> I believe it is the DTD (plus extensions) that makes the XML approach
> superior to the binary encoding. Without the DTD, XML is still slightly
> better than the binary encoding, but probably not enough better to win more
> converts.
Many of us believe that a special-purpose binary encoding (eg, IPP)
is *highly* undesirable under almost *any* circumstance. Recall
that at the June '96 IPP meeting (Nashua), a majority of people were
dead-set against a binary encoding of *any* kind, much less a special-
purpose one designed for a single application (ie, IPP).
Since XML is essentially a text-based encoding, I'd bet we'd see
a lot more "converts" than you might think.
...jay
PS: I've posted this reply (and your complete message) to the IPP
list, as I suspected you forgot to cc: the IPP DL. My apologies
if this was not your intention. ;-)
----------------------------------------------------------------------
-- JK Martin | Email: jkm@underscore.com --
-- Underscore, Inc. | Voice: (603) 889-7000 --
-- 41C Sagamore Park Road | Fax: (603) 889-2699 --
-- Hudson, NH 03051-4915 | Web: http://www.underscore.com --
----------------------------------------------------------------------
Robert Herriot wrote:
>
> I believe it is the DTD (plus extensions) that makes the XML approach
> superior to the binary encoding. Without the DTD, XML is still slightly
> better than the binary encoding, but probably not enough better to win more
> converts. For the DTD to work, it really needs the Schema rules from the
> "XML Data" document plus a few enhancements that I will propose. With such,
> I believe that it is possible to define something that addresses all (or
> nearly all) of what Section 15.3 in the model document does. I hope to send
> out some examples in the next few days. Note, I still don't think that
> printer vendors would implement a general parser with DTD embedded. Rather
> they would use the DTD to direct their hand coding of their parser and
> validator.
>
> Bob Herriot
>
> At 03:18 PM 2/13/98 , you wrote:
> >Paul,
> >
> >You've specified that no DTD is required, but that an
> >implementation could use a DTD as it saw fit, etc.
> >
> >How then is the "official" specification of the IPP
> >XML text supposed to be standardized? I would have
> >thought that the use of a DTD would be good for this,
> >even though *use* of the DTD by a client or server is
> >not required for protocol interaction.
> >
> >Note that I am assuming a DTD is itself a useful document,
> >not being very well versed in XML/DTD stuff.
> >
> > ...jay
> >
> >----------------------------------------------------------------------
> >-- JK Martin | Email: jkm@underscore.com --
> >-- Underscore, Inc. | Voice: (603) 889-7000 --
> >-- 41C Sagamore Park Road | Fax: (603) 889-2699 --
> >-- Hudson, NH 03051-4915 | Web:
> <http://www.underscore.com/>http://www.underscore.com --
> >----------------------------------------------------------------------
> >