[1439] in java-interest

home help back first fref pref prev next nref lref last post

Re: Interprocess communications

daemon@ATHENA.MIT.EDU (Michael Lorton)
Sat Sep 2 00:36:47 1995

Date: Fri, 1 Sep 1995 18:11:39 -0700
From: Michael Lorton <mlorton@eshop.com>
To: ser@jersey
Cc: java-interest@java.sun.com
In-Reply-To: <9509012047.AA00572@jersey.> (ser@jersey)


> 
> > Communicate through a class-static object and load using the "SRC"
> > argument in the HTML.
>                   -- Michael Lorton <mlorton@eshop.com>
> 
> I don't understand how using SRC will affect anything;

Actually, I don't either, I can only tell you empirically that if you
*don't* use SRC, each applet is put in its own name-space.  I am told
that SRC will fix this bug (feature?), but I have not tested it myself.

> do you mean have, say, 
> classes/port1, classes/port2, classes/port3... and have the same interface in 
> each of them with a static, synchronized, public String as the only member?  
> Will this ensure that each instance of the String will be disjoint with 
> interfaces in other directories, but unique in applets that implement the 
> interface from the same directory?

If I understand what you are saying, yes.

> It would be an interesting solution, but 
> far from ideal.  It feels like a hack.

It depends.  If you put an Observable in the class-static and the
Applets all Observed it, it would work quite elegantly

> What we really need are true interprocess communication ports that
> are oblivious to the host address and port number; something
>  indexed by a name.  Is this not possible?

You mean so that two otherwise unrelated applets could talk to each
other.  Sounds good to me, but no, I don't see how.

> --- SER

"To SER with love..."

M.
-
Note to Sun employees: this is an EXTERNAL mailing list!
Info: send 'help' to java-interest-request@java.sun.com

home help back first fref pref prev next nref lref last post