> Supporting an XML text based archival view of protocol messages
> would be a lot easier if protocol designers could help themselves
> to a useable binary XML encoding.FWIW. The transport mechanism has a great impact on the
viability of verbose formats. Some protocols *strongly* favor
getting all the data out in one packet, for example, or
reward packing everything into 512 byte blocks.In general, where parsing performance isn't *the* crucial
thing, text wins. In low-level protocols that are senstive
to data sizes, binary wins.
As always it comes down to Use Cases, and
- if readability by Humas requirement is important / prioritized then
Text based ML wins.
- If performance is important that usually Binary ML wins.
In the Financial world there are numerous Use Cases in which
relatively small update are sent in broadcast, multicasts, reliable
or not.
Here we have single writer - multiple readers where performance is
crucial
so switching from TextML to BinaryML makes a lot of sense.
Another 2 cents:
In communication between two application using generated low level
encoder and decoders
the actual messages are to a 99.99...% degree never viewed by humans
so BinaryML
should make sense here too.
/anders