On Thu, 2010-02-18 at 23:38 -0700, ara.t.howard wrote: > On Thu, Feb 18, 2010 at 23:33, Brad Fitzpatrick <[email protected]> wrote: > > Those aren't serious questions. Let's keep this list serious and on-topic. > > actually - in the two quite serious presentations i've been to on the > subject the question has come up and stumped the presenter completely > cold.
I think that it's a legitimate, pointed, and important question. My main reservation is that I think a better question is why NNTP wasn't the right foundation to build upon. The key element of SMTP is routing towards a given recipient. For peering-based broadcasting, building over NNTP seems to me to have been the more graceful option. > still the answer eludes the general programming public not versed on > this list. if you do not care, that it fine, but be aware that some > exceptional engineers have precisely the same questions. The best I can figure out so far is that it's mostly a generational thing. Ontogeny recapitulates phyoleny, sorta. They have to reinvent for themselves, apparently. That, combined with the fact that it became *easier* to express new protocol ideas as layers over HTTP because of the big burst in the development of scripting language HTTP libraries initiated by the dot-com boom. Knock out some python code, submit a standard, so to speak. Now, to be sure - engineering is an art that is supposed to respond to economic conditions and so the critics of your questions have a valid concern. *Maybe* the fact it was accidentally easier to express new protocol ideas this way justifies them...though I doubt it. The kids these days are mostly just blowing up the shoulders of giants rather than climbing up onto them, in areas like this. Sincerely: it makes me sad to witness. Especially when someone like Fitzpatrick with his cultish theory of poisonous people declares your very good question to be "not serious". Regards, -t > regards. >
