[
Lists Home |
Date Index |
Thread Index
]
Which is good stuff, Rich, and useful.
Do we need one more ring to rule them all? I question that it has to
come down to one binary type or will. XML The Text already
has a very large network effect. Coupling a set of smaller
network effects to that (sub nets of users) might actually
strengthen the overall system. It may not be necessary for
the W3C to do more than offer advice and classification or
best practice to that.
If there is a clear winner, great. If not, it isn't wasted
effort as long as the classes can be easily matched to the jobs.
len
From: Rich Salz [mailto:rsalz@datapower.com]
> The W3C XML Binary Characterization Working Group has the unenviable
> task of helping W3C decide whether we should be in the business of
> specifying some sort of efficient way to interchange XML that's
> different from gzip that HTTP can already use.
I think they (not really we, I don't do much:) are actually doing
much better then that. There are three major documents:
Use cases -- who would benefit from binary xml and why,
and what restrictions they place on the format
(e.g., must both sides know the schema?)
Properties -- characteristics to consider for an encoding
(e.g., can you sign it, do random-access, etc)
Measurements -- how to apply the above two and get a weighted
score
Anyone interested in the *concept* of binary XML should really take a look
at these documents.
|
