Archive for the ‘Connectivity’ Category

Bring Google Fiber to Boulder

Tuesday, March 23rd, 2010

I’ll get right to the punch line:  head over to http://boulderfiber.com, follow the instructions, and nominate Boulder for Google’s Fiber for Communities project.  Do it today, as nominations close Friday, March 26th.  As to why that’s a punchline, read on.

Google project is quite interesting; they are looking for a community (or communities) of between 50,000 and 500,000 in which to build out a fiber optic network that will deliver 1 Gbps (50 to 100 times faster than your average DSL or cable connection) to every business and home.   Google is not getting into the Internet business; they believe that universal high-speed access is necessary to drive technology (and, of course, their business) forward.  They are looking to fund a research project to demonstrate that it can be done and to determine the best way to do it.

The issue with deploying high-speed Internet access to everyone is more about business models and politics than technology.  In most places high-speed access comes from the phone company or the cable company. In most cases, these are the only companies that have a legal right to run cables through the street.

The situation is called a natural monopoly.  This happens when the infrastructure required to deliver a service is so costly that it’s difficult or impossible for competitors to enter the market.  Competition might actually increase the cost of the service and may be a logistical nightmare.  Consider multiple water companies (a situation that existed in nineteenth century Britain).  Each company has to build out it’s own costly network of water mains, and passes that cost on to the consumer.  Since more companies there are the fewer customers for each to have, the cost per customer is higher.  Then there’s the logistical issue of have six water mains running down any given street.  In these cases, it’s simpler, easier, and possibly cheaper to limit the business to one company.

While it may make practical sense for there to be only one phone company and one cable company running wires, the end result is two companies with little competition trying to shoehorn a business outside of, or even in conflict with (consider Internet phone and video service), their core business.

The solution?  Set up a new company whose only business is providing high-speed Internet that can focus on building the right network to provide it.   Like the phone and cable companies, give them to right to run cable in the streets, and require that they bring it to everyone.  This does create a new near-monopoly, but given that Internet service is practically a utility, like water or electricity, it makes sense.   In addition, an Internet utility would actually increase competition for phone and television.  1 Gbps can easily carry high quality phone service and video on demand.

You may be wondering if wireless is a better option.   As the name says, wireless doesn’t require running wires.   Without that physical infrastructure there’s no natural monopoly and there’s room for competition.  However, (affordable) wireless broadband technology hasn’t arrived yet.  When it does, by its nature wired access will always be faster.  And since wireless bandwidth is shared, there are limits to how many people can use it reliably, as anyone who’s ever tried to use their iPhone in downtown Boulder during the business day can attest.  I hope that wireless competition will arrive, in the meantime if someone is willing to build out a high-speed wired network, we should take them up on it.

Sounds like a no-brainer, let Google’s deep pockets build out what would be the best Internet access in the country and propel Boulder’s already exploding tech community into the stratosphere.   Good for business, who will have better access to the customers.   Good for the economy, no question that world-class Internet access will create jobs.   And, good for individuals, who will see better access and more services.  There’s only one group it’s bad for, Internet Service providers like, us.

So, why am I promoting this?  Because I believe it’s the best thing for Boulder, and I believe it’s the best way to deliver Internet in general.  Having started the very first ISP some 20 years ago, I have a good sense of where the Internet has been and where I think it should go.  In the end, someone is going to own the wires, let’s make sure it’s someone who’s providing Internet access on an unrestricted network.  Not a phone or cable company with other agendas.  Yes, this will hurt my business, but we have and will continue to diversify, and in the long term we will be helped by a grown Boulder economy.

So, go to http://boulderfiber.com,  follow the instructions, and nominate Boulder.   Not in Boulder?  Longmont and Fort Collins have initiatives  as well and you can nominate other communities at http://www.google.com/appserve/fiberrfi.

Nominate today, the deadline is Friday, March 26th.

Twilight New Moon!!!!!

Tuesday, November 24th, 2009

Has nothing to do with this post. In fact, I haven’t even seen it. This is just an example of how the ratings of a website can be inflated using the power of keywords. I’m betting that the number of people that read this post will double just by mis-direction of a title of a blog post. Of course a readership of two would be amazing so lets see what happens.

So how about this instead: Say you’ve just clicked a link to read a movie review and instead of the review you were hoping for you get “Cannot connect to website.” What’s the first step? If you’re running on com(*&^ a major internet provider’s network, it would seem the appropriate call would be to the support line to complain loudly! Not really -  the true cause of internet access more often than not is the equipment in a customer’s house. Typically, large providers will see that the line into the house is up and wish you a pleasant day and ask, via script, “Has X-corp has addressed all of your concerns today?” Of course they haven’t.

All they care about is that the line to your house is functional. That is where their concern ends. All you care about is accessing the web to read all about Robert Pattinson. The line into the house and the modem connected to it is where any ISP has to draw the line. This line, in network speak, is call the DMARC. The reason that line exists is because no ISP can know every combination of equipment that occurs past the modem (not even the über-awsome amazing Indra’s Net). Bearing in mind that our responsibility, like any ISP, officially ends after the modem, we do try to go the extra mile. Many of our customers know the amount of time that we spend trying to apply all that we learn to their home equipment to try to get it running and connected. I believe this is the difference between a local ISP and the “puppy-mills” of the internet connectivity world.

So the next time you can’t get to your favorite pictures of Anna Kendrick or Kristen Stewart, try rebooting your modem then your router as a first step. If that fails call your friendly local ISP; Indra’s Net!

I expect to hear plenty of teen angst vented against me for wasting time you could spend musing on Taylor Lautner‘s abs. Vent quick and return to the sea of web reviews before you’re skin sparkles and gives you away.

OMG, streaming movies!

Tuesday, November 17th, 2009

It’s finally happening - people are starting to watch movies online.  I know, I know…everyone has been streaming and downloading video for at least a few years now, right?  There’s video all over the net.  Every third blog post has an embedded video.  There’s YouTubeVimeo, and video podcasts.  There’s PlayStation Network, Xbox LIVE Marketplace, and Amazon VOD.  Folks are streaming to their TiVos and everyone has Rokus and Windows Media Center Home Theater PCs and Apple TVs.  Forget timeshifting cable with your DVR…everyone gets their TV via hulu or Boxee anyway.  Who needs Blockbuster when your new coffeemaker does Netflix Streaming?  iTunes!  …or, for the more frugal, Bi*Torrent!  Clearly, everybody’s been downloading and streaming movies and TV more or less constantly for some time now, right?

Nope.  The number of conduits for movies and TV over the Internet may have proliferated greatly in the last year or two, and streaming may be as familiar as cable to the geekier among us, but most normal humans haven’t been getting their new releases or favorite shows via their Internet connections.  I know, because I see how much data they download.  And I talk with other network operators who see how much data their users download.  And we compile aggregate stats on things like bandwidth usage patterns over the course of the day, protocol breakdowns showing streaming traffic rates at any given time, target analyses to determine how much traffic we’re pulling from Amazon’s servers or Apple’s Akamai nodes, etc.  There just hasn’t been that much movie and TV traffic.  …that is, until recently.

Streaming TV and especially streaming/downloading movies seems to have gone mainstream.  Over the past 6 months, we’ve seen a steady increase in streaming traffic of all sorts, but particularly in our upstream ingress traffic between 6 and 9pm, which has tripled.  That time period now represents the largest daily bandwidth peak we experience, and unlike most other traffic peaks in our upstream usage, this “streaming movie” peak is exacerbated on the weekends (a relatively sure sign of what’s going on).  Other ISPs are seeing similar trends.

The vast majority of our connectivity customers have 1.5Mbps or higher speed DSL connections.  More than a third have 3Mbps or better.  1.5Mbps is acceptable for real-time streaming of a standard definition movie.  3Mbps is about the minimum for streaming a 720p HD movie.  I fully expect to see more and more movie streaming and downloading from our users, particularly during and around the traditional prime-time period in the evenings.  As streaming video becomes increasingly popular, large ISPs will likely respond more and more with bandwidth caps and/or outright metered usage.  Indra’s Net has no plans to move in that direction, though we will be purchasing additional upstream capacity to satisfy this new demand for bandwidth.  It will certainly be interesting to see how this new model for entertainment distribution plays out over the coming months and years, and I will provide additional details about the movie and TV traffic we’re seeing (always anonymized in aggregate data, per our privacy policy) in a future post.