Geek Stuff

I just read an interesting article over on Ars Technica summerizing the debacle the FCC has gotten itself into when it asked for feedback from the public on how to define “Broadband”. You see, up until recently broadband was defined as any internet service with a bandwidth of 768Kbps or better. This made perfect sense. Anyone can understand that. Now the FCC decided to expand on that definition and asked two basic questions: 1) How should the speeds be determined (actual or advertised) and 2) Should the FCC take an “application-based” approach (meaning features not speed).

Here’s my take:

1. Broadband should be defined as a bandwidth range. Lets say a minimum average bit-rate. If ‘X’ service can provide ‘Y’-Kbps averaged out over a month, then it is a broadband service. There could even be classes of broadband. An example would be class-1 = 768k, class-2 = 1.5m, class-3 = 5m, and so on. That would still allow providers to use the term “Broadband” for differing levels of service, and also provide a reliable and easy to compare reference for consumers.
2. Defining “Broadband” as a set of features or capabilities will royally screw the consumer. Here’s why: First, it creates a loophole for ISPs to block services that don’t fall neatly into the FCC’s feature list. Second, it could allow ISPs to cap speeds to the bare minimum necessary for the said listed services. Third, it would open the doors for providers to tack-on “premium” capabilities to their broadband service. Hey, you’ve got Web and Email… Oh you want bit-torrent? that’s extra. Oh you need usenet? that’s extra. You can see where I’m going with this. Imagine the FCC made this list of basic capabilities. What happens as usage on the internet changes in the future?

I think that having the FCC define broadband as anything other than an actual bandwidth capability is asking for trouble. It’s not accurate for an ISP to be considered a broadband provider based on their advertised speeds. It would be much more fair to have the determination of broadband speeds be based upon an average of what the provider’s network actually supplied. Even more important is keeping the definition of broadband limited to bandwidth. Imagine we defined our roadways based upon their features. I’ve been on 2-lane highways and ridden 85MPH the whole way. I’ve also been on 4-lane city streets that have lights at every intersection and not gotten over 35MPH. Imagine that city street could be called a highway just because of the number of lanes it had.

I was lucky enough to be one of the original testers of cable-based broadband when it was first introduced to my neighborhood back in 1996 or so. Since that time – 13-years now – I have never once experienced an actual download that was equal to the advertised rate. As a tech person I get asked a lot of questions about one service versus another and what the “real” speed on that service is. It always leads to a long conversation about the differences between theoretical limits and RealWorld™ limits. It should be easier than that.

Ultimately I agree with the author, Matthew Lasar, in that broadband should be defined as a data rate and not a set of features.

Here’s the original article: Big cable to FCC: don’t define broadband by its actual speed – Ars Technica.