNigris wrote: >> >> >> Stephan Eggermont wrote >>> >>> >>> Don't you think it would be much less work >>> to port Christian Haider's work to Pharo? >> >> >> I was thinking the same thing. The usual Smalltalk duplication of effort >> seems to have been taken to new heights with PDF libs ;-) > > > I've loaded Artefact, browsed the code, tried the demos. > > I've not seen the code for PDF4Smalltalk, because I can't figure out how to > browse it without VW. However, looking at the examples at: > Â https://gitorious.org/pdf4smalltalk/pages/Examples > I get the (possibly mistaken) impression that the two frameworks do slightly > different things, but the final result is a PDF in both cases. > > Looking at the PDFDemos of Artefact, it's clear how text and tables can be > written. However, looking at the PDF4Smalltalk examples, it looks like you > have to position the text and tables in your own code. Of course, I may be > completely wrong about PDF4Smalltalk, since I've not seen the full code > base. > >

May 2, 2012