The prototype on which the patent application is based is on an IPC model
where the UA, based on information in the hypertext document, executes the
appropriate helper application, negotiates resource requirements, and allows
the app access to part of the rendered HTML window. Just the same way that
TkWWW (an early Tk/tcl-based web browser), upon needing to render a gif,
interacted with the external application xli to render inline. There is also
a proprietary protocol to generalize this across platforms. I believe that
there is only a reliable browsers are for the unix platforms at this time. The
complexities of cross platform development is the reason that others avoided
this issue earlier.
Let's hope that the USPTO realizes that smarter prototype applications in and
of themselves do not void prior art.
I do not believe that Eolas founders and UC mean to patent the other Eolas
products, most of which are the fruits of efforts of our group at UCSF, and
are only known to Eolas by our common employment history at UCSF, as we have
not packaged nor released them. [For some time, without my knowledge nor
consent, I was listed as one of the Eolas' /pub/people on their ftp site. I
have never been employed with Eolas nor have I ever worked towards software
patents.]
Of course, Eolas could put such technological speculation to rest by releasing
their patent request and license agreement with UC, since all relevant
information has published at one time or another on the web and since all of
the development work involved was done on the public dollar.
-marc