The attached file (BlackJava.0.1.N.s.gnutar.gz.crypt) is an encrypted version of BlackJava.0.1.N.s.gnutar.gz. BlackJava.0.1.N.s.gnutar.gz contains the source for my partial port of Java to NeXTStep (m68k). You have to use gnutar because some path names are >100 characters.
Aha! What's the reason for all this cloak and dagger stuff? Well, you have to register with Sun to receive a copy of their Java source. They're trying to track who has a copy and why. So I can only release the key to people who have also registered with Sun. Mail me at dave@prim.demon.co.uk with a copy of the automatic email you got when you filled in their source request Web page and I'll mail you the key.
I'm placing the sources here because the transatlantic Internet speeds are incredibly slow at present and it's costing me a fortune to mail them to people.
Many thanks to the site maintainers for permission to do this.
Dave Griffiths
PS: Here's a mail message I sent about the port.
From: David Griffiths <dave>
Date: Sun, 15 Oct 95 13:51:03 +0100
To: Domagoj Cosic <domagoj@toaster.cs.tu-berlin.de>
Subject: Re: Porting Status
cc: next-java@friday.com,David George <david@threewiz.demon.co.uk>
>Just today, I found that the Java source is (freely)
>available, and that there are already folks out there
>doing the NeXT port, and that there is this mailing list I
>just joined. I also have read that there exists a port by
>David Griffiths.
>
>I do not have much time and resources, but as I am
>intested in the final result (i.e. Java running on NeXT
>Step platform), I would like to help as much as I can. So,
>first of all, I ask anyone to inform me about the current
>project status and the issues still to be done.
As far as I can tell (and hopefully someone will correct me if I'm wrong), there isn't just one project, a number of people are attempting NeXTStep ports independently. I made a start on one myself just for fun and to learn something about Java. I called it BlackJava. It was fairly easy to get the runtime stuff working because Mach already provides support for threads and conditions. Note though that when I say "working", the approach I took was just to do the minimum necessary to get the compiler working. Then I put it all under ProjectBuilder (this was more difficult than porting the runtime!) and finally made a start on the Awt stuff. And I mean "a start". :) The only thing I've done is to implement the bit that creates windows, so you can actually run HotJava and a whole mass of windows appears on the screen (I didn't even bother to create them hidden - just wanted to see something on the screen!).
So there you go, pretty rough really. I was concentrating on being able to run HotJava as quickly as possible and ignoring any esoteric areas.
The interpreter and compiler seem to work OK though.
BUT, I'm due to start work in the States next week and will no longer have much time to work on this. David George would like to take over coordinating the port if anyone else wants to help out. The source has also been mailed to Bill Bumgarner <bbum@friday.com>, Dirk Schwarzhans <dirk@diaspar.fb10.tu-berlin.de> and vitek jan <jvitek@cui.unige.ch>. They are all registered with Sun as Java porters (a necessary condition to have a copy of the source).
So if you'd like to help port some Awt stuff, drop me (or David) a line. A lot of it should map fairly easily onto AppKit classes. (Hmmm...)
Oh, one other thing - Awt is supposed have changed a fair bit in the beta and the beta sources aren't out yet.
