February 2009 - Moose-dev

Media materials for Moose
by Simon Denier 11 Mar '09

11 Mar '09

Hi list I am looking for any media material (picture, screenshots of tools, slideshow, videos, podcasts) related to Moose. Some could be put on the website for the community, and I think they should be put in a file repository to be shared under a CC licence. Right now I am working on a poster for Moose and that's why I need such materials. -- Simon

2 1

Famix3 issue list
by Simon Denier 27 Feb '09

27 Feb '09

Dear list, After a fruitful meeting with Doru at Bern, here is the current list of issue to be resolved in the reference implementation of Famix3 (SqMoose), just so you know. Feel free to ask 38 Assigned [FAME] Handling of unknown attributes when importing MSE files (mismatched models) 39 Started Implement a Visitor for Famix3 40 Closed FAMIX3: review method organization to clean Famix-Core 41 New [MOOSE] port SCG algorithms to SqMoose 42 New [FAME] Add annotation/comment support in Fame metametamodel 43 New [FAME] add helper methods for multivalued attributes 44 New [FAMIX] Compability visitor for Famix2/famix3 45 Assigned [FAMIX3] NamedEntity#packagedIn should be derived and renamed to parentPackage 46 Assigned [FAMIX3] SmalltalkImporter: Classes/Types should be registered in ContainerEntity#types and not in Package#childNamedEntities 47 Assigned Smalltalk importer: check class registration 48 Assigned Smalltalk importer: fix stub 49 Assigned [FAMIX3] Smalltalk Importer: implement candidate invocations 50 Assigned [FAMIX3] Check attribute declarations are correct and synchronized 51 Assigned [FAMIX3] Remove generic setters from the model (from: to: belongsTo:) 52 Assigned Check current usage of NamedEntity#modifiers attribute 53 Assigned [FAMIX3] round-trip extraction/generation test with Fame -- Simon

2 1

Glamour browser
by Alexandre Bergel 26 Feb '09

26 Feb '09

With such a name, I couldn't resist to try. I downloaded Glamour and all the test pass (except one expected failure). What next? How I can open one? I skimmed the test, but I do not know where to start. Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

5 7

Re: [Pharo-project] Which Mondrian ?
by Alexandre Bergel 25 Feb '09

25 Feb '09

> I've loaded the mondrian form squeaksource and I'm trying it. I find > it quite different for the one used in OB. Yes, but very close to the VW version. We have a number of scripts and visualization made in VW that we would like to reuse. Moreover, in my opinion, the new version is simpler to use (and probably faster to render). The incantation to install the latest version of Mondrian in Pharo: -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= ScriptLoader new installer ss project: 'Mondrian'; install: 'MondrianLoader'. ScriptLoader perform: #loadMondrian -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= > I'll continue to try it as I'm reading Micheal Master thesis. I'm > trying on a simple graph first with tasks and node, but also I'd like > a special kinf of shape. Alex, I didn't found the way to compose > shape. Is it possible already ? Yes, you can compose them using MOFormsBuilder. There are numerous example in MOReadme, MONewFormsBuilderTest, and MOFormsBuilderTest (but some of them go red). An example that you can run with: MOReadme new flagExample -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= | builder view | builder := MOFormsBuilder new. "Assuming you want a grid made up of 3 columns and 2 rows" builder column; fill; column; fill; center; column; fill. builder row; fill; row; fill; row; fill; "For the German flag" row; center; "The EU text" row; pref. "For the French flag" "We draw a German flag" builder x: 1 y: 1 w: 3 add: (MORectangleShape new width: 30; height: 7; fillColor: Color black). builder x: 1 y: 2 w: 3 add: (MORectangleShape new width: 30; height: 7; fillColor: Color red). builder x: 1 y: 3 w: 3 add: (MORectangleShape new width: 30; height: 7; fillColor: Color yellow). builder x: 1 y: 4 w: 3 add: (MOLabelShape new text: [:aNumber |'EU ', aNumber printString]). "We draw a French flag" builder x: 1 y: 5 add: (MORectangleShape new width: 10; height: 20; fillColor: Color blue). builder x: 2 y: 5 add: (MORectangleShape new width: 10; height: 20; fillColor: Color white). builder x: 3 y: 5 add: (MORectangleShape new width: 10; height: 20; fillColor: Color red). view := MOViewRenderer new. view popupText: [:aNumber | 'my value is ', aNumber printString ]. view draggable. "Let's make nodes draggable, yeah!" view nodeShape: (builder shape). view nodes: (1 to: 27). view edges: (1 to: 27) from: #yourself to: [:each | each * each]. view layout: (MOCircleLayout new). view open -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= > I also wonder if we could imagine a SVGRenderer to use in conjunction > with seaside :) ? What do you think ? Yeah, that would be nice! > ps: David, concerning the graph button bar in OB, I'll suggest showing > it only in the comment view. It depends... I seldom press the comment button, except if there is a comment in it, which is rare. Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

1 0

FMMultivalueLink
by Laval Jannik 25 Feb '09

25 Feb '09

Hi all, In FAMIXScopingEntity, the method initialize is: --- initialize super initialize. childNamespaces := FMMultivalueLink on: self opposite: #parentNamespace:. parentNamespace := nil. globalVariables := FMMultivalueLink on: self opposite: #parentNamespace:. --- The class FMMultivalueLink does not exist. Does anyone know what it is and how to fix it? Thanks Jannik

3 3

MSE Incompability between Fame and iPlasma?
by Simon Denier 24 Feb '09

24 Feb '09

Hi everybody I'm trying to use iPlasma to parse Java source files, export the model as MSE, then import them in SqMoose using Fame. I ran into a couple of problems when doing this, specificall the MSE format built with iPlasma is not well parsed by Fame. Here is a sample generated by iPlasma: Beginning of MSE file: (Moose.Model (entity A) According to the MSE spec, the file cant begin directly with a name like that, should be ((Moose.Model (entity (....)) Besides, I dont see the point of this header, it works fine without it. (FAMIX.Namespace (id: 6) (name 'java::lang') ) B) Fame infers accessor 'smalltalk-style', ie it will call #name: to set the name of the Namespace entity in the example above. It works in Famix3 but not in Famix2, where the setter is declared as #setName: . It's alright with the inference rule of Meta (as I far as I remember) but not with the new one. We should add those accessors in Famix2 for seamless access. (FAMIX.Class (id: 3) (name 'Object') (belongsTo (idref: 6)) (isAbstract false) (isInterface false) C) Fame does not recognize the idref: syntax, only the ref: syntax. The MSE spec is inconsistent about that: the grammar only defines the ref: syntax as Fame does, but the doc talks about both an IDREF command and a 'REF command for metamodels only'. D) Finally, the biggest hurdle is that we dont know directly with which version of Famix the file was created. For example, the FAMIX package in SqMoose matches Famix3, and Famix2 is defined in a FAMIX2 package. This leads to a mismatch when trying to load the above file which was generated with Famix2 (unless batch editing all FAMIX -> FAMIX2) My suggestion is that each Famix metamodel should explicitely declare its version in its Fame package name, either Famix2 or Famix3, like that: FAMIXClass class>>annotation <MSEClass: #Class super: 'FAMIX3.Type'> <package: #FAMIX3> ^self Does it sound ok? -- Simon

6 11

Moose in Squeak is getting fast :-)
by Alexandre Bergel 24 Feb '09

24 Feb '09

Time millisecondsToRun: [| view | view := MOViewRenderer new. view nodes: (1 to: 50) forEach: [:each | view nodes: (1 to: 10) forEach: [:ea | view nodes: (1 to: ea * each). view gridLayout]. view gridLayout]. view open] => 18955 in Pharo => 23566 in VW :-) And I am sure there are rooms for improvement (IdentityDictionary are used with a small number of elements, a lot of OrderedCollection are used, ...) Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

4 3

Reordering nodes in MORoot
by Alexandre Bergel 22 Feb '09

22 Feb '09

Hi Mondrian friends, While doing a drag and drop, I am reordering the nodes in a MORoot: the dragged node is put at the end of the "root nodes" to make it the last to dislpay it on the screen. I am just wondering whether this will not interfere with other behavior. Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

2 2

Properties in SqMoose
by Alexandre Bergel 22 Feb '09

22 Feb '09

Dear SqMoose Crew, I have a question about properties. In FileImportTest>>testFiles, there are accesses to properties, e.g., self assert: (file propertyNamed: #TLOT) = 1. self assert: (file propertyNamed: #NOCh) = 0. (file propertyNamed: #TLOT) returns nil. It seems that the computing is not done. However, if I manually trigger the computation, getting the property returns the expected value as in: self assert: (file totalNumberOfLinesOfText = 1). "If I remove this line, the next one return nil" self assert: (file propertyNamed: #TLOT) = 1 Any idea ? It seems that the meta stuff is not able to retrieve the method that contains the property. Any idea why? Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

4 6

multi-line label
by Alexandre Bergel 21 Feb '09

21 Feb '09

Dear List, I was wondering why a UML diagram cannot be obtained through: -=-=-=-=-=-=-=-=-=-=-=-= | builder view | builder := MOFormsBuilder new. builder column; fill. builder row; fill; row; fill. builder x: 1 y: 1 add: (MORectangleShape new). builder x: 1 y: 1 add: (MOLabelShape new text: [:cls | cls name]). builder x: 1 y: 2 add: (MORectangleShape new). builder x: 1 y: 2 add: (MOLabelShape new text: [:cls | cls methodDict keys inject: '' into: [:sum :el | sum, el, String cr]]). view := MOViewRenderer new. view nodeShape: builder shape. view draggable. view popupText. view nodes: {MONExplicitIdentityNormalizer}. view layout: MOTreeLayout new. view open -=-=-=-=-=-=-=-=-=-=-=-= This is not polished, but having multi-lines label seems to be interesting... Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

3 2

Nice tool with unfortunate naming
by Thomas Schrader 21 Feb '09

21 Feb '09

Hi all, see: http://www.rosuda.org/Mondrian/ Not exactly as powerful but a clash of names nonetheless ... Regards Thomas -- mailto thomas j schrader at web de ____________________________________________________________________ Psssst! Schon vom neuen WEB.DE MultiMessenger gehört? Der kann`s mit allen: http://www.produkte.web.de/messenger/?did=3123

2 1

Is Moose-All really necessary?
by Alexandre Bergel 20 Feb '09

20 Feb '09

Dear SqMoose friends, i was wondering whether Moose-All is necessary since we have MooseLoader. The dependencies defined on Moose-All could be made explicit said in MooseLoader. Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

3 5

Time to load SqMoose
by Alexandre Bergel 19 Feb '09

19 Feb '09

Hi All, Am i the only one to notice the very long time taken by Moose to load? I do not understand how it could be so long... I have a "Writing definitions" for almost 10 minutes... Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

2 1

[VWMoose] Programmatically import a MSE file
by Simon Denier 18 Feb '09

18 Feb '09

Hi I'm trying to load a mse file into VW Moose without using the GUI. Below is my best shot right now: it doesnt fail but the created model is empty. I guess that "MooseImporter withActiveMetamodel" is too optimistic. Anyway, before I dive once again into the peculiarities of VW, what is the straightforward way to do that? I couldnt find the right API in Moose. model := MooseModel new name: (fileName asFilename tail removeSuffix: '.mse'). self halt. loader := MooseImporter withActiveMetamodel. loader stream: fileName asFilename readStream. result := loader asTask run. loader resolveDanglingReferences. model addEntity: result -- Simon

2 2

merge tool in VW :(
by stephane ducasse 17 Feb '09

17 Feb '09

Hi I wanted to merge the change made by veronicas and when I use the merge tool and apply resolution I get a DNU.... (I will not comment). So doru how do you do that normally? Stef

3 4

Re: [Pharo-project] SqMondrian
by Tudor Girba 14 Feb '09

14 Feb '09

Hi Alex, I guess this mail was supposed to go to moose-dev, too :). > I was wondering why is there a method originFor: in FormsBuilder. In > VW there is none. Why the builder should know about the bounds of the > element to display? I always though the shape was in charge of it. The method's name is wrong. The *For: methods take a Figure as an argument. In our case it should be originOf: because it computes the origin of a shape inside the context of the form. > Something else, both in Squeak and VW we have the following method: > -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= > FormsBuilder>>at: aX and: aY > "Utility method that makes other methods easier to read." > > ^self components > select: [:component | component gridX = aX and: [component gridY = > aY]] > -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= > > Is there a reason to not store the components in a matrix ? One reason > that I see, is that one could have a large compound contains only few > elements in it. But still, I find that strange. You could store it in a matrix. Cheers, Doru > Cheers, > Alexandre > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project -- www.tudorgirba.com "Problem solving should be concentrated on describing the problem in a way that is relevant for the solution."

2 1

Re: [Pharo-project] Implementing a DSM in SqMondrian
by Tudor Girba 14 Feb '09

14 Feb '09

Hi Alex, I guess this mail was supposed to go to moose-dev :). > I was wondering what the best strategy to implement a DSM will be. I > see two different strategies: > 1 - Building the matrix with a formsBuilder. Each shape contained in > the builder is therefore a cell of the DSM. > 2 - Having an MONode for each cell. > > I would favor option 2, but a special layout need to be defined right? In VW, you can already do that through "formsBuilder asLayout". I used this to build an evolution matrix that can be align nicely both vertically and horizontally. > Doru, you said once that Mondrian may not be adapted to build DSM. I > do not understand (or even feel) why. May you comment on this? You can, but it won't be optimal. The reason is that for any significantly sized matrix you will not want to represent each cell as an object, but rather treat the complete matrix as one smart object that optimizes the display. Cheers, Doru > Cheers, > Alexandre > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. -- www.tudorgirba.com "Beauty is where we see it."

2 1

[SqMoose] FMPragmaProcessor subclass: #MSEPragmaProcessor
by Simon Denier 13 Feb '09

13 Feb '09

I was looking at the pragma processor in SqMoose and noticed that there is a subclass of FMPragmaProcessor called MSEPragmaProcessor. Now this class does nothing except override one method from FMPragmaProcessor, which is almost a copy-paste of its overridden method. The only difference between the two methods is that one have an early abort condition while the other continues even if the condition is met. Now MSEPragmaProcessor is packaged in Moose-Finder (and is used by mooseDescription). I wonder what is the intent behind this class? -- Simon

2 1

Just a thought...
by Alexandre Bergel 12 Feb '09

12 Feb '09

Dear friends, I am now sitting in a room in which people are doing demos of their software. A team does some sensor wireless networks, and they have a fancy animation looping on a screen. Some of us are doing visualization (including me, even if I am a newbie), I think we should be able to produce some eye-catching animations. Maybe something can be done around code city. A .mov will be welcomed... I will work on something in the near future, as soon as I have time. Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

2 3

Layers in Mondrian?
by Alexandre Bergel 10 Feb '09

10 Feb '09

Dear List, The MOViewRenderer contains a list of layers (instances of MOLayer). Is there a particular use for layers? what was the idea behind? Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

3 4

[Mondrian] MORectangleShape>>mightHaveChildren
by Alexandre Bergel 09 Feb '09

09 Feb '09

Dear List, It appears that a rectangle shape in the Squeak version of Mondrian may or may not have children. Conceptually, when a rectangle cannot have children? I am trying to figure out whether this something we should keep. Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

2 2

History flow visualization
by Adrian DOZSA 04 Feb '09

04 Feb '09

IBM - History flow visualization (visualizing the editing history of Wikipedia pages) link: http://www.research.ibm.com/visual/projects/history_flow/index.htm Same concept could be applied for code source also. ------------------------------ dipl.ing. Adrian DOZSA Software Engineer mail: adi.dozsa(a)gmail.com web: http://adi.dozsa.googlepages.com/

3 2

Mondrian and events
by Alexandre Bergel 02 Feb '09

02 Feb '09

Dear list, I enhanced Mondrian to support announcement based event. Here an example: -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= | view | view := MOViewRenderer new. view subscribe: MOMouseEnter do: [:ann | (ann element notNil and: [(ann element isKindOf: MORoot) not]) ifTrue: [MOPopup show: ann element printString] ifFalse: [MOPopup remove] ]. view subscribe: MOMouseDown do: [:ann | ann inspect]. view nodeShape: (MORectangleShape new width: 30; height: 40). view nodes: (1 to: 5). view open -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Two effects: - putting the mouse over a box display a popup window. - clicking on a box open an inspector I propose an extension. I think this respects what was done in VW. But maybe one wants to have different event handler regarding the objects inserted. For example: -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= | view | view := MOViewRenderer new. view nodeShape: (MORectangleShape new width: 30; height: 40). view subscribe: MOMouseDown do: [:ann | ann inspect]. view nodes: (1 to: 5). view subscribe: MOMouseDown do: [:ann | Beeper beep]. view nodes: (5 to: 10). view open -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Nodes from 1 to 5 will open an inspector when clicked. Nodes from 1 to 10 will beep instead. Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

4 6

Mondrian and Squeak: tooltip
by Alexandre Bergel 01 Feb '09

01 Feb '09

Hi again, Tool tips are a kind of shape. I found that strange. In VW, a view answers to the message interaction, returning an event handler. I like this approach. Is there a reason why in Squeak it is so different? Cheers, Alexandre _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

2 1

Mondrian in Squeak
by Alexandre Bergel 01 Feb '09

01 Feb '09

Dear List, I've tried the following: -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= | view builder | builder := MOFormsBuilder new. builder column; left; pref. builder row; left; row; left; row; left; pref. builder x: 1 y: 1 add: (MORectangleShape new). builder x: 1 y: 2 add: (MOLabelShape new). view := MOViewRenderer new. view node: #blah using: builder shape. view open -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= A shape is simply added to the builder. If you execute the following code, you will not get what you expect (a text 'blah' below a rectangle). If you manage if get the hand back (Cmd-.), then you will see that the message width is sent to an instance of MOLabelShape. This method does not exist, thus raising an exception. I have several questions: - How can a shape knows the width of the string it is supposed to display, if the shape does not know about what to display? The "what to display" is provided in the display:on: method. - In VW, the LabelShape>>width returns a widthBlock. This variable does not exist in Squeak. Is there a reason why? Is there something very different regarding labels from the Squeak and VisualWorks version? A more general question, tt seems that the strategy to compute width and height of label in VW differs in Squeak. Any reason for this? Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

1 0

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

Moose-dev February 2009