I've recently had some quite strange race condition in the applet code. Looks like init() and start() methods have been called by different threads in pre-2 java plugin version. What do you think, is this possible?
Java plugin version 2 for all of init/start/stop/destroy hooks shows that those hooks are run off plugin2.applet.Plugin2Manager$AppletExecutionRunnable. Code/logs available on request, if needed.
UPD #1: My current version is java 6 update 24, linux. I've seen no problems with my config (all hooks are run from the same thread), but maybe there're some older, 1.5+, versions of Sun's Java Plugins which do this differently (windows, preferrably, but any OS would do).
UPD #2: As for the "pre-2 java plugin version" - see the links. There was a rewrite of Java plugin across Java releases 1.5 and 1.开发者_C百科6.
- Ken Russell Discusses The Rewritten Browser Java Plug-In
- Sun’s Promise: Reliable, Portable, Functional Java Plug-In
Thanks, Anton
..What do you think, is this possible?
In some versions of some browsers running some JREs on some OS', anything is possible with applets.
精彩评论