[1345] in java-interest
Re: Commercial view of Java ...
daemon@ATHENA.MIT.EDU (Rawn Shah)
Wed Aug 30 17:14:18 1995
From: Rawn Shah <rawn@rtd.com>
To: Michael.D.Byrne@norwest.com (Michael D. Byrne)
Date: Wed, 30 Aug 1995 11:28:42 -0700 (MST)
Cc: java-interest@java.sun.com
In-Reply-To: <30438a8b2006002@nwhub1.norwest.com> from "Michael D. Byrne" at Aug 29, 95 04:45:48 pm
:
: So, what would I like to see in Java?
:
: 1. Persistence of objects at the workstation in a way more elegant than cache.
Interesting but I do not think necessary for every instance. I think it
may cause unnecessary load for temporary objects. Unfortunately, there isn't
multiple inheritance. If there was, subclass Object to become persistent
and inherit this as well as any other class to make it work inside your
prog.
:
: 2. Automatic versioning of applets!! Get a new one if & only if you need
: to, automatically.
This I like. Versioning of applets may be quite important in some cases.
Unfortunately, the problem comes around to the fact that who will decide
how to create the version numbers if there are several copies running
around. Another one: In the case of a Web applet, do you enter the
version number at the APP tag or inside the class? If inside the class,
it'll download the entire class to look at the version number anyway.
:
: 3. People use a whole desktop, not necessarily just your application. Need
: OLE or some good way to move applet data into a spreadsheet or
: correspondence system or word processor (and no, I don't want to get a new
: Java-based word processing package when we have 20,000 people to retrain.)
: Co-existence with Microsoft.
YES! YES! Sorry about the excitement. I've been yelling about this over
and over again. We need some kind of communications between classes at the
user interface and network level, not just the code level. I don't know
about reimplementing OLE because it uses a lot of elements that may not
be useful. I think we may need to look at and possibly implement several
object management features like OLE and CORBA.
:
: 4. Maintainable, supportable, debuggable, testable code and tools. Fancy
: coders are more often a liability than an asset in the long run to my kind
: of company. So, no operator overloading. Other non-rocket-scientist
: members of the team must be able to easily read/modify the code.
:
I believe that is coming down the road. One think I was wondering is if
in the light of the drag-n-drop world of today, anyone is interested in
working on a visual coding environment?
: 5. A good way to (authenticated and encrypted) get to "legacy" data --
: ODBC, DRDA.
More interfaces to create.
:
: 6. A better way to build end-to-end applications; maybe a
: publish-and-subscribe approach like ISIS or Technekron integrated into Java.
I'm not too familiar with this. Do you mean an automated class
distribution to current apps?
:
: As far as I can tell, its Java versus Blackbird, and the winner may have
: nothing to do with technical elegance. (Otherwise, Multics and ALGOL would
: still be here.)
True; unfortunately in the larger PC world, MS is still (*barf*) king.
:
:
: Michael D. Byrne (Michael.D.Byrne@norwest.com)
: Voice: 612-667-0765 Fax: 612-667-9416
: Speaking for myself, not my company...
: V V
: AL%B000012$N (A very obscure string!)
:
: -
: Note to Sun employees: this is an EXTERNAL mailing list!
: Info: send 'help' to java-interest-request@java.sun.com
:
--
Rawn Shah RTD Systems & Networking, Inc.
Vice President 4003 E. Speedway Suite 123, Tucson AZ 85712
rawn@rtd.com (520) 318-0696 FAX: (520) 322-9755
----------------------------------------------------------------------------
-
Note to Sun employees: this is an EXTERNAL mailing list!
Info: send 'help' to java-interest-request@java.sun.com
