Re: WebVAT, example code

Benjy Cline <benjycline@...> writes:

> 
> Greetings,
> 
> Two tidbits for you:
> 
> 1.  Several people want to use JRex to access the pixel
> coordinates of rendered objects such as buttons and text
> fields.  JRex doesn't appear to expose the Gecko's page
> renderer's Frame Tree data structure.  However, I found
> an article about WebVAT that is an open source tool
> that uses a modified JRex and Gecko to expose Frame
> Trees.
> 
> I can't find the source code, but here are the links to
> the article and another blurb:
> 
> http://www.cs.sunysb.edu/~jmahmud/ICWE2007.pdf
> http://www.acm.org/src/subpages/borodin/YevgenBorodin.htm
> 
> I don't need this facility; however, I thought others might
> be interested in the article.
> 
> 2.  I found this source code for a JRex-based browser:
> 
http://svn.openqa.org/fisheye/browse/floyd/trunk/src/java/jrex/floyd/jrex/JRexBr
owser.java?r=45
> I'm not sure what the project is all about; however, it
> has methods that illustrate how to programmatically generate
> click events.  Someone might be able to use that as an
> example.
> 
> I haven't actually tried the code.
> 
> Benjy
> www.benjysbrain.com
> 


I wouldn't bother trying to follow up on anything from Ramakrishnan's group at 
Stony Brook, since they are very secretive and refuse to release any useful 
code. Consider the case of WebVAT, which was reimplemented based on Mozilla's 
browser. As far back as 2007, they promised to release WebVAT as open source ( 
see http://www.cs.sunysb.edu/~borodin/publications/ICDAR07.pdf ), but 3 years 
have gone by, and they still haven't done it. Ditto for HearSay, CSurf, and all 
of the other minor variations of the same algorithms that they have been 
publishing over and over again at tons of different conferences ( 
http://www.cs.sunysb.edu/~hearsay/publications.htm ) 

Bottom line: all smoke and mirrors. Move along, there is nothing to see here.