I am having just been in NYC for a couple of board meetings (see I do work from time to time).  Last week, I was in Tokyo which I am found of saying is like New York without the dirt and crime but that is not really fair. New York is like no other place in the world.  And while many find it expensive, I think it is the cheapest place in the world per experience. When I am in NYC for business, I normally stay at the W at Times Square.  They usually give me a corner room  approx. 50 stories high where I can look right down on Times Square.  Which brings me to the purpose of this post.

Most hotels in the USA charge for high speed access.  Frankly, I think this should be included in the room but then again, I don’t think they should charge $1.50 per minute for phone calls, or five dollars for a glass of water.   I can only imagine what the charge for the so called intimacy package but I certainly hope the quality of that product is better than their broadband offering.    I measured the broadband at the W as something less than 250kbits and sometimes not more than 125kbits. The charge for this is about $20 bucks.  At first I thought that it might be this low because of congestion because even if the hotel has a dedicated line, all the guest rooms are probably connected to one big LAN and are sharing bandwidth.  But even in the middle of the night, there as not much more bandwidth. The crazy thing was that Sling was running as it normally does when I am viewing my tivo over the net at around 350kbits.  In addition to slow bandwidth, latency was horrible.    I have decided to get a high speed  EDVO connection from Sprint.  That is also a shared medium and probably sucks during the day but it will pay for itself in a just a few days of not paying the hotel for their awful broadband.  And I can use it at the airport instead of paying T-Mobile.  And I can use it when I am in the car going to and from the Airport.  Now you Blackberry people don’t care about any of this I imagine. Which brings me to one of my favorite subjects:  The sorry state of broadband in the USA. 

 Brian Roberts, the CEO of Comcast and someone I have get respect and affection and which I hope is returned (for it was me that introduced Comcast to the idea of cable modems back in 1993), announced that Comcast will be offering 150 mbit bandwidth with DOCIS III and this should be available in 2-3 years (probably in areas where Comcast has competition for Telco’s deploying fiber).  What they are doing is converting more of the channels that are “wasted” on analog video, to be used for downstream IP. The best way to think of the cable network is like a highway with lanes and a maximum speed limit.  If there is very little traffic you can go at the maximum, but if there is a lot of traffic you can crawl.  Now for most people the times they may want to drive on the highway are the same times that others want to drive on the highway (like going to and from work) so the average speed that drivers experience is much lower than the speed limited posted.  When it gets really bad, they add lanes to the highway (which is what Comcast is doing).  But if traffic grows faster than the highway grows then we still end up with traffic jams.  And guess what? Traffic on the internet is going to grow faster than they will add lanes.  In fact, the Comcast announcement means that they can only guarantee about a mbit per home unless they drive fiber deeper into their networks and reduce the size of nodes something that would be expensive and take a lot of time to do (it requires things like ditch diggers and pole line men).   One of the things that has made a shared network like cable work, is that only a small number of consumers are on the computer at any one time and the usage model was been browsing which is point, click, wait, read, and do that all over again.  When a system is highly interactive there is a lot of dead time using we are talking about 15 years playing games.  But video is changing all that and in addition the number of users in a home is increasing and the amount of time people are spending on the net is changing.  So while, I am very glad that Comcast is adding this capability (which by the way, they could have done a long time ago in my opinion), it is just an other bandage.  And of course none of this deals with upstream bandwidth which is very limited for cable.  The issues here is the P2P will not work if deployed in large numbers within cable systems but they guys that are developing P2P do not seem to care or maybe they just do not know.  So DSL is point to point (for the last mile).  While does not offer the potential bandwidth that cable has, it should give a consistent amount of bandwidth.  In a sense, it is more like a train (like a Japanese Bullet Train) only slower).  Right now consumers do not understand all this and I don’t think they will every understand it technically but they will certainly understand when their video is flaky.  I am just surprised that all the investors in video distributions networks do not seem to understand the issue or maybe they thing they can get in and out before the problems become obvious. Check out what GigaOM has to say on this topic here.  The comments are more technical and worth reading. 

You may wonder why I did not speak about BPL (broadband over powerline).  Since I am a director of Current Group, the leader in BPL, I do not feel that it is right for me to comment on this other than to say that I believe the BPL will be very competitive with Cable and Telecom offerings.