[vtk-developers] [GSoC 2014] Supporting a Visualization Grammar

[Marco Cecchetti]

Hi Jeff,
thank you for your reply. Well indeed, the "huge" adjective was referred 
to the VTK framework, but I agree that implementing all Vega's features 
would be a too big project for the GSoC time-frame.

As probably you noticed I posted a proposal. It lacks a list of 
deliverables and a tentative time-line. I will provide them on a few 
days, as soon as I get a deeper understanding of the subject.
In order to achieve this goal I would need a bit of your help, so I want 
to share some thoughts and questions I run into by the last mail.

Vega accepts data sources in several formats, such as the JSON format. 
Moreover Vega's grammar is in JSON format, too. Hence, a JSON 
parser/reader is surely needed for this project. I can see that a third 
party JSON library (jsoncpp) is already included in the VTK 
distribution. I was wondering if the project should exploit the jsoncpp 
library or should implement a native VTK solution for parsing files in 
JSON format (something like vtkXMLParser for xml). Using the jsoncpp 
library would help to have a longer time for developing the Vega's 
grammar and  hence to provides support for a a larger subset of Vega's 
features. On the other hand a native solution could provide more 
flexibility and it would not depend on external libraries.
What is your point of view ?

I browsed several VTK class interfaces, such as vtkContextView, 
vtkContextScene, vtkContext2D, vtkContextItem, vtkActor2D. What is the 
different role of a vtkContextItem respect with a vtkActor2D ?
Why most charting classes are present both as subclasses of 
vtkContextItem and as subclasses of vtkActor2D (e.g. : vtkChartXY vs 
vtkBarCharActor, vtkAxis vs vtkAxisActor, and so on) ?
Vega's Mark elements implementation in VTK should be based on 
vtkContextItem or vtkActor2D ? According to the feeling I got of 
charting support in VTK, I think that they should be implemented as 
subclasses of vtkContextItem, but I am a bit confused.

Kind Regards,
-- Marco



Il 20/03/2014 19:02, Jeff Baumes ha scritto:
> Marco,
>
> Thanks for your interest. It's true that Vega is a large system. We
> would be looking for proposals to tackle an initial framework that
> supported rendering some subset of Vega, and yes options include
> leveraging charting capabilities in VTK. I could see this being used in
> particular for rendering axes, which are abstracted in both Vega and VTK.
>
> Jeff
>
>
> On Thu, Mar 20, 2014 at 1:36 PM, Marco Cecchetti <mrcekets at gmail.com
> <mailto:mrcekets at gmail.com>> wrote:
>
>     Hi,
>     I am a student in mathematics at university of Pisa. My university
>     courses are essentially on mathematical subjects, however I applied
>     for a course in programming and data structure and a course in
>     numerical analysis. I have already applied to GSoC in the past years
>     and I am interested in the "Supporting a Visualization Grammar"
>     project. My main programming languages are C++ and JavaScript.
>
>     I built the VTK framework and create a Gerrit account for review
>     tasks. Could you point me to a list of bugs or patches I could work on ?
>
>     I read the Vega library documentation and played a bit with the
>     online editor, moreover I started analyzing some VTK examples, but
>     it is really a huge framework! I noticed that there are some
>     plotting/charting classes and I was wondering if they could be
>     useful for this project.
>
>     I know that I should have contacted you earlier but I was busy with
>     my courses. I have not yet write down a proposal (I know that I have
>     only 24 hours left!) because I have not yet all details for planning
>     a sensible project schedule. Do you mind if I write down a proposal
>     without a time-line ? We could talk about this one and other details
>     later.
>
>     Kind regards,
>     -- Marco
>
>