Gephi forums

This is the thread for asking more details about the WebGL proposal.

AFAIU, this project is not related to original gephi project

in this we have to start a new project, by developing an efficient network visualization library for the web using WebGL. am I right ?

Hi,

I understand that this project has been scrapped from the GSoC ideas list, but if there is any splinter of a chance that Gephi's accepting a proposal on this, I think I'd still try... (but please let me know if there's no point to do so)

Have a question though: how is the user expected to use this library? I get the impression that the user will be able to associate say, a gexf file, then use the library to define graphics attributes of node/edge, layout properties and preferred metrics calculations of the graph? So in this case the library means to abstract interaction from the user i.e. the library will provide a web-based display, but with API for the user to work with network properties -- is this the right understanding?

Hi Juliana, I'm sorry to say that we will not go for this proposal. We have a limited number of 'slot' Google is giving to us, and also a limited number of mentors. We also believe WebGL is still a very young technology and waiting for better standard and support is relevant.

The proposal was not going so far and mostly concentrate on the visualization issues. However questions remains on the best way to provide an efficient network visualization platform on the web. For us we believe the first step is to let users have the same level of interaction Gephi can have on the desktop, but without the analysys platform. It's a bit what was done with the GexfExplorer project, which provide mostly read-only features, based on what GEXF file contains. I would also cite this project, http://cytoscapeweb.cytoscape.org/. It shows great effort to bring to the web complex features and consistency between desktop and web applications. But unfortunately it stuble on efficienty, try to load a 100 nodes graphs!

Hi mbastian, thanks for the reply. Guess this isn't the best timing =( In any case, thanks for the heads up on GexfExplorer and Cytoscape. Perhaps next year will be a better time for Web Gephi!

Hi dev,

First of all congratulations for get selected on GSoC 2011 as a mentoring organization. I'm interested in the project idea 'Web based network visualization using WebGL'. As I understood, the objective of the project is to come up with an abstraction layer on top of an efficent library (i.e. PhiloGL) to layout large networks loaded as GEXF/JSON. I'm reading about PhiloGL & WebGL these days.
It's great to see a similar idea 'Use of Hardware Accelerated Techniques' in Javascript Infovis Toolkit (Sencha Labs) GSoC project ideas list too. May be Gephi & JIT can collaborate in GSOC 2011 to make this idea a reality (Just a thought ).
I would like to know whether this project's scope cover considering other alternatives like d3.js and processing.js too for visualizing large networks on the web. I would like to discuss more on this idea with you all in this forum & mailing list and come up with a good work plan for the summer.

While trying to mess around with WebGL to get a feel for what it is I'm getting myself into it I found that Chrome doesn't have support enabled for WebGL enabled by default in Ubuntu. You can read more about getting WebGL set up here.

Buyer beware, I didn't actually use those instructions. Some Googling discovered you could enable WebGL in chrome with:
That didn't work, but some more Googling discovered that it had been added to a blacklist and you could get around that blacklist with:
Hopefully this saves someone some time.

May be Gephi & JIT can collaborate in GSOC 2011 to make this idea a reality

Yeah very good idea, I sent an email to the author. I'm happy they are doing GSoC as well, it's an excellent project !

I would like to know whether this project's scope cover considering other alternatives like d3.js and processing.js too for visualizing large networks on the web

We have considered this (see this topic), these librairies offer great features and flexibility but they are not performant enough for large graphs. The limits would be around 5000 nodes, and the experience limited, as zooming and panning speed would be reduced.

joshua wrote:While trying to mess around with WebGL to get a feel for what it is I'm getting myself into it I found that Chrome doesn't have support enabled for WebGL enabled by default in Ubuntu.

Thanks joshua. It was working in Chrome 9 but stopped working after an update to Chrome 10. Seems some graphic card drivers with OpenGL are heavily buggy. So engineers at Google and Mozilla have decided to blacklist certain models of graphic cards (ATI, Intel etc).

joshua wrote: That didn't work, but some more Googling discovered that it had been added to a blacklist and you could get around that blacklist with

You can find a blog post here I wrote about getting WebGL to work in Chrome 10+ and Firefox 4. Hope it'll be beneficial to many, who are having the same problem.

Since this idea was scrapped last year I would like to double check how the Gephi team views this idea as a GSOC project this year? I would hate to write a proposal for it only to later find it scrapped due to other priorities in the team

Cheers,
Urban