From: Jules Richardson <julesrichardsonuk at yahoo.co.uk>
It's a shame that there wasn't a widely-adopted
data transfer protocol that
was truly bi-directional.
True, though there are and have been many good not-so-widely adopted
solutions.
Both FTP and HTTP largely revolve around moving data
in one direction only. Even given the predominant client/server model of the
time, it's a shame that the "main" protocols weren't designed with the
foresight that one day we might want to move just as much stuff *to*
a site as we were taking *from* it.
Client/server is still essentially the model for the vast majority of
internet use: it's almost all asymmetric. As in, depending on the site, 100
to 1000 times or more as much data going one way than the other. Even
profiling sites like YouTube, where people actually do upload large amounts
of data, individual clients download far more.
HTTP was designed for the common case in it's problem space. It shouldn't
be surprising that it's not optimal for other cases.