This is the mail archive of the
mailing list for the Java project.
Re: Overview of interpreted debugging design
On Fri, 2005-05-27 at 13:14 -0600, Tom Tromey wrote:
> Keith> The back-end is segregated into several packages which deal with the
> Keith> most important areas of functionality:
> Wow, there's a lot more than I realized.
Lots of bookkeeping...
> I'm thinking the review can center on a few things:
> - Is there code that can/should be in classpath, and if so should we
> try to do that first in order to avoid renaming everything later.
It might be able to go into classpath. I'm hoping to make a back-end
which is independent of the VM, but, of course, I don't actually *know*
that this is possible. I just think it is.
If it _is_ possible, then the next logical question is: does classpath
want such functionality or even want this version of it?
> - What changes are needed to existing parts of libgcj
Ah, yes. There are changes. Some will be non-trivial. The most invasive
parts are all event notifications and thread handling. I'll be looking
for extra eyes on this stuff.
> - General style conformance; second look for big gotchas
Threading issues. I don't typically use threads often, so any extra
eyeballs appreciated here.
I guess with that said, I'll post a patch for something...