RE: "Binary XML" proposals

[Miles Sabin <MSabin@interx.com>]

David Brownell wrote,
> Miles Sabin wrote,
> > David Brownell wrote,
> > > Binary formats are bad because they tend towards being 
> > > proprietary, and that's the last thing that should happen to 
> > > the world's next "intellectual commons".
> > 
> > True in the document world, perhaps. But not so obviously true
> > in the protocol world. For example, DNS question and answer 
> > payloads are an example of an open, structured, binary format.
>
> I'm fully aware.  But you also ought to consider exactly how
> open and extensible DNS is -- by seeing whether you can
> get to two hands when you count implementations (BIND,
> and hardly any other servers), and extensions (rare).

If you mean particular identifiable implementations, then no,
not unless I'm allowed to count BINDs 4, 8, and 9 separately.
But there are many, many more: client only implementations
embedded in OSs, web browsers and other network clients and 
servers; and specialized servers developed and used by various 
segments of the network infrastructure community (CDN, firewall, 
NAT).

Interoperability isn't simply due to a lack of diversity.

> Basically, every binary RPC protocol I've ever seen has been
> converted, sooner or later, into a conduit for proprietary
> platforms.  Fragmenting a previously-unified (XML=text)
> world by creating a binary variant seems a fine start, for any
> organizations wanting to head that direction.
<snip/>
> There's also the "out of sight, out of mind" issue.  Once things
> get binary, the number of people who can detect mistakes (much 
> less shenanigans!) declines by orders of magnitude. That means 
> that interop becomes more fragile; which also pushes things 
> towards proprietary behaviors/bugsets.

I understand your concern, and I share it. But I think you're
overestimating the extent to which text is a defence. Beyond the
denizens of this list, what proportion of the XML aware would
realize that there was something wrong with,

  <evilvendor:binary chunksize="16"
  >slkdjfweuoiwurjs</evilvendor:binary>

and would reject it in favour of the base64 equivalent if all
they ever had to deal with were evilvendors parsers?

Cheers,


Miles

-- 
Miles Sabin                               InterX
Internet Systems Architect                5/6 Glenthorne Mews
+44 (0)20 8817 4030                       London, W6 0LJ, England
msabin@interx.com                         http://www.interx.com/