August 2014 - Pharo-users - lists.pharo.org

Trying to understand the "display: children:" messages of Moose browsers.

by Offray Vladimir Luna CÃ¡rdenas

Hi, I'm trying to understand the workings of the "display: children:" message of Moose Browsers. Based on the Help Browser example, I have write a node object which contains a header, a body and a ordered collection of Children, wich are other nodes. Now I would like to use moose browsers to navigate a tree made of such nodes. For that I am modifying the "packagesIn:" tree of the PBE2 book example in the Moose chapter. I have something like this: -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= UbakyeBrowser>>packagesIn: constructor | defaultTree | defaultTree := UbakyeNode new. defaultTree becomeDefaultTree. constructor tree display: [:treeHeaders | treeHeaders := defaultTree headers]; children: [:currentNode | self headersOf: currentNode ]; format: #asString -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= The display part is working just because I'm sending a collection of node names (headers), but I don't know how to relate the current selected node with the things that should be shown by the "children:" message. Any help on this is appreciated. By the way, I would like to share my code and point to it, instead of sharing just cut and pasted snippets, because may be I'm doing tons of errors in a lot of places. I have already asked in a separate thread about how to sync local cache and SmalltalkHub repos. Cheers, Offray

Aug. 2, 2014

Re: [Pharo-users] Syncing local cache and SmalltalkHub

by Sven Van Caekenberghe

Hi Offray, Once you create your StHub repository and define it in your image, you can just copy versions (mcz files) from your package cache to the new repository using the Copy button in the Repository browser (which you get by selecting your local package cache in the Monticello browser and clicking the Open button). HTH, Sven On 02 Aug 2014, at 21:25, Offray Vladimir Luna CÃ¡rdenas <offray(a)riseup.net> wrote: > Hi, > > I have my outliner buggy mockup on local cache, and I would like to share the code and point to it online using SmalltalkHub for my questions. I have followed the chapter on Source code management with Monticello y PBE2, and it says that you can have your code on SmalltalkHub, but says nothing about how to sync from local cache to your SmalltalkHub account. Can some one could point me the way to do it? > > Cheers, > > Offray >

Aug. 2, 2014

MooseTechnology failing again. Any way to download Moose Book?

by Offray Vladimir Luna CÃ¡rdenas

Hi, These days, MooseTechnology.org has being failing continuously and I can not read any documentation on Moose browsers, except for the one in PBE2. There is any way to download this book to see it off-line, hopefully in pdf format? Thanks, Offray

Aug. 2, 2014

Syncing local cache and SmalltalkHub

by Offray Vladimir Luna CÃ¡rdenas

Hi, I have my outliner buggy mockup on local cache, and I would like to share the code and point to it online using SmalltalkHub for my questions. I have followed the chapter on Source code management with Monticello y PBE2, and it says that you can have your code on SmalltalkHub, but says nothing about how to sync from local cache to your SmalltalkHub account. Can some one could point me the way to do it? Cheers, Offray

Aug. 2, 2014

Re: [Pharo-users] Roassal2 orthogonal lines, attach points and self-lines.

by Alexandre Bergel

Good point. I will work on this today Alexandre > Le 02-08-2014 Ã 5:29, Peter UhnÃ¡k <i.uhnak(a)gmail.com> a Ã©crit : > > Hi Alex, > > is there a (simple) way to reset element's shape? > When I add a new control point to RTMultiLine I need the trachelShape to be updated accordingly, however it is created only once and I don't see any way how to reset it apart from completely removing the element from canvas and recreating it. > > I tried subclassing RTEdge (and altering RTElement>>addConnectedEdge: assert to isKindOf:) and setting trachelShape to nil, however that wasn't enough anyway since the shape is apparently also stored also elsewhere (TRCanvas I would assume). > > So do I have to remove it and re-add it or am I missing something? > > Also, is there some document(atiton) explaining how all the things interact together (Roassal, Trachel, updates, redraws, ...) or is source code the only thing at the moment? > > Thanks, > Peter > > >> On Thu, Jul 31, 2014 at 3:22 PM, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: >> > I am familiar with RTBezierLine (I mentioned it in the original question), but what I was wondering about was whether controllingElements: should be added to RTMultiLine (so it will be handled internally just like in BezierLine), or leave it as is and let the user handle it from outside - as demonstrated by the example I posted. >> >> Users of the API will decide :-) >> The way you are interacting with us is the right way. Keep asking questions and requesting for new features! >> >> Alexandre >> >> > >> > >> > On Thu, Jul 31, 2014 at 8:24 AM, stepharo <stepharo(a)free.fr> wrote: >> > this is cool :) >> > >> > >> > On 30/7/14 20:12, Alexandre Bergel wrote: >> >> Hi Peter! >> >> >> >> Iâve just added the orthogonal vertical and horizontal lines in Roassal2. Here are some screenshots. >> >> >> >> <Mail Attachment.png> >> >> <Mail Attachment.png> >> >> >> >> These last script has been made with: >> >> -=-=-=-=-=-=-=-=-=-=-=-= >> >> | v | >> >> v := RTView new. >> >> >> >> v addAll: ((RTEllipse new size: 20; color: (Color red alpha: 0.3)) elementsOn: (1 to: 20)). >> >> >> >> shape := RTMultiLine new. >> >> shape orthoVertical. >> >> shape color: (Color blue alpha: 0.3). >> >> >> >> RTEdge >> >> buildEdgesFromObjects: (1 to: 20) from: [ :n | n // 3 ] to: #yourself using: shape inView: v. >> >> >> >> v elements @ RTDraggable @ RTPopup. >> >> >> >> RTTreeLayout on: v elements. >> >> v open >> >> -=-=-=-=-=-=-=-=-=-=-=-= >> >> >> >> Cheers, >> >> Alexandre >> >> -- >> >> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >> >> Alexandre Bergel http://www.bergel.eu >> >> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >> >> >> >> >> >> >> >> On Jul 26, 2014, at 3:57 PM, Peter UhnÃ¡k <i.uhnak(a)gmail.com> wrote: >> >> >> >>> Hi, >> >>> >> >>> is there any support for orthogonal lines in Roassal2? I've seen something similar (ROOrthoVerticalLineShape), however nothing of that sort in Roassal2. I would imagine it to behave similarly to RTBezierLine - having controllingElements to specify the corners. (While ROOrtholines did it automatically I was looking for something more manual, so it can go in both directions - thus the controllingElements). >> >>> >> >>> Second thing I wasn't able to figure out is whether it is possible to create line from element to itself. All *AttachPoint classes aren't really able to cope with that. The best result I got was from Vertical/Horizontal AttachPoints (combined with BezierLine), however I would like to be able to specify which side it would start on and end on. Ideally not just middle of sides, but arbitrary part of the shape (I imagine that might be hard for non-rectangular shapes). >> >>> >> >>> Thanks, >> >>> Peter >> >> >> > >> > >> >> -- >> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >> Alexandre Bergel http://www.bergel.eu >> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >

Aug. 2, 2014

Re: [Pharo-users] Finalization question

by Ben Coman

Udo Schneider wrote: > All, > > I just stumbled over something which thought was pretty easy - turned > out it isn't as easy as I thought or I'm simply _______ (fill in > expression of choice like "too dumb", "blind", etc.) > > I have an object which starts a background process to do it's job. > This object holds a reference to the running proccess in an instance > variable. > > My goal is to #terminate that proccess once the object isn't needed > anymore (before GC). > > Using WeakRegistry for the object itself doesn't work. E.g. > > MyObject>>#initialize > super initialize. > WeakRegistry default register: self. > myProcess := self startProcess. > > MyObject>>#finalize > myProcess terminate. "Doesn't work. myProcess is already nil" > > I think this is due to all the ivars of the object already being nil > when #finalize is begin called. As far as I read the code this is due > to Object>>#executor returning a shallow copy of the object and thus > being by intention ... > > So how to proceed? Did I miss something obvious? Is overwriting > #executor the way to go (in terms of "does it work" and "doesn't it > break something")? > > Best Regards, > > Udo > > > > > I don't know much about finalization of Weak constructs, but maybe an alternative. I assume you have something like.. startProcess [ "do stuff" ] fork. so how about... startProcess [ [ "do stuff" myProcess = nil. ] whileFalse ] fork. cheers -ben

Aug. 2, 2014

Finalization question

by Udo Schneider

All, I just stumbled over something which thought was pretty easy - turned out it isn't as easy as I thought or I'm simply _______ (fill in expression of choice like "too dumb", "blind", etc.) I have an object which starts a background process to do it's job. This object holds a reference to the running proccess in an instance variable. My goal is to #terminate that proccess once the object isn't needed anymore (before GC). Using WeakRegistry for the object itself doesn't work. E.g. MyObject>>#initialize super initialize. WeakRegistry default register: self. myProcess := self startProcess. MyObject>>#finalize myProcess terminate. "Doesn't work. myProcess is already nil" I think this is due to all the ivars of the object already being nil when #finalize is begin called. As far as I read the code this is due to Object>>#executor returning a shallow copy of the object and thus being by intention ... So how to proceed? Did I miss something obvious? Is overwriting #executor the way to go (in terms of "does it work" and "doesn't it break something")? Best Regards, Udo

Aug. 2, 2014

Re: [Pharo-users] Roassal2 orthogonal lines, attach points and self-lines.

by Peter UhnÃ¡k

Hi Alex, is there a (simple) way to reset element's shape? When I add a new control point to RTMultiLine I need the trachelShape to be updated accordingly, however it is created only once and I don't see any way how to reset it apart from completely removing the element from canvas and recreating it. I tried subclassing RTEdge (and altering RTElement>>addConnectedEdge: assert to isKindOf:) and setting trachelShape to nil, however that wasn't enough anyway since the shape is apparently also stored also elsewhere (TRCanvas I would assume). So do I have to remove it and re-add it or am I missing something? Also, is there some document(atiton) explaining how all the things interact together (Roassal, Trachel, updates, redraws, ...) or is source code the only thing at the moment? Thanks, Peter On Thu, Jul 31, 2014 at 3:22 PM, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > > I am familiar with RTBezierLine (I mentioned it in the original > question), but what I was wondering about was whether controllingElements: > should be added to RTMultiLine (so it will be handled internally just like > in BezierLine), or leave it as is and let the user handle it from outside - > as demonstrated by the example I posted. > > Users of the API will decide :-) > The way you are interacting with us is the right way. Keep asking > questions and requesting for new features! > > Alexandre > > > > > > > On Thu, Jul 31, 2014 at 8:24 AM, stepharo <stepharo(a)free.fr> wrote: > > this is cool :) > > > > > > On 30/7/14 20:12, Alexandre Bergel wrote: > >> Hi Peter! > >> > >> Iâve just added the orthogonal vertical and horizontal lines in > Roassal2. Here are some screenshots. > >> > >> <Mail Attachment.png> > >> <Mail Attachment.png> > >> > >> These last script has been made with: > >> -=-=-=-=-=-=-=-=-=-=-=-= > >> | v | > >> v := RTView new. > >> > >> v addAll: ((RTEllipse new size: 20; color: (Color red alpha: 0.3)) > elementsOn: (1 to: 20)). > >> > >> shape := RTMultiLine new. > >> shape orthoVertical. > >> shape color: (Color blue alpha: 0.3). > >> > >> RTEdge > >> buildEdgesFromObjects: (1 to: 20) from: [ :n | n // 3 ] to: #yourself > using: shape inView: v. > >> > >> v elements @ RTDraggable @ RTPopup. > >> > >> RTTreeLayout on: v elements. > >> v open > >> -=-=-=-=-=-=-=-=-=-=-=-= > >> > >> Cheers, > >> Alexandre > >> -- > >> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > >> Alexandre Bergel http://www.bergel.eu > >> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > >> > >> > >> > >> On Jul 26, 2014, at 3:57 PM, Peter UhnÃ¡k <i.uhnak(a)gmail.com> wrote: > >> > >>> Hi, > >>> > >>> is there any support for orthogonal lines in Roassal2? I've seen > something similar (ROOrthoVerticalLineShape), however nothing of that sort > in Roassal2. I would imagine it to behave similarly to RTBezierLine - > having controllingElements to specify the corners. (While ROOrtholines did > it automatically I was looking for something more manual, so it can go in > both directions - thus the controllingElements). > >>> > >>> Second thing I wasn't able to figure out is whether it is possible to > create line from element to itself. All *AttachPoint classes aren't really > able to cope with that. The best result I got was from Vertical/Horizontal > AttachPoints (combined with BezierLine), however I would like to be able to > specify which side it would start on and end on. Ideally not just middle of > sides, but arbitrary part of the shape (I imagine that might be hard for > non-rectangular shapes). > >>> > >>> Thanks, > >>> Peter > >> > > > > > > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > >

Aug. 2, 2014

Re: [Pharo-users] Data Science with Pharo?

by volkert＠nivoba.de

Am 01.08.2014 um 17:38 schrieb Thierry Goubier: > Le 31/07/2014 15:29, Alexandre Bergel a Ã©crit : >>> I am successful using Pharo + Roassal/Mondrian in analyzing and >>> visualizing IT architecture models. And it is real fun. :-) >> Glad to ear this :-) >> >> Yes, the vision I see behind Pharo, Moose, Roassal is to provide a >> solid alternative to R. We are not there yet, but we will get there. > Cool. I'm interested by that objective :) > > Thierry > Me too BW, Volkert

Aug. 1, 2014

Re: [Pharo-users] Roassal2 RTDraggable grid and synchronized movement

by Alexandre Bergel

This is an excellent start. Do you need the grid? If yes, then I can add this to Roassal. It is easy to define an interaction that define a grid. I have added event to signal the beginning and ending of a dragging-and-dropping. Here is an example: -=-=-=-=-=-=-=-= | v e | v := RTView new. e := (RTEllipse new size: 30) element. e @ RTDraggable. v add: e. e when: TRMouseDragStart do: [ :evt | Transcript show: evt printString; cr. evt element trachelShape color: Color red. v signalUpdate ]. e when: TRMouseDragEnd do: [ :evt | Transcript show: evt printString; cr ]. v open -=-=-=-=-=-=-=-= Let me know whether this helps. Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. On Aug 1, 2014, at 2:36 PM, Peter UhnÃ¡k <i.uhnak(a)gmail.com> wrote: > Hi Alex, > > While the "snap to grid" might be complicated I thought of simpler solution that might be achievable purely at Roassal level without the need to change Trachel classes. You can look at Umlet, since I am trying to copy it's grid behavior here. > Basically you don't move objects with mouse immediately but only after fixed increments. However I've encountered a problem with TRMouseDragging>>step, since I would probably need to continuously sum the individual steps and then reset them when dragging has started. > > Here is a _very_ crude prototype: > --------------------- > "i subclassed RTDraggable and changed initializeElement: method as follows" > initializeElement: element > | sum gridSize | > gridSize := 20. > sum := 0 @ 0. > element > when: TRMouseDragging > do: [ :e | > | d t | > t := RTGroup withAll: groupToDrag. > (t includes: element) > ifFalse: [ t add: element ]. > d := element view canvas camera distanceFromPixelToSpace: e step. > sum := sum + d. > "move element only in increments of gridSize" > d := gridSize * ((sum x / gridSize) truncated @ (sum y / gridSize) truncated). > sum := sum - d. > t translateBy: d. > e signalUpdate ] > --------------------- > > and here is a demo utilizing it > > ----------------------------- > | v e1 e2 line gridSize b1 b2 b3 b4 edge | > v := RTView new. > > "draw grid" > gridSize := 20. > > (-20 to: 20) do: [ :i | > b1 := RTBox element. > b2 := RTBox element. > b3 := RTBox element. > b4 := RTBox element. > b1 translateTo: (i * gridSize) @ -1000. "top side" > b2 translateTo: (i * gridSize) @ 1000. "bottom side" > b3 translateTo: -1000 @ (i * gridSize). "left side" > b4 translateTo: 1000 @ (i * gridSize). "right side" > "vertical line" > edge := RTLine edgeFrom: b1 to: b2. > v add: edge. > "horizontal line" > edge := RTLine edgeFrom: b3 to: b4. > v add: edge. > ]. > > e1 := (RTEllipse new size: 30; color: Color magenta) element. > e1 translateTo: -20 @ 0. > e1 @ RGGridDraggable. > > e2 := (RTEllipse new size: 30; color: Color magenta) element. > e2 translateTo: 20 @ 0. > e2 @ RGGridDraggable. > > (line := RTLine edgeFrom: e1 to: e2) shape color:Color black. > > v add: e1; add: e2; add: line. > > v open. > ----------------------------- > > Obviously it needs a lot of fine-tuning and such. Since RTInteraction (RTDraggable) instances are not created directly the gridSize would have to be stored probably class side. Also there is a problem with resetting the "sum" variable when dragging starts. (There is no TRMouseDragStart event as far as I've seen). And also drawing the grid itself the way I've done is very brutal and doesn't really work if I want to move the canvas camera (perhaps image pattern as a background would work better). > > What are your thoughts on this? > > Peter > > > > > On Fri, Aug 1, 2014 at 4:07 PM, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > Hi Peter! > > > While playing with RTMultiLine I thought that it would be nice to have option for RTDraggable elements to snap to a grid / move only by fixed increments - for example the position could only be divisible by 10 (grid size would be 10), so dragging from 20@10 to 26@4 would jump to 30@0. Is something like this possible with current implementation? > > > I first thought this would be easy. But actually it is not. > The way to achieve the snap on grid behavior is to have an indirection in the Trachel classes. Each trachel shape knows its position (the position is contained in the matrix attached to each shape). The snap-on-grid means you need to slightly process the matrix coordinate. Maybe there are other way to do this, but it does not look like. Changing Trachel because of this seems a bit overkill... > > Cheers, > Alexandre > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > On Jul 31, 2014, at 3:58 PM, Peter UhnÃ¡k <i.uhnak(a)gmail.com> wrote: > > > While playing with RTMultiLine I thought that it would be nice to have option for RTDraggable elements to snap to a grid / move only by fixed increments - for example the position could only be divisible by 10 (grid size would be 10), so dragging from 20@10 to 26@4 would jump to 30@0. Is something like this possible with current implementation? > > > > The second question is: is it possible for element to subscribe for position updates of another element? (I.e. element A watches for movement of element B and moves accordingly). I know I could add both of them to a composite shape or RTGroup (RTDraggable>>groupToDrag:), however I would prefer something more loosely coupled. > > > > In RTDraggable>>initializeElement: i found > > " element when: TRMouseDragging do: [ :e | e element translateBy: e step. e signalUpdate ]" > > however it is commented out so perhaps there some other way? > > > > Peter > > >

Aug. 1, 2014