Roassal3 and Bloc
Locked
 
|
Hi Alex,
I forgot to ask at ESUG... but what are the plans for moving towards Roassal3 and Bloc? Because currently I am very hesitant in engaging with Telescope which may or may not solve some of my problems... but the code is really, really hard to read for me. So from this perspective it might be more fruitful for me to spend energy on Bloc, provided there will be some more gradual movement from 2 to 3. Thanks, Peter _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
> Because currently I am very hesitant in engaging with Telescope which may or may not solve some of my problems... but the code is really, really hard to read for me.
Yeah, Telescope code is a bit obscur time to time. > So from this perspective it might be more fruitful for me to spend energy on Bloc, provided there will be some more gradual movement from 2 to 3. Well… The goal of Roassal 3 is to embrace Bloc and all the good things it brings, which includes fancy events, OSWindow and other rather technical but important points (eg., supercool event handlers). In addition the GT team is helping Alain, which is somehow a quality label for me. Alex told me that during the next two months Bloc may be severely changed. In any case, Roassal3 should remain largely compatible with Roassal2. The API will be roughly the same, so migrating code should be rather pretty straightforward. If you are ready to invest now in Bloc, then you are very welcome! Widgets are missing I believe. What are you plans now? Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
Well… The goal of Roassal 3 is to embrace Bloc and all the good things it brings, which includes fancy events, OSWindow and other rather technical but important points (eg., supercool event handlers). In addition the GT team is helping Alain, which is somehow a quality label for me. My understanding was that this would basically replace Trachel layer or so? So if I wanted to have some nice shapes and visual objects I could compose them in Bloc? Or am I mistaken?
So this would mean that work on Roassal3 will start in two+ months? In any case, Roassal3 should remain largely compatible with Roassal2. The API will be roughly the same, so migrating code should be rather pretty straightforward. I don't think that this is either possible nor welcome, because if Bloc introduces different ways to use events and such, you would have to create an extra layer just to hide Bloc... but will see how this will evolve. Peter _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
> My understanding was that this would basically replace Trachel layer or so? So if I wanted to have some nice shapes and visual objects I could compose them in Bloc? Or am I mistaken?
Yes, Bloc will replace Trachel. But since Bloc has more than Trachel, Roassal will be slightly (positively) impacted. > So this would mean that work on Roassal3 will start in two+ months? Maybe earlier. I started to work on it actually. I already have something that runs actually :-) Cheers, Alexandre > In any case, Roassal3 should remain largely compatible with Roassal2. The API will be roughly the same, so migrating code should be rather pretty straightforward. > > I don't think that this is either possible nor welcome, because if Bloc introduces different ways to use events and such, you would have to create an extra layer just to hide Bloc... but will see how this will evolve. > > Peter > _______________________________________________ > Moose-dev mailing list > [hidden email] > https://www.iam.unibe.ch/mailman/listinfo/moose-dev -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
In reply to this post by Peter Uhnak
Hi Peter, I don't know exactly what are your problems but if it is to create really high level visualizations with lot of interactivity, Telescope is definitely the good tool to use. You should take the last version (https://ci.inria.fr/moose/job/Telescope/lastSuccessfulBuild/artifact/Telescope.zip) and look at the TLDemos class methods to have examples of the way to create or use Telescope visualization. If your problem is about visualization but not yet managed by Telescope; ask me and I will look at it. The Telescope philosophy is to provide a stable API with a smart model that ease the life of the user; that's why we prefer to implement less things but have them well designed. Cheers, 2015-07-20 18:58 GMT+02:00 Peter Uhnák <[hidden email]>:
Guillaume Larcheveque
_______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
That was my understanding so I wanted to look at it. But without any documentation it was hard for me to even begin. Usman Bhatti posted in another thread link to a Telescope paper, so I'll read it tomorrow and hopefully it will give me a little bit of insight.
Well we have already heavily invested in Roassal and this is not going to change, since we are continuing to work even on quite low level stuff (like bendable lines, elements resizing, attach points, ...). But I wanted to see, if Telescope could help us with some high-level abstractions, which are currently quite painful for us. I even started implementing some Eclipse GEF ideas... So I'll read the paper and I'll see... Peter _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
the paper you mention was written very quickly and might not be that much helpful At some point we will try to write something better Don't hesitate to come back to us with questions nicolas On 21/07/2015 01:55, Peter Uhnák wrote:
_______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
In reply to this post by Peter Uhnak
Hi Peter, The paper is not the best way to see what Telescope is able to. 2015-07-21 1:55 GMT+02:00 Peter Uhnák <[hidden email]>:
Guillaume Larcheveque
_______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
Just my 2cents:
As far as I understand one of the main features of Telescope are “Visualizations” that essentially are models that you set up and then they spit out a view on demand. Now builders in Roassal were designed to be the same. Here is the first page of agile visualization book, that is about builders and was first written as an IWST paper that was not accepted: http://www.slideshare.net/GuillaumeLarcheveque/telescope-introduction-and-example. It tells that there is a #renderIn: method that should be overwritten by a builder developer and which accepts and view and does what the builder should do. Now some time after this documentation was written, people started develop “stateless” builders that don’t collect all data but instead instantly draw something. I started to complain about this already at that time… Now as I see a situation now is that Telescope is a part of Roassal that Roassal voluntarily refused to be. From my point of view the only reason to do this stateless builders is because they are easier to implement, you don’t have to create single building method and so on. Otherwise they are much harder to use because you have to remember order. Uko
_______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
Writing visualization that are stateless is indeed harder. Yuriy, you said it well.
The paper is here: https://dl.dropboxusercontent.com/u/31543901/AgileVisualization/Builder/0201-Builders.pdf Alexandre > On Jul 21, 2015, at 10:17 AM, Yuriy Tymchuk <[hidden email]> wrote: > > Just my 2cents: > > As far as I understand one of the main features of Telescope are “Visualizations” that essentially are models that you set up and then they spit out a view on demand. Now builders in Roassal were designed to be the same. Here is the first page of agile visualization book, that is about builders and was first written as an IWST paper that was not accepted: http://www.slideshare.net/GuillaumeLarcheveque/telescope-introduction-and-example. It tells that there is a #renderIn: method that should be overwritten by a builder developer and which accepts and view and does what the builder should do. > > Now some time after this documentation was written, people started develop “stateless” builders that don’t collect all data but instead instantly draw something. I started to complain about this already at that time… Now as I see a situation now is that Telescope is a part of Roassal that Roassal voluntarily refused to be. From my point of view the only reason to do this stateless builders is because they are easier to implement, you don’t have to create single building method and so on. Otherwise they are much harder to use because you have to remember order. > > Uko > >> On 21 Jul 2015, at 09:44, Guillaume Larcheveque <[hidden email]> wrote: >> >> Hi Peter, >> >> The paper is not the best way to see what Telescope is able to. >> >> Look at the demos and the presentation Anne did at Esug: http://www.slideshare.net/GuillaumeLarcheveque/telescope-introduction-and-example >> >> 2015-07-21 1:55 GMT+02:00 Peter Uhnák <[hidden email]>: >> >> I don't know exactly what are your problems but if it is to create really high level visualizations with lot of interactivity, Telescope is definitely the good tool to use. >> >> That was my understanding so I wanted to look at it. But without any documentation it was hard for me to even begin. >> Usman Bhatti posted in another thread link to a Telescope paper, so I'll read it tomorrow and hopefully it will give me a little bit of insight. >> >> If your problem is about visualization but not yet managed by Telescope; ask me and I will look at it. >> >> Well we have already heavily invested in Roassal and this is not going to change, since we are continuing to work even on quite low level stuff (like bendable lines, elements resizing, attach points, ...). >> >> But I wanted to see, if Telescope could help us with some high-level abstractions, which are currently quite painful for us. I even started implementing some Eclipse GEF ideas... >> >> So I'll read the paper and I'll see... >> >> Peter >> >> _______________________________________________ >> Moose-dev mailing list >> [hidden email] >> https://www.iam.unibe.ch/mailman/listinfo/moose-dev >> >> >> >> >> -- >> Guillaume Larcheveque >> >> _______________________________________________ >> Moose-dev mailing list >> [hidden email] >> https://www.iam.unibe.ch/mailman/listinfo/moose-dev > > _______________________________________________ > Moose-dev mailing list > [hidden email] > https://www.iam.unibe.ch/mailman/listinfo/moose-dev -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
In reply to this post by Uko2
Hi,
On Tue, Jul 21, 2015 at 10:17 AM, Yuriy Tymchuk <[hidden email]> wrote:
So, how do you tell builder that you want to color or add interaction to a "group" of nodes and not all the nodes in the visu? Groups are first-class entities in Telescope. Composites are first class entities with customizable interactions.
What is the stateless builder because the paper does not mention it or may be I missed something? If it is about parameterizing the visu, then telescope offers this mechanism with the help of blocks and selectors and the visualization is therefore disconnected from a particular usage. Moreover, the model takes care of updating only the concerned nodes and not all of the visualization.
Telescope was conceived to write reusable visualizations and it wasn't destined to replace Roassal from the beginning. But there are some interesting concepts in Telescope that you might see taken up in Roassal 3. Telescope should be seen as high-level composable model for a visualization and we have created visualizations that are highly interactive. We have done things like showing connections amongst several levels (e.g. packages, classes, methods) and it wasn't too difficult to do with a clean model and composite shapes with interactions. Telescope is a rich model for visualization and it does not replace the rendering features of Roassal (at least not for now). regards. usman
_______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
I’m not saying that Telescope does not have cool features that are not available in Roassal. What I say is that I believe that builder is made to ease people lives and the natural evolution of builder are Telescope Visualizations. Some people my not agree with me, but I think that it’s strange that Roassal builders didn’t move that way, and as there are already this cool features in Telescope they are not integrated into Roassal.
Uko
_______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
On Tue, Jul 21, 2015 at 11:40 AM, Yuriy Tymchuk <[hidden email]> wrote:
The we are on the same page :)
I know Roassal team is thinking about it :)
_______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
«
Return to Moose
|
1 view|%1 views
| Free forum by Nabble | Edit this page |