Re: OpenDoc viability

Ron Resnick (resnick@interlog.com)
Fri, 4 Apr 1997 07:21:04 -0500 (EST)


On Fri, 4 Apr 1997 01:21:55 -0800 Peter Bayley (bayleyp@dnc.net) wrote:
>
> I read your web page on:
> <http://www.interlog.com/~resnick/javacorb.html>
>
> You are most probably aware of Apple dumping OpenDoc and CyberDog
> development.
>
> Currently CILabs has the source code, and are despirately searching
> a way to keep the technology viable. Their time is running out, funding
> will dry up very soon.
>
> There are three solutions I have thought of.
>
> 1) Port to Java.
>
> 2) Get a buyer of the technology(who will no doubt port it to their
> object model).
>
> 3) Finish the UNIX port and allow Linux enthusiasts to keep it going.
>
> The first solution is obviously the most preferable. The second
> solution could be done if a N|C developer like Oracle wanted to use OpenDoc
> as a flexible user-end component architecture. I don't know how possible
> the third option is, given I don't know if IBM has released it's AIX source
> code.
>
> Personally I have been trying to get the word out that the world's
> best component architecture is available to anybody who wants it--CHEAP. It
> is difficult finding the persons who would be most interested, perhaps you
> know somebody? Certainly some of the OMG members could be interested.
>
> PS: Know any distributed object environments for Solaris?
>
>
>

Hi Peter,

Thanks for reading my views in the "Bringing" paper (javacorb.html). As I'm sure
you could tell, it's pretty antiquated by now - written last spring, which in
web terms is a few centuries :-). There's a more modernized view of the same
paper, which stresses Beans more and OpenDoc less, in the Jan/Feb 97
Dr Dobbs SourceBook.

Yes, I am most certainly aware of Apple's corporate churn in general, and
their OpenDoc dump decision in particular. In fact, I belong to two
mailing lists which have been discussing the relevance and impact of these
topics in great depth recently. You may want to check out the archives of:
FoRK: http://xent.w3.org/FoRK-archive/winter96
dist-obj: http://www.infospheres.caltech.edu/mailing_lists/dist-obj/distobjgroup.html

In particular, you may want to see the discussions of Quilt (a followon to Bento)
on FoRK.

As to your proposed alternatives for how to proceed with OpenDoc in a post-Apple
scenario, I must say that I lean heavily towards a flavour of 1) as well. In fact,
I had quite long ago (last summer, I would guess) moved my interest focus away
from OpenDoc towards
JavaBeans. What I would like to see, and I think will happen, is that some of the
better concepts and ideas from OpenDoc make their way into Beans. Not
necessarily a code port, as I think you intimate, but something higher level
than that. I think Beans is the natural successor to OpenDoc.

As to OMG or ORB vendors getting interested in funding OpenDoc: not very likely
would be my call. I think OMG & CORBA have a survival fight of their own
ahead of themselves - they have to tone down their expectations that
we will all live in a CORBA world, and more realistically expect to be
an (important) source of ideas to the really dominant architecture (currently
being duked out between Java and Microsoft). My trusty informant on such matters
(Mark Baker) tells me
that OMG is actively looking for what to do for their business object frameworks
with the cancellation of OpenDoc, with the likely outcome probably a move
towards Beans.

As to your quest to "get the word out" on OpenDoc's availability, I hope
that by posting this response into the mailing lists I'll help you in
your publicity efforts, although I think the folks on these lists pretty
much know the score here already :-).

As to the PS. Um, unless I'm missing something obvious, Java runs very happily
(most happily?) on Solaris. Doesn't Beans count as a d-o environment
here? I think it's what Sun would love you to use on Solaris.
Best of all, you get 'write once, run anywhere' behaviour too, so
you can do your thing on Solaris or anywhere else you like....

Regards,
Ron.