[
Lists Home |
Date Index |
Thread Index
]
Claude L "Len" Bullard top-posted:
> Good enough. Can I get that without the XML syntax?
I have no idea what "that" you're talking about,
but assuming that by "that" you mean this:
> [Joe English]
>> XML's basic building blocks of elements, attributes,
>> and text are flexible enough to accomodate a much
>> broader range of data. Only a few things fit in a
>> regular table, but a lot of things can fit in a tree.
then sure: you can use S-expressions. As any Lisp fan
will tell you, S-expressions can do everything that XML
can do, can do it better, and moreover did it first 40
years ago.
(FWIW I don't fully believe that myself, but that's
what all the Lisp fans say. Certainly S-exprs and the
lowly CONS cell are just as good a foundation as anything
XML has to offer, and S-expression syntax has been used in
this problem domain before. Dunno why the idea hasn't gained
as much traction as XML has.)
But I get the impression that you're looking for something
*binary*, 'cause, well, everyone knows that *binary* is
*faster*. I don't believe that either, but if that's what
you want, take your pick: ASN.1 BER/DER/PER/*ER, ONC XDR, COM,
DCOM, any number of CORBA implementations, or an even greater
number of ad-hoc all-purpose binary serialization formats.
Or invent your own; everyone else does. (The reason why none
of these has gained as much traction as XML is left as a
(one-star) exercise to the reader. Multiple answers accepted.)
OTOH, if you buy the premise that elements, attributes, and text
are in some way the "right" set of building blocks -- well,
then why do you want that without the XML syntax? That's like
saying you want S-expressions except without all the parentheses.
--Joe English
jenglish@flightlab.com
|
