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

[Pharo-project] UIManager default chooseDirectory:
by Alexandre Bergel June 10, 2011

June 10, 2011

Hi! Pharo1.3a Latest update: #13214 Steps to reproduce: 1. do a printIt: UIManager default chooseDirectory: 'Choose a directory to store the image'. 2. It returns nil before even you select a directly. this is obviously wrong. This has recently changed. Before chooseDirectory: was waiting for you to select a directory before returning a value, the directory corresponding to what you have selected It would be cool to fix this. But I am not knowledgable at modal windows... Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

2 1

[Pharo-project] Have you played with Ameba?
by HwaJong Oh June 10, 2011

June 10, 2011

"Load package" Gofer new squeaksource: 'DaliotsPlayground'; package: 'Ameba'; load. "run objects" x := Ameba example. "kill objects" x do: #delete. -- View this message in context: http://forum.world.st/Have-you-played-with-Ameba-tp3585624p3585624.html Sent from the Pharo Smalltalk mailing list archive at Nabble.com.

5 8

[Pharo-project] [update 1.3] #13258
by Marcus Denker June 10, 2011

June 10, 2011

13258 ----- Issue 4375: TesRunner: DNU #index http://code.google.com/p/pharo/issues/detail?id=4375 Issue 4380: Browse hierarchy from an inspector generates a DNU http://code.google.com/p/pharo/issues/detail?id=4380 -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

1 0

Re: [Pharo-project] Tab widget improvements
by Marcus Denker June 10, 2011

June 10, 2011

On Jun 8, 2011, at 3:55 PM, Gary Chambers wrote: > Attached, some improvements to the TabGroupMorph etc. having had feedback from Doru. > > Provides scrolling buttons when insufficient room for all tabs and displays only tabs for which there is room! (excepting at least one). > > Can optionally set wrapScrolling on the TabSelectorMorph to allow wraparound scrolling via the buttons. > Do we put that in 1.3 or 1.4? http://code.google.com/p/pharo/issues/list?thanks=4376 -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

5 16

[Pharo-project] Deleting packages in squeaksource.com?
by HwaJong Oh June 10, 2011

June 10, 2011

Hi, I want to delete some packages that are in my squeaksource repository. They are either empty or old or part of aborted project. I want to keep my repository neat and focused. Couldn't find any buttons in Monticello Browser nor links in squeaksource.com. Plz Help. Best Regards HwaJong Oh -- View this message in context: http://forum.world.st/Deleting-packages-in-squeaksource-com-tp3587292p35872… Sent from the Pharo Smalltalk mailing list archive at Nabble.com.

3 2

[Pharo-project] FFI versus Alien versus NativeBoost (was FFI versus Aloen)
by Chris Cunnington June 10, 2011

June 10, 2011

// >What shortcomings? I believe there are three outlined in the description on SqueakSource: "It supports faster call sequence, return code setting, and callbacks to execute smalltalk code from the FFI routine." There's something about that last one: callbacks. I think FFI gives a return value, but you can't create a callback. > NativeBoost is faster than FFI, but NB is a whole other ball-of-wax, and I'd like to support it eventually as well as FFI/Alien. Another way to look at it is via what Goran said: in terms of age. FFI is oldest and most well known. Alien less so. NativeBoost is Igor's project and I don't think is widely adopted yet. (Igor's going to show up and squish me like a bug, but oh well... ) Igor did say something interesting about FFI versus plugins at Deep Into Smalltalk. I believe he wrote Native Boost because he was tired of endlessly recompiling vms with plugins. I don't think there is any standard vm with NB in it, which means that's exactly what you're going to need to do to use NB. FFI has a reputation as something that's a bit touchy. It looks a little easier than it is. The prudent choice would be for you to use FFI and curb your ambition away from Alien or NB until later. There are lots more examples of FFI around than Alien and NB combined. OpenSophie has copious code examples. http://opensophie.org/ I have to stop now. I'm only making it worse for when Igor arrives... must go! :p Chris

2 1

[Pharo-project] DNU in "Browse hierarchy" from Inspector
by Diego Fernandez June 9, 2011

June 9, 2011

The issue is reported on: http://code.google.com/p/pharo/issues/detail?id=4380 With a fix attached to it.

1 0

Re: [Pharo-project] [Pharo-users] Pharo Talk ESUG 2010
by Marcus Denker June 9, 2011

June 9, 2011

On Oct 7, 2010, at 1:23 AM, GermÃ¡n Arduino wrote: > hehe, thanks! > > You will come to Smalltalks 2010 ? Yes, Mariano, Luc and me. > > I was thinking exactly in ask permit to use this one (and other > similars) to present Pharo in other contexts, for example in localgroups, etc. Normally all my talks are CC non-commercial share-alike... only that I often forget to put it in the slides... > That's perfectly allowed and even more, a ppt/keypoint template should be available. 2 days ago Marcus borrowed this one for my PhD first year evaliuation ;) Ok, I have put the keynote (zipped): http://marcusdenker.de/talks/10ESUG/ -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

4 3

Re: [Pharo-project] [GS/SS Beta] [ANN] Google Analytics Tracker, a small tool
by Lukas Renggli June 9, 2011

June 9, 2011

Very cool. I especially like the XPath parser :-) Lukas On 9 June 2011 13:41, Nick Ager <nick.ager(a)gmail.com> wrote: > Hi > This looks great - looking forward to integrating it. > Cheers > Nick > > On 8 June 2011 21:50, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: >> >> Hi, >> I want to announce the release of Google Analytics Tracker, a small tool >> we make for getitmade.com (Nick Ager's project), to allow us to get >> analytics data. >> Right now is very simple, it just give us page visits data, but can be >> used as a foundation to get a lot more info. >> >> You get it by: >> >> Gofer it >> Â Â Â Â squeaksource: 'gat'; >> Â Â Â Â package: 'ConfigurationOfGoogleAnalyticsTracker'; >> Â Â Â Â load. >> (Smalltalk at: #ConfigurationOfGoogleAnalyticsTracker) load. >> >> To see how it works just see GAAccess class>>#example. >> >> It uses Zinc and PetitParser, so if gemstone can load both packages, it >> will work without any problems (I think it will, but I will check it >> tomorrow... maybe with a different configuration). >> >> You will be able to add to your pages something like the image I'm sending >> :) >> >> Cheers, >> Esteban >> >> >> >> >> > > -- Lukas Renggli www.lukas-renggli.ch

3 2

Re: [Pharo-project] [Pharo-users] About GUI tests [was: TWM: Multiple World integration]
by laurent laffont June 9, 2011

June 9, 2011

> > > >> > >> I need to write tests for this. A cool thing is that with TWM I've > learned a lot on how to test GUIs, Worlds, ... > > > > I'm interested in these tests. > > Me too. > Some words about what I've learned. As TWM move / resize windows, it's better to create a World dedicated to the tests, so you don't see all your windows moving while testing (and if you use Autotest note that tests are run as soon as you save a method, quite fun :). So in TWM have a look at WMInANewWorldTestCase>>setUp and some of its subclasses like: - WMInANewWorldTestCase |_ WMWithTwoBrowsersAndTwoWorkspacesTest>>setUp |_ TWMBarWithTwoBrowsersAndTwoWorkspacesTest>>setUp If I merge the setUps: testWorld := PasteUpMorph new initAsTestWorld: WorldState new; yourself. testWorld bounds: (0@0 extent: 200@200). windowManager := TilingWM forWorld: testWorld. browser1 := self newBrowserIn: testWorld. browser2 := self newBrowserIn: testWorld. workspace1 := self newWorkspaceIn: testWorld. workspace2 := self newWorkspaceIn: testWorld. Then it should be easier to understand something like: WMWithTwoBrowsersAndTwoWorkspacesTest>>testTileVisibleWindowsShouldPlaceThemInTwoColumnsAndRows windowManager tileVisibleWindows. self assert: (0@0 extent: 100@100) equals: workspace2 bounds. self assert: (100@0 extent: 100@100) equals: workspace1 bounds. self assert: (0@100 extent: 100@100) equals: browser2 bounds. self assert: (100@100 extent: 100@100) equals: browser1 bounds. or WMWithTwoBrowsersAndTwoWorkspacesTest>>testTileLastUsedWindowsShouldTileOnlyLastTwoWindowsUsed workspace1 activate. browser2 activate. workspace2 activate. browser1 activate. browser1 minimize. windowManager tileLastUsedWindows.. self assert: (0@0 extent: 200@100) equals: browser1 bounds. self assert: (0@100 extent: 200@100) equals: workspace2 bounds. Another example: WMWithAnOBBrowserAndDockingBarsTest>>setUp super setUp. obbrowser := self newWindowFor: #OBSystemBrowser. taskBar := TaskbarMorph new openInWorld: testWorld; adhereToEdge: #bottom. leftBar := DockingBarMorph new adhereToLeft; width: 25; openInWorld: testWorld. Note that in Core there's no OB, so tests are expected to fail: WMWithAnOBBrowserAndDockingBarsTest>>expectedFailures ^ (Smalltalk at: #OBSystemBrowser ifAbsent: nil) ifNil: [self class testSelectors] ifNotNil: [#()] Now if I want to test the content of a window / morph, I use methods you can find in Morph, protocol submorphs-accessing. For example I want to find the menu that have a given icon. Here's an helper: TWMBarWithTwoBrowsersAndTwoWorkspacesTest>>menuWithIcon: aForm ^ self twmBar dock findDeepSubmorphThat: [:aMorph| (aMorph isKindOf: MenuItemMorph) and:[aMorph icon = aForm]] ifAbsent: [nil]. so I can test that the Other Tools menu in TWMBar have an entry named "New TestRunner": TWMBarWithTwoBrowsersAndTwoWorkspacesTest>>testOtherToolsMenuShouldContainsEntryForOpeningSUnitTestRunner self assertMenuWithIcon: twmBar otherToolsMenuIcon containsItemNamed: 'New TestRunner'. I can also check that clicking on a given entry do the action I want. For example, if I have a Browser browsing class String, there should be an entry in the Browsers menu. If minimized, when I click on it, the Browser should be placed on top and expanded: TWMBarWithTwoBrowsersAndTwoWorkspacesTest>>testClickOnBrowserTwoMenuItemWhenMinimizedShouldActivateIt |menuItem| browser2 minimize. workspace1 activate. menuItem := self itemNamed: 'String' fromMenuWithIcon: twmBar browsersMenuIcon. menuItem target perform: (menuItem selector). self assert: browser2 equals: SystemWindow topWindow. self assert: (windowManager visibleWindows includes: browser2). Finally, if you use some settings, it's better to backup them in the setUp and restore them in the tearDown, so you keep your Settings while testing them :) WMWithTwoBrowsersAndTwoWorkspacesTest >>setUp ...... backupNumberOfWindowsSetting := TilingWM numberOfLastUsedWindows. TilingWM numberOfLastUsedWindows: 2. WMWithTwoBrowsersAndTwoWorkspacesTest>>tearDown ..... TilingWM numberOfLastUsedWindows: backupNumberOfWindowsSetting. Ideas welcome :) 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 > > Doru > > > >> Laurent > >> > >> > >> > >> On Jun 7, 2011, at 10:33 PM, laurent laffont wrote: > >> > >>> With the help of my local developper group, I've integrated Sean > DeNigris work and now Tiling Window Manager supports Multiple Worlds. > >>> > >>> Here's a screencast that shows the progress. (My English seems a little > less catastrophic than the previous screencast :) > >>> > >>> http://magaloma.blogspot.com/2011/06/multiple-worlds-for-twm.html > >>> > >>> > >>> Cheers, > >>> > >>> Laurent Laffont - @lolgzs > >>> > >>> Pharo Smalltalk Screencasts: http://www.pharocasts.com/ > >>> Blog: http://magaloma.blogspot.com/ > >>> Developer group: http://cara74.seasidehosting.st > >>> > >> > >> > >> > > > > > > -- > www.tudorgirba.com > > "Yesterday is a fact. > Tomorrow is a possibility. > Today is a challenge." > > > > >

1 0