Monday, April 25, 2011

LUS Fiber — GigaFest Announcements (Updated)

LUS Fiber kicked off "Gigafest" today with an announcement that LUS Fiber was upgrading our systems connection to the internet from 1 gig to 10 gigabits/second. Huval says that this is several years earlier than LUS had planned to make that move and the early ten-fold increase is a testament to its users finding ways to make use of the big pipe LUS has made available.

You need to stop and ruminate on that for a minute. LUS is bragging on the fact that its users are using a lot of bandwidth. They are crowing about making a 10x increase in the size of the connection to the larger internet that they have to buy to sustain their customers' usage. Now you might think it nice but not all that remarkable that a business should be proud that their customers find their product so useful that they have to upgrade their supply system to cope with demand. If you think like that you are still operating in the regular, competitive, "free enterprise" part of the american market, NOT the telecomm segment. In duopoly-land providers from AT&T and Cox to Verizon and Time-Warner are constantly complaining that their users are trying to use too much of "their" bandwidth and insisting that their customers need to be throttled down and capped at miniscule amounts to make sure that the "bandwidth hogs" don't ruin everyone's experience. It is downright refreshing to hear from a provider who is happy to upgrade their system to meet demand—and who doesn't accompany any upgrade to the network with some sort price hike and incessant whining.

Welcome to the land of community-owned broadband.

And that, the advantages of a community-owned network, was one of the themes of today's kick-off presentation to the media. The others, as I saw it, were the impact on businesses and the role of latency.

Community Ownership
Both Mayor-President Durel and LUS head Huval emphasized the advantages of a locally owned network—but in characteristically different ways. Durel made clear, during his brief remarks, that his emphasis had always been on the potential for economic development that he saw in an LUS Fiber network. He saw the examples of business usage that were highlighted in the presentations as a realization of his hopes. Huval, as you might expect of a utility head, emphasized that LUS was keeping its covenant with the community by providing fast, cheap and reliable services—underlining that by saying it was true that Lafayette has "the fastest, cheapest internet in the US." That's a pretty bold claim and on a megabits per dollar basis I think that's true. LUS' tiers are the best values I've been able to find for the speed and capacity they represent. You won't find a cheaper 50 meg symmetrical connection (or even an asymmetrical one I think) anywhere in the US. (You can get onto the internet for cheaper—but LUS doesn't sell anything less than a 10/10 meg symmetrical connection. Real broadband. Those that are cheaper are much less capable, asymmetrical or capped at some ridiculously low monthly maximum.)

Business Uses
The real focus of the GigaFest event is on business recruitment—LUS is apparently starting a major push to recruit more of Lafayette's small and medium size businesses. Announcements and promotions have been going on in the background for a week or more. (Apparently Cox thinks that's what LUS is up to also: Sunday's Advertiser carried a prominent classified ad that solicited for salespersons to work in their small business sections—with or without experience. The troops are massing on both sides of this battle.)

The presentation includes a slick, locally produced video that highlighted local businesses that are making good use of LUS Fiber. (LUS should make that video available on its website; it is convincing.) There were folks who loved how they could work from home, a coffee shop case, an application in medical records and medicine, a web design house, and a church that does massive video uploading. The recurring theme was that the speed of LUS Fiber made it much easier to do their job. Some of the background info that Alcatel-Lucent provided in its flashy surround environment of many (84?) monitors on all four walls gave some technical context as to why these users found their experience on LUS so superior. Sheer speed is part of the explanation; symmetrical upload and download was another. But the hard-to-explain but oh-so-important part was Latency.

Latency has an involved technical explanation. But what is important to understand is that your perceived speed, the speed that actually matters to you, is composed of both throughput—what we usually call "speed"—and latency. Latency involves the time it takes to make or confirm a successful communication. A call and response: "Are you there?—Yes I am here." Only once that connection is made does throughput size (bandwidth) comes into packet-based systems each packet's success is bracketed by such a call and response. If it fails the information packet is resent. Latency and throughput are conceptually separable. You can have a "skinny," slow pipe with very quick response or latency on one network and a "fast," big pipe with very, very slow response-latency times. Depending on how you are using those network either type may be perceived as slow or fast. Examples: a video stream that uses big packets will seem slow over the skinny pipe, no matter how fast the latency. But a game or a home working session that relies on many quick back and forth connections and so uses many small packets will stutter and feel slow no matter how fat the pipe if the latency is high.

We didn't used to have to explain this stuff. That was because networks were getting both "fatter" bandwidth and lower, quicker latency at the same time. So it was easier to just peg it all on bandwidth or speed...and sell the public a 768 kbps or a 15 mbps package—bigger was better. But it was always a misleading sort of shorthand and now things have changed, at least in Lafayette. The old copper-based DSL and Coax that the telephone and cable companies are reusing to provide us with data services are reaching their limits—and those limits are different for speed/bandwidth than they are for latency. Latency is much more resistant to improvement and that fact is beginning to show as bandwidth numbers are improved without improving latency. The fellow from Alcatel noted coax cable introduces a latency of around 45 milliseconds as it exits the first neighborhood node. But a fiber to the home network has much lower latency and you can count on only 10-15 ms of lag to be introduced into the local network. So even for the same size 30 meg connection a fiber-based network will have lower latency...and working from home or gaming sessions will feel much smoother and quicker. Channels change quicker on IP-based video systems. Your connection to netflix is smoother and your interface connections feel a lot more responsive.

The best of all possible worlds is, of course, to have a system with both a fat pipe/big bandwidth/high speed and low latency—you want the interface to Netflix to feel smooth and you want the big video packets to flow down a nice fat pipe... What the man from Alcatel was trying to say is that here in Lafayette we have such an ideal system. There just aren't very many places where you can get both kinds of speed in one package. But we can here

End Notes
You can check out LUS' (newly redesigned!) bit on the Gigafest event. (You can register through LUS to attend one of the demonstrations.) And you can check out channel 10's coverage. The media was there in force, so there will have likely been stuff on the local news channels and it will appear in print media in the morning. More here as it appears.

Update 4/26/11: The Advertiser has an overview article from the pen of the new general business beat reporter. There are two infelicities in the article that the tech savvy Lafayette reader will note: the translation of bits into bytes (telecom uses bits; storage bytes, the two are seldom translated into each other) and the claim in the final paragraph that Huval had announced the cost of the project in December last (the cost was established long ago, she's probably trying to reference the completion of the project). The Advocate has a nice, large picture with a paragraph-long cutline on the front page of the business section; unfortunately that's not available online.

You can also take a look at Alcatel-Lucent's press release which includes the following quotable quote from Joey Durel: "We are now only one of a handful of communities in the world with this level of accessible Internet capacity – and only one of the few in the world to have a system like this which is owned by its citizens. That is the differentiating factor – the success of LUS Fiber is passed on to and enjoyed by all Lafayette's citizens."

Friday, April 01, 2011

Why no Google Fiber for Baton Rouge? (Updated)

The folks in Baton Rouge are probably asking themselves why they didn't get Google's gigabit fiber network. After all for a while the Facebook page "Bring Google Fiber to Baton Rouge" had more fans than any in the country and there seemed a pretty large groundswell of support. There was a heavily rewritten AP article in the Advocate that interviewed a BR Chamber officer and recounted the history of local public involvement. (Unfortunately not online check p. B-6 of 3/31/11 paper.) It all seemed so hopeful. The Chamber held out hope that Baton Rouge might get in on the second round.

I don't think so. At least not until we put our own house in order.

Here's at least one reason that Google avoided Louisiana:

See Kansas? It's Green. Texas and Arkansas are Red. Louisiana is a sickly Orange. Google is only going to green states. This map has nothing to do with solar energy or recycling. The green denotes a place where there are no state-wide legal barriers to a community building and owning its own fiber-optic network. Red states absolutely forbid it. Louisiana is among those who are hostile but do not completely outlaw the idea. (Witness Lafayette's ongoing battle.)

Google wants to strike out and do something truly different. They are frank about thinking the Cox's and AT&T's of this nation haven't done a good job and that local communities can do better and should be helped to do so. Google has no reason in the world to go to a state that tries to make the sort of community involvement they count on illegal.

They aren't coming to Louisiana until the "(un)fair competition act" is abolished. If Baton Rouge (or New Orleans, or Shreveport or Bossier, or any of the other Louisiana cities that applied) want to have a shot at Google's second round the first thing they have to do is get their own house in order.

Repeal Louisiana's (un)fair competition act...

(Check out the great map at from which I grabbed the above illustration. It chock full of valuable, if depressing, information.)

Update 4/1/11: Stacy Higginbotham, tech journalist extraordinaire over at GigaOm, covers the Texas version of this story. Apparently Austin had a very credible, widely supported effort to get their city picked. The local organizer thinks:
“Austin caught their eye for all the right reasons, and we had support at the highest levels with the involvement of the mayor and the city manager, but given the Texas limitations on municipalities getting involved in network, there was only so far we could go,” Rosenthal said. “So I look at the Texas Legislature, because they really put us in a box with regard to Google, and every response the city gave had to be measured within that box.”
Yup, I expect he's exactly right. Texas forbids muni networks. Google is doing this to encourage muni networks. The are NOT going to pick a city in a state with lousy laws that forbid what they are trying to get other municipalities to do. That's only common sense.

Update 4/4/11: Take a look at what the paper in Kansas City, Kansas thinks were the reasons that its city made the cut. The story, understandably, tends to focus on drama and secrecy but there are some very interesting nuggets in there about the underlying factors that might have favored KCK once the first cuts were made.

Update 4/4/11, 8:15 PM: As part of the ongoing discussion in the comments I reviewed the Louisiana law constraining muni networks. There I found what I thought I remembered: The law explicitly includes the sort of public-private partnership that Google is undertaking in Kansas City. So anyone who is murmuring that Google could do a project similar to the KCK one in Louisiana simply has not read the law. You can bet that Google has. See the element of the law which defines a public-private partnership as one that must adhere to all aspects of the law at RS 45:844.47 B(3): "Through a partnership or joint venture." If Baton Rouge wants Google to consider them in the second round they'll want to repeal this law first.