RE: Moving system properties to gnu.classpath.*

[David Holmes]

I wrote:
> At present there is only one
> classloader that documents that it will do a package access
> check, and that is the classloader returned by URLClassLoader.newInstance.

Bah! I can't even read straight this morning. According to the article the
system classloader *does* perform this check - not that this seems to be
documented anywhere.

So Classpath probably should define a system class loader that does this
too.

That said, given the check only comes into play when a security manager is
installed, and it may be avoided by a user-defined classloader, does this
actually provide the protection that was desired? (And what exactly was the
protection anyway?)

David