[FoRK] A Radical New Router

Stephen Williams sdw at lig.net
Wed Jul 22 15:49:23 PDT 2009


Adam L Beberg wrote:
> Bill Stoddard wrote on 7/22/2009 12:46 PM:
>> Stephen Williams wrote:
>>> This shouldn't be radical, it should have already been done.  Nice 
>>> solution to "fixing" TCP by working with it.  They carefully drop a 
>>> single packet per TCP flow to drop transmission rate without going 
>>> exponential.
>>>
>>> http://www.spectrum.ieee.org/computing/networks/a-radical-new-router/1
>>>
>>> sdw
>> I was an SNA hack in a past life and I had the exact same 'WTF' 
>> response when I read this in my hardcopy Spectrum that came in the 
>> mail a while back.   Hardly newsworthy much less 'radical'...
>
> WTF-fatigue took me long ago. People are now working on and VERY 
> excited about things I was taught (in books for god sake) in the early 
> 90's that people figured out and mastered by the 70's.
>
> Of course, industry is the same, look at Google. They make all their 
> money on AdSense which they bought from someone else, and 90% of the 
> employees - their parents were in diapers when people first used stuff 
> like Google docs.
>
> I used to pimp-slap uppity nerds with Engelbart's Mother of all 
> Demo's, but I stopped caring... let them get all excited that they 
> discovered multicore, or state caching (the excitement over caching 
> right now is mind-boggling), or a Web 2.0 version of dumb terminals, 
> or whatever, it really doesn't hurt anything.
>
I think it's OK generally, although I do get irritated frequently.  
Especially when something I was trying to do finally becomes viable, 
long after I moved on.
Many times, the addition of new technology, or new bandwidth / cpu / RAM 
/ displays / form factors make things suddenly much more useful.  Other 
times, the whole industry was hobbled by stubborn ignorance of better 
ways simply because incremental evolution from the status quo coupled 
with lack of imagination or ability to try new paradigms doomed most to 
inefficient, painful solutions.

sdw



More information about the FoRK mailing list