"Aaron M. Renn" <[EMAIL PROTECTED]> writes:
> I'm now the proud owner of a glibj.zip file that is 449,109 bytes --
> about 5% of the size of Sun's latest runtime library. Way to go
> guys. Just say no to bloat.
Wow. One of these days I might get around to trying to build all of
Classpath myself. :) Don't forget that java.awt.* is a TON of code, so
that could make our size grow by just a weee bit.
> -- GtkToolkit and Test in the GTK peer implementation didn't
> compile. I think this has something to do with the fact that we
> don't have a working java.awt yet. Anyhow, this is still a work in
> progress so I removed them from the build.
The GTK+ peers are being developed with Sun's VM and classes.zip.
This is good, as it means we can soon (and I use that term loosely)
post a screen shot of some nice big Java program like HotJava running
with GTK+ widgets. I can't wait for that to work.
And FYI, the GTK+ peers are a lot faster than the Motif peers under
GNU/Linux -- A LOT faster. Clicking, typing, and other input stuff is
actually responsive.
> I'm gonna get the GTK+ CVS latest and greatest soon so I can play
> with this super cool stuff. These are fully Gnome integrated right?
The peers rely solely on GTK+ code. Gnome functionality tends to fall
outside what the AWT provides. I'd eventually like to create a
gnu.gnome.* package namespace to house Gnome functions and integrate
that well with the existing AWT. For example GnomeMenu would extend
java.awt.Menu but allow the addition of pixmaps in the menu items.
I commonly uncover bugs in the GTK+ CVS tree -- as such, there's
usually a lag time between my bug fixed copy which can run the GTK+
peers, and what's available in the main CVS tree.
Points of interest in getting the GTK+ peers to run:
1) There's an extremely nasty deadlock bug in the current GLIB
codebase. You'll need to apply glib-rao-981218-0.patch.gz from
ftp.gtk.org/pub/gtk/patches in order to fix it. (that patch still
hasn't been moved from /incoming as of this post, but it should
show up in /pub/gtk/patches shortly -- if not, email me and I'll
send you a copy)
2) You'll need the latest greatest Blackdown JDK with native thread
support -- and you'll need to turn those native threads on.
3) Build GLIB with pthreads enabled.
Unfortunately, the requirement on pthreads cannot be removed until we
switch over to JNI 1.2 code (ie. start using 1.2 VMs for development).
--
Paul Fisher * [EMAIL PROTECTED]
