on 28/3/02 11:44 PM, Matt Denton at [EMAIL PROTECTED] wrote:

>> The plug-in API is published, so it is certainly possible to do - would
>> be
>> much better done within the engine though.
> 
> Hey Ben, not sure what you mean here, do you mean somehow play Flash
> directly, without using the plug-in?  I guess QTime is played from

No, I meant it would be better for the engine to host browser plug-ins,
rather than

> I wonder how hard it would be to get an XCMD or XOBJ or DLL to read a
> web-browser plug-in?

Because of the facilities that the plug-in API (at least the original embed
one, I'm not familiar with the ActiveX) provides for a plug-in to call back
to the host (browser), it would be much more functional if the hosting
support was built into the MetaCard engine, rather than into an XCMD which
has a limited integration with Rev/MC.

> That to me sounds like a great idea, as I'm sure there are overhead and
> compatibility issues with the plug-in path.  Would that mean licensing issues?

There are two ways to go: Macromedia have published the format, so one can
write one's own player (or authoring system) from scratch, with no licensing
issues - but I wouldn't recommend it!  Alternately you can license the
player code from Macromedia - I know nothing about what terms they offer.
This would be more plausible; but my vote would be to put the effort into
creating the general mechanism to support browser plug-ins, rather than
specifically integrating the Flash player.  (Or if Scott and the team do
want to write original code, SVG support...)

Anyway, that's what I meant: better to have plug-in support in the engine
not an XCMD.

> Sounds like it was well worth it, and possibly fun too.  Out of
> interest, how long did the play-plug-in it take from start to finish?

Hard to tell exactly, but I've just gone back and checked my time report.
In 1996, I seem to have recorded 36 hours on this before reporting that it
was working for 68K plug-ins on the 68K version of my engine; and a total of
57 hours when I wrapped it up with the 68K and PPC support in 68K and PPC
engines, tested on some plug-ins.  Scarily however I then find that I spent
some 40 hours in 1998 getting the molecular model viewer plug-in working on
Mac and Windows engines (and I'm pretty certain that a colleague had already
implemented basic plug-in support on Windows before that).  I think this was
partly because there were some nasty bombing issues, but it's still more
than I'd remembered.  But then again, a better programmer might be much
faster (that's not false modesty - I was operating close to the edge of my
depth, especially on Windows).
 
  Ben Rubinstein               |  Email: [EMAIL PROTECTED]
  Cognitive Applications Ltd   |  Phone: +44 (0)1273-821600
  http://www.cogapp.com        |  Fax  : +44 (0)1273-728866


_______________________________________________
improve-revolution mailing list
[EMAIL PROTECTED]
http://lists.runrev.com/mailman/listinfo/improve-revolution

Reply via email to