A Baby/Bathwater Kind of Thing

The Washington Post’s lead gadget writer, Rob Pegoraro, graced us with the benefit of his expertise yesterday in a column on the FCC’s Open Internet order (FCC votes for a half-measure on net neutrality;) In short, he’s not happy.  His FCC post is actually more closely related to the frustrations expressed in a preceding post reviewing the video calling services provided by a game controller, the Xbox Video Kinect, than he apparently realizes.

Pegoraro fails to find satisfaction with video calls over the Kinect:

The major disappointment here was the horrendous quality of the video, considering that I had about 5 million bits per second of upstream bandwidth at each end of the test. The footage looked unmistakably pixelated. 

He notes that fellow gamers experience video problems as well. He might try video Skype as an alternative, but alas, that allegedly peer-to-peer service is down today due to a system-wide software problem with the “super-nodes” that run its directory service. When Skype’s on vacation and Kinect’s too pixelated to be of any use, the would-be video caller might fall back on the Cisco Umi system built on a conferencing-oriented network, but Pegoraro finds the price of $21.95/month service “completely out of whack for the mass market,” which would apparently prefer Skype’s blank screen or Kinect’s ghostly images.

Cisco’s Umi outperforms its competitors in large part because it’s built on a premium network instead of a consumer-grade system that cuts every corner in pursuit of a great experience with Netflix and the Web. Even a 5 Mbps upstream connection isn’t enough to provide a high-quality end-to-end experience on a download-oriented system.  Netflix works great on consumer broadband systems because it only has to download bits, which it can that from a nearby CDN server, but CDNs don’t solve the video calling problem. We need premium networking services such as the one that Cisco provides to do that.

Premium networking can be provided over a consumer-oriented broadband circuit easily enough: Triple play services provide voice and video streams with higher quality service than end-to-end Internet access, after all. Networks can’t boost the quality of everything any more than the people of Lake Woebegone can make all the children  above average, so wouldn’t it be a boon to consumers if we could designate a portion of our upstream bandwidth for special video call treatment when we want to get on the Kinect and call a friend? A temporary, targeted boost just for the packets that need it seems like just the ticket to improve quality without breaking the bank.

Well no, not according to Pegoraro, who rejects this sort of service with the Extra Snark that’s the hallmark of confused criticism:

The FCC’s rules would prohibit wire-line broadband Internet providers…from “unreasonable discrimination” against particular sites but would allow them to charge sites more for faster delivery…(Note that wired and wireless providers alike generally testify that they have no plans to discriminate against anybody online; they merely want the legal flexibility to continue not doing such a thing.)

Let’s try to kinect the dots for Mr. Pegoraro: You’ll generally run into poor video call quality at times of the day when the path through the network of networks that connects you and your video buddy is heavily loaded. Today’s Internet was designed for web surfing, and lacks a robust, reliable technical system for moving video packets in both directions without random delays and relatively high packet loss. Such features of the Internet’s technical system don’t adversely affect web surfing, but they’re hell for video. Premium video streaming services overcome this problem through a system that bypasses most of the Internet, but their system only works for canned content. Premium video calling services overlay the Internet with a set of extra-capacity circuits dedicated to video calling.

We know how to improve the Internet so that it can handle a mixed load of web and video calling traffic at a modest price, but we need the engineering freedom to use “discrimination” to do so. The Kinect can only work as well as the Umi for a lower price if the FCC green-lights these systems. Throwing bandwidth at the problem isn’t enough; as Pegoraro found out, even 5 Mbps for an application that needs less than one doesn’t cut it. It’s not the bandwidth, it’s the quality.

The trouble with democracy is that we get the kind of rules from our government that we ask for; if our gadget lovers cum policy advocates continue to denounce the technical and economic systems that enable gadgets to work well, we’re not going to have a happy fun time on the Internet. 

Be careful what you ask for, you just might get it.

 

Print Friendly

About the author

Richard Bennett is an ITIF Senior Research Fellow specializing in broadband networking and Internet policy. He has a 30 year background in network engineering and standards. He was vice-chair of the IEEE 802.3 task group that devised the original Ethernet over Twisted Pair standard, and has contributed to Wi-Fi standards for fifteen years. He was active in OSI, the instigator of RFC 1001, and founder, along with Bob Metcalfe, of the Open Token Foundation, the first network industry alliance to operate an interoperability lab. He has worked for leading applied research labs, where portions of his work were underwritten by DARPA. Richard is also the inventor of four networking patents and a member of the BITAG Technical Working Group.