> From: Peter Reilly [mailto:[EMAIL PROTECTED] > > > No, the optimization is that if the <typedef/> does not define a > classpath, a cached classloader is used. This > cached classloader gets set once (the first call). > > Jose's more general caching solution souds more promising. >
Peter, I could not work on this during the weekend, but will try to post it tonight. Some additional questions I have (for improvements): 1) It seems that we want a static cache, instead of per project. I can do this, but I am a little concern with IDEs keeping classes longer than we want. An alternative is to pass the parent cache to <antcall/> subproject (or a clone of it). Just like we do with some other system-like properties. Suggestions. 2) Looking at the code for createClassLoader(Path), as I mentioned before, we never seemto pass getCoreLoader() as the parent of the classloaders being created, if we were, we would be reusing the parent all the time and maybe gaining some savings. Is this kind of a very scarry change to do? Some of the complaints about the bootclassloader being used, may be fixed by doing such a change. 3) Since we do not need to support JDK1.1 anymore, can we get rid of the 1.1 specific reflexion magic in createClassLoader() and use an AntClassLoader2 directly? Opinions? I think I can incorporate some of your input in this regard quite easily. So let me know what you think is best. Jose Alberto --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]