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

[Pharo-project] Updated Configuration packages for PharoMorphicView

by S Krish

Details are posted in: http://skrishnamachari.wordpress.com/2012/02/09/pharo-morphic-view-framewor… hopefully cleaner and easier for anyone to load and check out.. PharoMorphicView-Core-App : is a basic package and PharoMorphicView is the full pack. Trying to now wade through the Specs / Presenty / existing capabilities which I can merge with but overall very interesting first time run up with Morphic Experimental Run, but quite usable overall.... plugs in hand in glove with existing primitive morphs. Easy too to get some newbies started with Morphic UIs.. PS: Once comment was nice to hear.. Few new trainees found working in Pharo easier than starting up with VisualAge.. !..

Feb. 9, 2012

Re: [Pharo-project] Feedback on Pharo 1.4 Debugger

by Eliot Miranda

Has anybody tried to partially highlight the other shared activations? i.e. in the below if you selected any of the updateReport activations (home or blocks within it) the other activations could be highlighted in a lighter colour. On Thu, Feb 9, 2012 at 9:37 AM, Ben Coman <btc(a)openinworld.com> wrote: > The debugger in Pharo 1.4 is looking very nice - cleaner and easier to > follow. I wonder if the following might be an additional improvement... > > If you consider a class has a method subpart which has a block subpart, > then the "natural flow" of these would continue from left to right. > Referring to [1] - having the "Block" text on the left breaks this sense > of "flow". What I've drawn in [2] feels more natural to me. > > Possible advantages: > * It is easier to see the five SystemReporter items line up. > * Balances brevity with giving a bit of context to distinguish nested > blocks. > > [1] NewDebuggerStackView(1).png (attached) > [2] NewDebuggerStackView(2).png (attached) > > cheers, -ben > -- best, Eliot

Feb. 9, 2012

[Pharo-project] Feedback on Pharo 1.4 Debugger

by Ben Coman

The debugger in Pharo 1.4 is looking very nice - cleaner and easier to follow. I wonder if the following might be an additional improvement... If you consider a class has a method subpart which has a block subpart, then the "natural flow" of these would continue from left to right. Referring to [1] - having the "Block" text on the left breaks this sense of "flow". What I've drawn in [2] feels more natural to me. Possible advantages: * It is easier to see the five SystemReporter items line up. * Balances brevity with giving a bit of context to distinguish nested blocks. [1] NewDebuggerStackView(1).png (attached) [2] NewDebuggerStackView(2).png (attached) cheers, -ben

Feb. 9, 2012

[Pharo-project] Smalltalk on Bacon conf?

by Janko MivÅ¡ek

Hi guys, Maybe this two day conference "by developers for developers" is another opportunity to show strengths of Smalltalk? Bacon http://devslovebacon.com/ BACON is a two-day, two-track technology conference on things developers love. 20-21 April, 2012 London. Thirty-two tasty sessions on topics including web development, machine learning, astronomy, and electronic music. Call for speakers Bacon is a conference by developers for developers. We are looking for a wide range of talks that focus on things developers love, whether digital or analogue. Are you a developer by day and electronic music superstar by night? Have you been working on some super interesting code and want to share your findings? Each speaker will have half an hour to deliver their talk and to gain the eternal admiration and adoration of their peers. Best regards Janko -- Janko MivÅ¡ek Aida/Web Smalltalk Web Application Server http://www.aidaweb.si

Feb. 9, 2012

Re: [Pharo-project] FSFilesystemTest>>testReferenceTo fails on Windows

by Ben Coman

Ben Coman wrote: > >> >> C:\TestC1> E: E:\> mkdir C:TestC2 >> > note this should be two lines > > C:\TestC1> E: E:\> mkdir C:TestC2 > > > damn try again. sorry for the noise. C:\TestC1> E: <newline> E:\> mkdir C:TestC2

Feb. 9, 2012

Re: [Pharo-project] FSFilesystemTest>>testReferenceTo fails on Windows

by Ben Coman

> > C:\TestC1> E: E:\> mkdir C:TestC2 > note this should be two lines C:\TestC1> E: E:\> mkdir C:TestC2

Feb. 9, 2012

[Pharo-project] [JOB, Academia] Postdoc Chile/France

by Marcus Denker

Postdoc position in software engineering at University of Chile in cooperation with INRIA ===================================================================== A post-doctoral research position is available with the Pleiad Lab (University of Chile) and the RMoD Group (INRIA Lille Nord Europe). The position is targeted for research on software maintenance and software visualization. Validation of research hypotheses will be carried out using the Pharo programming language. Experience with it, or any other Smalltalk dialect, is a plus (but not necessary). The postdoctoral researcher will be expected to develop independent research initiatives within the scope of the position, design and contribute to the research experiments conducted by Pleiad and RMoD. There is no teaching duty associated with this position. Both Pleiad and RMoD offer a creative, productive and multicultural staff and students, with many collaboration opportunities. Pleiad and RMoD have an extensive expertise and an advanced infrastructure related to software analysis, data mining, testing, meta-modeling and language semantics. The position is funded by INRIA for a period of 12 or 18 months. In order to foster the collaboration between France and Chili, the candidate will have to share his time between Lille, France and Santiago, Chile. The candidate must hold a PhD in computer science, finished max 2 years ago. Contact Persons: Alexandre Bergel and StÃ©phane Ducasse More about... - PLEIAD: http://pleiad.dcc.uchile.cl - RMOD: http://rmod.lille.inria.fr - Pharo: http://www.pharo-project.org - Moose: http://www.moosetechnology.org (permanent link of this call: http://rmod.lille.inria.fr/web/pier/blog/2012-02-09) -- Marcus Denker -- http://marcusdenker.de

Feb. 9, 2012

[Pharo-project] [update 1.4] #14323

by Marcus Denker

14323 ----- Issue 5239: Add explicit entries for squeaksource / gemstone repos http://code.google.com/p/pharo/issues/detail?id=5239 Issue 5237: as: LinkedList does not behave as expected http://code.google.com/p/pharo/issues/detail?id=5237 Issue 5266: Behavior >> methodDict should answer empty method dict rather than nil http://code.google.com/p/pharo/issues/detail?id=5266 -- Marcus Denker -- http://marcusdenker.de

Feb. 9, 2012

[Pharo-project] FSFilesystemTest>>testReferenceTo fails on Windows

by Ben Coman

I have loaded ConfigurationOfFilesystem v2.0.3 into Pharo-1.3-13315 on Windows 7. All tests are green except FSFilesystemTest>>testReferenceTo which hard codes the unix root folder... self assert: (filesystem referenceTo: '/') isRoot. However, guessing at the semantics of isRoot, simply changing this to... self assert: (filesystem referenceTo: 'C:\') isRoot. is not sufficient since... FSAbsolutePath>>isRoot ^ self size = 0 has size = 1. Looking into this further, it appears that the test for absolute path is incorrect in FSWindowStore>>pathFromString: aString. It is not the drive letter and colon that make it an absolute path, but whether the third character is a slash as well. Consider this sequence from a new command window ... C:\> mkdir C:\TestC1 C:\> cd TestC1 C:\TestC1> E: E:\> mkdir C:TestC2 E:\> C: C:\testC1> cd \testC1\testC2 C:\testC1\testC2> Definitively... [1] "Section 'Fully Qualified vs. Relative Paths' >> If a file name begins with only a disk designator but not the backslash after the colon, it is interpreted as a relative path to the current directory on the drive with the specified letter. " [1] http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247%28v=vs.85%… Without knowing the design philosophy or architecture of FS I'll just wildly speculate on on implementation.... by FSDiskStore having an instance variable "volume." For Unix this would remain empty. For Windows this would be 'C:' or 'D:' (without the slash). Then once this volume is stripped from the front of the path, the super method could be used, and (^ self size = 0) in FSAbsolutePath>>isRoot would be true. For example... FSWindowStore>>pathFromString: aString | class | class := ((aString at: 1) isLetter and: [(aString at: 2) = $:]) ifTrue: [ volume := (aString at: 1) asString , ":" . ^ super pathFromString: (aString copyFrom: 3 to: (aString size)). ] ^ super pathFromString: aString. FSStore>>pathFromString peek = $/ would also need to be replaced by peek = self delimiter. OR... perhaps... FSDiskStore>>pathFromString: aString self volumeFromString: aString. super pathFromString: ( aString copyFrom: (self volume size) + 1 to: (aString size) ). FSDiskStore>>volume volume ifNil: [ volume := '' ]. FSDiskStore>>volumeFromString ^self. FSWindowStore>>volumeFromString: aString ((aString at: 1) isLetter and: [(aString at: 2) = $:]) ifTrue: [ volume := (aString at: 1) asString , ":" . ] "also handle UNC volumes here" Also are UNC paths under consideration? Here the volume '\\server\share' might be extracted from the front of the path and isRoot would be True at this location. From [1] ... "The "\\?\" prefix can also be used with paths constructed according to the universal naming convention (UNC). To specify such a path using UNC, use the "\\?\UNC\" prefix. For example, "\\?\UNC\server\share", where "server" is the name of the computer and "share" is the name of the shared folder. These _ prefixes are not used as part of the path itself. _ " Anyway, that is as far as I can take it. Obviously lots of implications I haven't considered. Hope this helps. cheers, -ben

Feb. 9, 2012

Re: [Pharo-project] Enhancement: Browsing SLICE dependencies with tree view in Monticello Repository Inspector

by Mariano Martinez Peck

Hi Ben. I think this was fixed in Pharo 1.4. Can you try and if true, close the issue? Thanks On Thu, Feb 9, 2012 at 1:47 AM, Ben Coman <btc(a)openinworld.com> wrote: > I originally thought something was wrong when I could not browse the code > of a SLICE using Monticello. After hunting around I found [1] saying this > is expected behaviour. However I then suggest the following enhancement: > > Have the right-pane of the Monticello Repositoy Inspector be a tree view, > such that dependencies are the children which can then be <Browsed> > together from the one place, rather than having to chase them manually. > This seems more intuitive usage, especially for those not familiar with > what slices are. (I had to hunt around for 5 minutes to find out) > What do you think? > > cheers, -ben > > [1] http://code.google.com/p/**pharo/issues/detail?id=1606<http://code.google.com/p/pharo/issues/detail?id=1606> > says " <Browse> in MC only browses the contents of the *selected* package > - not of dependent packages. Therefore the slice package always looks > empty, because it actually is empty." > > > > > > -- Mariano http://marianopeck.wordpress.com

Feb. 9, 2012