joanna bujes wrote:
>
> (I thought HTTP was big because it could get you through fire walls,
> but ravi, please correct me if I'm wrong.
>
no, you are quite right -- HTTP is/was used as a fallback transport for
various applications (such as audio/video streaming), even though it was
not well-suited for them, because, as you suggest, firewall
administrators permitted HTTP into the intranet.
i was referring to the additional effect of these extremely abstracted
web based network solutions. many of these are quite heavy duty network
applications but, imho, in their object oriented/over-abstracted design,
they carry the blackbox model of the protocol stack too far. protocols
can and should be fine-tuned to particular applications (i admit i am
being a little vague here).
i use "transport protocol" in a loose sense above. HTTP is not really a
transport protocol -- its an application protocol. perhaps i should not
make this loose reference, since this is exactly what i am complaining
against: the use of HTTP as a transport protocol for all applications.
i.e., HTTP as the default and only application layer protocol -- whether
it is ready to perform such an important role is questionable (and afaik
has not been determined. one could even argue that the opposite, i.e.,
HTTP is a poorly designed application protocol, has been shown to be a
valid conclusion).
> Oh, and that "IP over XML" was hillarious.)
glad you liked it!
--ravi
