No subject
Tue Jun 4 12:17:20 MDT 2013
context usually refer to individual windows, not to HTML frames. An applet that
started in its own window will stick around until you close the browser - unless
you manage to destroy this window from code. This is an observation from
experience - not based on theory. I am assuming the JVM that the browser provides
for your applet is truly a separate sandbox. It gets loaded on demand but isn't
unloaded until you close your browser.
Check this out to verify my statement:
http://www.archenergy.com/AECHome/ckv/oac/oac.html
The applet in the separate window will stick around until you close it or until
you close the browser. You can point your browser to any other page if you like -
you can even start multiple instances of the oac applet.
FYI - I haven't done serious applet development in close to 2 years since browser
support for this platform seems to have a dark future. So at least my experience
is rather dated.
> - The 'semantic' of the back button is different for different browsers.
> Sometimes you leave a frame, sometimes you leave the entire frameset.
This is what I am talking about. The implementation of JavaScript is browser
specific, especially for more sophisticated actions. So you will often end up
first checking for the browser ID and OS and then you invoke your JS code for the
appropriate environment.
> If've done some java stuff that relies on JavaScript/Java functionallity,
> but this was allways the most problematic spot in the application.
> Also, many users switch of JavaScript/Java interaction (and with _very_
> good reason, this 'feature' has a long history of security problems).
>
> Ralf
>
> _______________________________________________
> Web Page: http://lug.boulder.co.us
> Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug
This response is to clarify my previous message and is not at all intended to
contradict Ralf's point. We mostly simply looked at different aspects of the
problem.
BJUG is still at the same address (bjug-discuss at avitek.com) - but unlike BLUG the
BJUG disucssion list is not very active. One more link:
http://www.javaranch.com/ (watch the fly go right through the bull's nose)
:-)
Ferdinand
--
Ferdinand Schmid
Architectural Energy Corporation
http://www.archenergy.com
(303) 444-4149
More information about the LUG
mailing list