Sunday, January 27, 2008

Why US connectivity is lagging

We have discussed the decline of the US in terms of connectivity. A recent article in the Economist asks why some developed nations are doing better than others, concluding that "The best explanation, in fact, is that broadband thrives on a mix of competition and active regulation, to ensure an open contest".

In South Korea, where half the population lives in large apartment buildings, each owns its own internal cabling and allows rival operators to install equipment in the basement. Tenants choose which to use. In Japan, politicians pressured the dominant operator, NTT, to connect people's homes by high-speed fiber lines. The Communications Ministry has indicated that it will make NTT open those lines to rivals.

In theory, France, Canada and the US also passed laws requiring incumbent telephone companies to allow competitors access to their local lines. Those laws have succeeded in generating competition in France and Canada, but not in the US where regulatory and court decisions have enabled the telephone companies to keep competitors out.

Average advertised download speed in Japan is 95 megabits per second. How does that compare to the speed in your home? What sorts of applications would such high speed make possible? In Denmark, the Netherlands and Switzerland, over 30% of inhabitants have broadband. In the US it is around 22%. How might that impact the economy?

WiMax deployment begins

We have seen that DSL and Cable are the two most common technologies for home connectivity, but there are alternatives. WiMax wireless is beginning to roll out.

TDS Telecom has just announced the availability of fixed WiMax service to 55,000 households and 10,000 businesses in Madison Wisconsin. Residential service is $45/month for 2 Mbps, $50 for 4 Mbps, and $55 for 6 Mbps.

Note that this is connectivity to a fixed location, an alternative to DSL or cable. Mobile WiMax, which will compete with cellular service, has not yet been deployed.

TDS also offers DSL connectivity in Madison. The rates are $30 for 768 Kbps, $40 for 1.5 Mbps, and $50 for 4 Mbps. At these rates, which would you prefer? Are both available throughout the Madison area?

Monday, January 21, 2008

Infrastructure services -- for developers, not users

We have discussed many Internet services intended for applications like editing documents or developing simple applications like blogs or online databases. These are generally intended for users -- people who use the Internet in their work or play, not professional developers.

We can contrast these user services with infrastructure services, which are intended for professional developers building user services or custom applications. Amazon has been a leader in this field. They learned to operate vast, efficient, reliable server "farms" in support of their shopping site, and decided to allow others to use the same infrastructure. Amazon Web services include virtual servers, storage space for large amounts of data, and storage and retrieval of data from flat files.

3Tera's Applogic offers a graphical user interface for designing a virtual server farm and deploying applications. In this example, the drag and drop interface was used to create a site with four firewalls (gray), a load balancer (yellow), two Web servers (blue), a MySQL server (red) and servers storing configuration files and logs (orange).



Each of these is an independent virtual server running Linux. Developers can configure these machines and load software on them as if they were in sitting on their desks. They create "dashboards" to monitor them, and can increase or decrease capacity in seconds. (Check out this Applogic video demonstration).

These services allow one to develop applications without hardware or datacenter cost. They bill for resources used -- CPU time, storage, bandwidth. That means there is essentially no cost while an application is being developed and debugged since there is no traffic. When the application goes live, the capacity, and hence cost, grow and shrink dynamically depending upon utilization.

Storage and computation are not the only infrastructure services. For example, Amazon offers a service for handling monetary payments from computers or users to other computers or users and Lypp.com offers telephony services. For example, a developer can give the user the ability to initiate a call between their telephone and a company. Lypp servers establish the call, interface with the switched telephone network, and bill for minutes used. Again, there is no traffic, so no cost, until the application is put in use.

Infrastructure services are the latest step in server evolution. Would infrastructure services appeal more to large or to small and medium sized organizations? Why? We offer courses today on programming for vendor-specific platforms like ASP.NET. Will we one day offer a course in programming for Amazon Web Services? Would you like to take such a course?

Best practice: what to do when a service outage occurs

We have discussed the pros and cons of relying on service providers. Reliability is one issue, and, when something goes wrong, you want fast action and continuous status reports. 37Signals, provider of Basecamp project management and other services, responded well during a recent two hour service interruption. The following log was posted on their Web site, fully informing users of the problem and progress toward its solution.

  • All systems are currently offline as we're experiencing network outage from our provider. We're working on it right now. No data has been lost, all our machines are still working, but they're not accessible from the internet. Sorry for the inconvenience — 10:03am CST (16:03 GMT) on January 18, 2008.
  • We have located the problem to be with the load balancer setup. A new unit is being installed. We should be back shortly. Again, we're terribly sorry for this disruption of service — 10:28am CST (16:28 GMT) on January 18, 2008.
  • The technicians at our service provider are still working on installing the new load balancer. We're breathing down their neck as heavily as we can. And we profusely apologize for this unacceptable interruption of service — 10:56am CST (16:56 GMT) on January 18, 2008.
  • The load balancer has been swapped and is currently being configured. We should be in the home stretch now. Again, we're incredibly sorry for this disruption. This is not how Fridays are supposed to be — 11:18am CST (17:18 GMT) on January 18, 2008.
  • Our service provider is still working on the configuration of the new load balancer. We're on their case every few minutes to get updates. It's hard not to be very disappointed that a simple load balancer replacement can take this long for someone who's supposed to be the best in the business. We're out of new ways to say we're sorry, so we'll just say it again: We're so sorry for this — 11:50am CST (17:50 GMT) on January 18, 2008.
  • The latest from our service provider is that the reason it's taking so long is because the configuration for the load balancer lives on a CF card that has also gone bad. The configuration is currently being rebuilt from scratch — 11:57am CST (17:57 GMT) on January 18, 2008.
  • Now we finally have access to the servers with remote access. We should be able to bring the services back one by one now. Looks like there's finally light at the end of the tunnel — 11:59am CST (17:59 GMT) on January 18, 2008.
  • Basecamp is now back. We're bringing back everything else we can as fast as possible. It might take a few minutes for the DNS access to update, but it's coming back — 12:05am CST (18:05 GMT) on January 18, 2008.
  • All the products should be coming back online now as soon as the DNS updates. It seems like we're coming out of the woods entirely7 — 12:13am CST (18:13 GMT) on January 18, 2008.
Could an outage like this occur if the user ran their own application? Would in-house staff have responded as well as 37Signals staff did to this outage?

Thursday, January 10, 2008

Video of a Google Android cell phone prototype

We have seen that a dumb, open network encourages investment and hardware and software innovation. We have contrasted this with the telephone and cellular networks, where application and hardware innovation are controlled by the network operators. We've also followed Google's efforts to open the cellular network by possibly purchasing spectrum and by developing and distributing Android, an open source mobile phone operating system bundled with Google applications like maps and mail.

Google has released some videos demonstrating prototypes of Android phones. For example, the following video shows a touchscreen phone reminiscent of the Apple iPhone.



(You can find other Android-related videos at the Android Web site linked to above or by searching on Youtube).

The most important feature shown in the video is not clever hardware or software -- Apple already did that with the iPhone -- but the open platform. The video begins and ends with Google co-founder Sergey Brin inviting developers to build Android applications. Based on his Internet experience, he feels that openness will be critical to the success of Android, and has kicked things off with a ten million dollar contest for application developers.

If he is right, third party applications will spur Android sales, and each Android phone sale will spread Google applications -- and ads. Not satisfied with being on your desktop, Google also wants to be in your pocket.

After viewing the prototype video, can you think of applications you would like to have on your next cell phone? Would you be willing to watch targeted advertisements in return for free cell phone service?

Saturday, January 05, 2008

The OLPC XO, a LAN machine in the Internet era

We have talked about the One Laptop per Child (OLPC) computer called the XO. Sales have been slower than hoped and Intel has left the OLPC coalition because they were unwilling to stop selling their low cost Classmate PC in developing nations. An Intel sales person tried to unhook the OLPC sale to the government of Peru, and OLPC head Nicholas Negroponte compared that to McDonald's competing with the World Food Program.

But, the Classmate is proving a strong competitor to the XO in spite of its higher price. Is Intel selling it below cost to eliminate a competitor or is the Classmate a better product in the eyes of the education ministers and other government officials who are the target market. (Both are being marketed, appropriately, as ubiquitous educational infrastructure).

I've not seen the Classmate, but have an XO, which I find disappointing. The hardware is innovative and appropriate for kids and rural areas in developing nations, the environment for which it was designed, but, it is a pre-Internet machine. We have outlined the evolution of application development and delivery platforms from batch processing to the Internet, and the XO is still primarily a LAN machine in the Internet era.

It can connect to the Internet via WiFi (if you have an open access point), but the browser cannot handle mp3 and Flash files, foregoing tons of kid-oriented material. Such shortcomings may be overcome with software upgrades, but that will make a more fundamental problem apparent -- XO desktop applications like image, audio and video processing will seem lame compared to their Internet-based counterparts. You can't keep them down on the farm once they've seen PBS Kids or Jumpcut.

One can argue that most XO users will be on XO LANs in locations without Internet connectivity so they won't know what they are missing. That is true, but sad -- second rate technology for the third world.

Mr. Negroponte discussed Internet connectivity in a presentation at the MIT Emerging Technologies Conference in 2005, stating that

It is not a solved problem, but there are many people and many systems working on it. Wifi, WiMAX, GPRS, 3G, 4G, fiber, on and on and on. ... (Connectivity) is happening. It doesn't need me, it doesn't need MIT, it doesn't need the media lab.
Negroponte believed that local interest, competition and regulatory reform would take care of the connectivity issue. Unfortunately, as I have argued elsewhere, global competitiveness and regulatory change is not leading us to connectivity, but to a growing gap between developed and developing nations.

Uruguay, an OLPC customer, recognizes the need for connectivity. They will be providing connectivity to OLPC schools. I hope the XO is a better Internet access device by the time those Uruguayan children get theirs.

Where the Internet is not available, the XO seems appropriate. It is designed for the LAN. The concept of connecting to and sharing with local XOs and a local server is "baked in." The only flaw there may be speed, which I was not able to test. (A mesh network with single radios in each machine could be mighty slow).

Connectivity is beyond the control of OLPC, but I was also disappointed with aspects of the user interface. Kids who are forming a mental model of the machine and network need immediate feedback. When a kid is exploring, every click and gesture is an experiment. The cursor must always change shape when over a hot spot, every click has to be executed immediately or, if that is not possible, an "hourglass" or explanation of what is happening needs to appear, etc.

If the target market is the LAN, the XO is a reasonable alpha-test in need of rapid software upgrades. The only way to achieve that in a timely manner is for a strong open source development community to form around the machine. OLPC would do well to encourage the formation of a social network of developers and support them with Windows and Mac-based development tools.

In the (slightly) longer run, it needs to make the transition from a LAN machine to an Internet machine, and, as in Uruguay, the rollout needs to be coupled with connectivity.