[e2e] It's all my fault

Durga Prasad Pandey dpsmiles at MIT.EDU
Wed May 16 18:39:01 PDT 2007


If your intention was to be purely provocative, from what you have  
written below, you sure did succeed. Being a network engineer, you  
cannot be unaware of the tremendous impact academic research has, and  
will continue to have in every field, and most strikingly in  
networking. Would you consider the Internet to be one of the greatest  
inventions of the last century? (In my opinion its number 1, and I'd  
be surprised if it wasn't one of the top inventions in your list  
too). Now if you are really unaware of the academia's contributions  
to the creation and design of the Internet, and of the "DC folks"  
support and participation that made it possible, may I present  a  
recommended reading for you:


Further, let me be bold enough to propose an analogy. The role of of  
genetic mutations in evolution is widely recognized. Similar is the  
role of new ideas in science and technology, that you so happily seem  
to ridicule. You might argue that a good percentage of what gets  
published doesn't get implemented, especially immediately. Yes, there  
are papers whose ideas are just not very good, or practical. But what  
does end up working out and making a difference could look very  
"Platonic technology"ish in the short term. Lots of wiser and more  
experienced people on this forum and elsewhere will tell you how  
their ideas were ridiculed at first, and then turned out to be  
staggering successes.

Regarding source routing, may I recommend Clark et al's Tussle paper  
if you haven't read it yet: http://www.sigcomm.org/sigcomm2002/papers/ 
tussle.pdf where they make a really interesting analysis of how the  
Internet represents a constant tussle among different stakeholders.  
One of the examples they quote is that of source routing. Source  
routing represents choice for a customer of routing their packets  
through providers they like. It doesn't seem to make economic sense  
for ISPs under current business models, especially when the consumer  
is a residential user - that situation and model might change in the  
future. Perhaps source routing,  as currently defined needs to  
change. But if you're arguing for ISPs constraining any choice the  
user has with respect to how their data flows depending on their  
short term economic thinking, I think you aren't making a good  
argument. The basis of economics is the concept of customer and  
utility, and technologists who ignore this might well get swept away.

Yes, your position is quite explicit. :)


On May 16, 2007, at 6:38 PM, Vadim Antonov wrote:

> On Wed, 16 May 2007, Ken Calvert wrote:
>> You are right, one can contemplate.  But as soon as one
>> starts talking about it, lots of folks with a firm grasp of
>> the status quo start saying it'll never work, there's no
>> market for it, etc.
> Here you see the division between engineers who design and build  
> things
> which make economic sense and academics who think of pure Platonic
> technology existing in economic vacuum.
> And, yes, network engineers do have to deal with the spillover of bad
> ideas from academia - things which would never get into protocols and
> designs if somebody took trouble to evaluate their economic  
> feasibility
> before sneaking them into standards.
>> The only way to overcome that is to build something and use it  
>> (what I
>> think Reed was talking about).
> Yep. *Build* something. To do that you need way more than a neat  
> idea -
> you need capital, you need business plan, you need customers who  
> actually
> wish to buy the product. You need to spend years of your life  
> working like
> hell.  And if you were wrong, you don't get anything for your  
> troubles.
> Or you may convince some bureaucrats in DC to give you lots of  
> money they
> have taken from us under the threat of jailtime and violence so you  
> can
> play with your pet idea.  A lot of people resent that, you know?
> As long as you want to go the first route I can only wish the best  
> luck,
> and offer some advice - do not talk much about the ideas you intend to
> implement, there's a lot of sharks in this water.
> If your plan is to organize another federally funded playpen, I  
> (and other
> engineering people) will do everything to shoot the proposed neat idea
> down, before it becomes another excuse for looting more from us.
> I'm all for discussing various neat tricks and gimmicks as a pure  
> mental
> excercise, contemplating possibilities, and such.  But I draw the line
> when someone starts to talk about implementing his ideas at my  
> expense. I
> have neat ideas of my own - and wish to spend my resources on  
> playing with
> them.
> Did I make my position clear?
> --vadim
> PS. Sorry for the off-topic.

More information about the end2end-interest mailing list