This page and the listed discussion groups about Harvard are not affiliated with or sponsored by Harvard University or the Harvard Alumni Association.

vendredi, juillet 31, 2009

Why keep it twice? - search engineering esperance your-oo

Leapfrog Google? The search industry is bent down. Imagine next week
the Yahoo to Redmond flights where heads hit the seat upright pillows
dreaming of this upheaval. You are about to receive a magic vessel
infusion. When you lift your head from this email you'll have the
power to start your own company with two 'o's in its name. What part
of Africa, the Ukraine, India, Europe, China or the US will fuel the
your-oo headquarters? You'll be lauded for your green solution. No
data center near Grand Coulee or wave energy farm was needed. Why keep
the data twice?

The energy to store the data in your search results was paid for once.
You store it on your server we can access now. Why pay to copy and
keep a reflection of what is on your server in other locations too?
And what about latency? The copies are never accurate on demand.

So. The answer? Build an open source or proprietary search project
that leaves a stub of your data indexed on your server. It is your
data not theirs. Each server advertises. Create a billing system that
pays server providers where this footprint is above a certain onerous
threshold. Your data center costs for your-oo reduce to staging areas
around the world that temporarily hold indices and common high volume
results. The Redmond-Mountain View business models are disrupted.
You're a genius. Google survives with unbridled aspirations reduced.
They own the translation layer to the data on your server. Redmond
consortium dies replaced by you. Valid revolution starts as impossible
fantasy.

Perry
perry.gregg@post.harvard.edu
Cell: (510) 684-4152
Skype: perry.gregg

--
Sent from my mobile device