[1831] in java-interest
Re> java-interest-digest V1 #169
daemon@ATHENA.MIT.EDU (Sid Conklin)
Mon Sep 18 16:57:03 1995
Date: 18 Sep 1995 10:33:02 -0700
From: "Sid Conklin" <sid.conklin@nora.stanford.edu>
To: "java-interest@java.sun.com" <java-interest@java.sun.com>
Ditto on Phil's viewes. I'll take a 99% solution anyday!
Knowing that all platforms have these classes allows the developer an easy
way to communicate with the host clients and servers in a
platform-independent fashion. I was really counting on these classes, now I
will have to roll my own(*#$*%#@).
Thanks I'm getting off my box now,
Sid Conklin
Stanford University
------ From: java-interest@java.sun.com, Mon, Sep 18, 1995 ------
From: phil@bolthole.com (Philip Brown)
Date: Sun, 17 Sep 1995 18:27:34 -0700 (PDT)
Subject: Re: JDK Question
>
> :
> : I think tcp/ip networking is rather platform independant.
> : To say someone will only want to use network functionality if they are
> : running on a sun, is silly.
>
> That's a very naive viewpoint of how networking actually exists in the
world.
> TCP/IP has several different programming API standards: Berkeley Sockets,
> SVR4 TLI, XTI, Windows Sockets, MacTCP (if you can actually call it that).
Yes, I know that.
However, that doesn't stop the need to have a standard way of utilizing
sockets for java.
> The Java classes NetworkClient, etc. are an even more generalized
> viewpoint of network communications. It would take far too long to create
> yet another universal API set for networking.
I haven't gotten all the etails on NetworkClient, and its server counterpart.
However, they sound like what 99.999% of most folks need. Therefore, they
should be provided to everyone as a standard now, with extensions in the
future, as the need comes up.
-
Note to Sun employees: this is an EXTERNAL mailing list!
Info: send 'help' to java-interest-request@java.sun.com
