Question for you smarty-pants out there! :)

[warcrow]

**THE FOLLOWING IS A MBA PROJECT FOR A CLOSE FRIEND**

Let's say you had some software that would act as some software that would/could act as a buffer for on-demand video SD/HD. This software was designed to dynamically adjust the size of data packets, the size of the buffer, and network connectivity speed to determine the best Q.O.S. for the user/client.

What, in your opinion, is the best way to make money with this technology? My proposal is to sell this type of tech (it has to be good, but this is all hypothetical so lets just say this is really really great software), and sell it to Comcast/Roadrunner/Cox as a better solution to whatever they're using. But I think this is too obvious. Where else could something like this be utilized?

Do you think NASA or someone in the private sector could utilize this (rememeber, this software is really really incredible -it's all hypothetical!) with video transmissions from space?

I'm curious what everyones suggestion will be.

 

[TuxDave]

Crosspost?

http://forums.anandtech.com/messageview...atid=38&threadid=2010828&enterthread=y

 

[warcrow]

Your detective skills are quite impressive young padwan!

 

[esun]

Get VC support to set up an Internet video site, sell it off once some huge company decides to buy you out.

 

[gsellis]

I think you and your bud need to go to NAB to research this

 

[engineereeyore]

Not sure I see a whole lot of point in the dynamics resizing. It seems awful complicated and unnecessary. Buffer sizes are the size they are because they've been shown to provide optimal performance at that size. So what are you supposed to really gain from this? The packets are UDP, so there is not resending of lost data. You increase the packet size and drop a packet, and that makes a big difference.

Question two. Are you talking about streaming video, or HD cable video? Streaming video is the only one that uses packets, as far as I know.

EDIT: Here's a paper that was written on it. Paper Since most routers keep information such as this already, I think the extra software would be overkill.

 

[esun]

I think it's the spirit of the idea that's important, not actually whether it works in real life. Maybe they designed an ASIC that does all this special video stuff very quickly in a manner better than any existing network interface does it.

 

[DrPizza]

I'm not sure how your hypothetical software is supposed to do anything for NASA - suppose you have a probe that's 20 minutes away (by the speed of light) - how's it supposed to calculate what packet size is ideal; or rather, how is this going to improve over simply determining this well in advance (before it's launched?) I can't think of a market other than comcast/etc.