[Date Prev][Date Next][Thread Prev][Thread Next] - [Date Index][Thread Index][Author Index]

Re: AO40rcv and the Kenwood TS-2000

> Date: Wed, 04 Apr 2001 20:30:07 -0400
> From: Margaret Leber <maggie@voicenet.com>
> Subject: Re: [amsat-bb] AO40rcv and the Kenwood TS-2000
> Tony Langdon wrote:
> >> We don't need a protocol, only a set of capabilities that 
> >> need to be available,
> >> and a set of object methods which support them. Java is 100% 
> >> network capable.
> > 
> > Which sounds like an application layer protocol of some sort...
> Java *has* protocols for calling methods on remote objects, so there's 
> no contradiction here.
> 	[...]

Well, sort of.  I wouldn't call them particularly open or interoperable.

A group of us tried using the Java object serialization facility to
transfer Java objects containing data between Java applications on
different machines.  While it seemed like a neat idea at the time,
perhaps even a little bit elegant in an object-oriented sort of way,
it was largely a mistake.  If we were to do it all over again, we would
design a simple protocol to encapsulate the data, rather than relying on
the magic of object serialization.  We just ran into too many problems.
Java did a tremendous amount of work behind our back, such as chasing
through all the objects that were referenced by the object being serialized.
Bad things happened when we recompiled the software and then tried
to read archived data (stored in the form of serialized Java objects).
The use of serialized Java objects seemed to require more coordination
between developers than would have been the case if we were using
a simple protocol (e.g., more cases of "Ok, I just changed something
so everyone has to recompile their code").

Using Java RMI (remote method invocation, or calling methods on remote
objects) also pretty much assumes that the remote software is written
in Java.  A reasonable protocol would be language-independent.

At this point, I am not sure whether we are talking about technologies
in general or a specific application of technologies.  However, for
interconnecting amateur radio applications across a network, I recommend
that some simple (extensible) protocol be developed (even if both ends of
the connection are written in Java).

Via the amsat-bb mailing list at AMSAT.ORG courtesy of AMSAT-NA.
To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org