The Protocol [was: Syntax] is API Fallacy

[Ingargiola, Tito]

Hi,

	[ Jonathan Borden states: ]
>        You are suggesting using the DOM to model your distributed objects.
> This is a bad idea. You are correct in your conclusion that this is a bad
> use of granularity, though I've never tested this out myself because it is
> evident by a quick analysis of round trips. 
> 
agreed

> What I had suggested was to use
> XML to marshal objects by value (MBV) which specifically gets around the
> network roundtrip killer.
> 
Actually, the idea of using XML as the means of achieving MBV doesn't seem
too good to me.  

Why?  First, one of the great attractions of a distributed object system is
precisely that I'm not typically zapping objects about the network, but am
rather dealing with the more interesting objects in different address spaces
by (remote) reference.  So, as an an application developer using corba, I
really don't care too much how my corba implementation (as provided by Iona
or whomever) does its business (e.g., marshalling).  I just care that it
does it correctly and, hopefully efficiently.

The question then becomes: "does a corba implementation developer (e.g.,
someone writing orbix at Iona) want to use xml as the underlying
representation for the marshalling of objects?"  I still think the answer to
this question is likely "no."  It'll likely be less efficient than other
such implementations, and adds little value otherwise.  

Finally, I'm not convinced that a corba/RPC-style of remote method
invocation is typically well suited to web-oriented interactions -- most
things on the web can be treated naturally and efficiently as streams.  Why
add the extra weight (which seems to buy me little) of corba to a domain in
which it seems a mis-fit?


>     Furthermore I am suggesting that by enabling use of web native
> protocols
> under currently available distributed object systems we can move toward a
> better integration of distributed objects and the web.
> 
I certainly share your enthusiasm for a better integration between
distributed objects and the web.  My experience thus far tells me that
layering corba (or some other rpc -like mechanism) on top of more ubiquitous
streaming protocols costs a lot and yields little benefit.  However, efforts
like http-ng offer the promise of such an integration without carrying some
of the baggage that might come with trying to shoehorn a big heavy thing
like corba on top of something so umm svelte ;-> as http and friends.
Naturally, your mileage may vary...  Best regards,

	Tito.


xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)