This is the mail archive of the java-discuss@sourceware.cygnus.com mailing list for the GCJ project. See the GCJ home page for more information.
>>>>> "Terry" == Terry Sikes <tsikes@netcom.com> writes: Terry> From the sound of things you're planning on implementing GUI Terry> support at some point(?). Our plans here aren't finalized. Terry> Is there a web page outlining your implementation plans and Terry> timelines in more detail than "Cygnus Implementation Plans for Terry> the Java(TM) Platform"? I don't think so. I'm pretty sure we wouldn't make any sort of timeline available for net releases anyway. This is more or less "GNU Standard". Terry> I looked around but didn't find anything... I'm particularly Terry> interested in Swing support on Linux. Wouldn't it be possible Terry> to bypass much of the AWT implementation if the approach was to Terry> get Swing running first (no native peers)? What is the status Terry> of JDK 1.2 support in general? I don't really know anything about AWT or Swing. So I can't answer those questions. We've been targetting JDK 1.1, but we have some JDK 1.2 stuff in there. Still, the JDK 1.2 support isn't complete (or even really consistent right now). Terry> What GC approach are you taking? Have you considered Terry> implementing "pluggable" GC? We're currently using the Boehm conservative GC. We've written our own mark function, so Java objects are scanned precisely. Alex wrote a paper proposing pluggable GC as the appropriate choice for embedded Java. So I'd guess it is safe to say that we'll be doing that. This won't be in the first net release, though. Before you ask, we do already have pluggable threads. We currently have three thread implementations: * No threads * Quick Threads (this is a longjmp-style cooperative thread system) * POSIX threads (not heavily tested, but it does work on my Debian box) Writing a new thread system port is fairly simple. Tom