Found: One Troll

Fuck ATM. Seriously, who the fuck uses that anymore?

I got this link in my inbox today. Head over there if you feel like testing the pressure seals on your arteries. At a time when our national debt is skyrocketing out of control, a time when ISPs are struggling to meet our margins because people have decided that high speed internet isn’t as important as, say, food or shelter, the Federal Government wants to help us all get broadband internet! Let’s see what they’re on about.

Goal 1: At least 100 million U.S. homes should have affordable access to actual download speeds of at least 100 megabits per second and actual upload speeds of at least 50 megabits per second.

Well, at least it’s only 1/3 of the population. DOCSIS3 cable is technically capable of providing those speeds, but not with any kind of guarantee and not with a high oversubscription rate. Fiber to the home can do it a lot more easily, but not affordably except in areas with a relatively high population density. It’s branching out, but you’re still not going to see folks in Sedona, AZ with FTTH for at least a few years, probably longer if the current recession continues.

Goal 2: The United States should lead the world in mobile innovation, with the fastest and most extensive wireless networks of any nation.

Great idea. I’d love to see how we’re going to pay for it.

Goal 3: Every American should have affordable access to robust broadband service, and the means and skills to subscribe if they so choose.

I don’t even know what this means. There are a number of Federal subsidies to encourage ISPs to build their networks out into otherwise unprofitable and remote areas, but are they also suggesting that we help people finance their internet connections?

Goal 4: Every community should have affordable access to at least 1 Gbps broadband service to anchor institutions such as schools, hospitals and government buildings.

This one is actually doable. As long as we’re building leagues of (government subsidized) fiber across the US, it’s trivially easy to get 1Gbps service for any given city. The cost is only a few thousands of dollars per site compared to the millions required for the other goals. The ISP that I work for provides Metro Ethernet to a large number of cities, schools, and hospitals.

Goal 5: To ensure the safety of Americans, every first responder should have access to a nationwide public safety wireless network.

And, folks, here is the real goal. The Government would like to turn Internet access into a public utility so that they can regulate it like one. It will be done under the guises of “safety” and “fairness.” Innovation in the Internet sector will slow to a halt as it comes under increasingly byzantine regulation, drafted by people who couldn’t tell you the difference between a bit and a byte and think a “firewall” is a bad movie starring Harrison Ford.

Goal 6: To ensure that America leads in the clean energy economy, every American should be able to use broadband to track and manage their real-time energy consumption.

This is just bullshit environmental fuzziness to gloss over the “safety” they just shoved up your internet tubes.

Here’s the deal.

The Internet is not a basic human right. It’s awesome (in the original sense), I love it, and it keeps me employed. But it’s optional. It’s a luxury, just like cable TV or owning a car. The Internet, cable, and cars are useful tools. They make us more productive, they can help us stay informed, and an informed population is a more free population, but we don’t have the money to give everyone in the country free Internet, free cable, or free cars. Trust me when I say that we also don’t have the money to subsidize the upgrades necessary to accomplish the government’s goals; they’ll have to come at the speed the market is capable of supporting.

Frankly, I don’t understand this drive to be “#1!” at teh intarwebs. Fine, Japan has 100M symmetrical access across the country. Whoop-de-doo. I frequently see municipal governments and schools with 100M pipes that don’t use more than 10% of their available bandwidth. If you want people to be “connected” to surf the web and download some pr0n movies and stuff, 5M is sufficient. 100M is overkill for greater than 99% of users. If we let the market run its course, we will get 100M to residences across the states in a few years when there is an actual demand for the service. There’s no need for the .gov to pump in billions of dollars for negligible benefit.

Frak, wasting this kind of money just pisses me off.

Do you wish you could feed your cats from across the world? Well, check this out:

This guy built a pet-feeding device out of a cool machine and an old Cisco switch. The details of how he did it are here. (Via Jeremy)

I have arrived in Detroit for NANOG 47. This is the first event I’ve attended where there is an actual PGP key signing party. I think that makes this the nerdiest place I’ve been since college.

Yesterday I passed the CCIE Routing and Switching Lab Exam on my first attempt. I guess I owe myself an AR-15.

I’ve scheduled my CCIE Lab exam for 26 August in RTP, NC. Wish me luck.

I passed the CCIE Routing and Switching written exam today with flying colors. What a difference a good night’s sleep can make. The only thing now between me and a unique CCIE number is the lab exam, which is a little like saying the only thing between me and millions of dollars is a bank vault door. I’ve got another post on IPv6 in the works based on some of the stuff I learned at Cisco Networkers from folks who’ve actually deployed it on a large scale (mostly from Europe and Asia). Hopefully I’ll get that up by the end of the week.

Sorry about not updating. There’s been lots to write about but no time to do so.

I’ve gotta say, flying from Phoenix to San Francisco is like travelling from Arrakis to Caladan. It might seem odd, but the first thing I noticed about San Francisco is the water. It’s everywhere! Imagine that.

I registered for the convention late so I ended up in a different hotel than my co-workers. They’re in a hotel about a mile closer to the convention Moscone Convention Center but I got a room larger than a broom closet and in a place that serves a great steak. I figure I got the better deal. Both hotels charge for internet, though. $13 a day when I’m only awake in my hotel room for 2 hours a night is pretty crappy, even if it does go on the company card.

I’ve been doing a lot of reading on the Kindle instead. Can’t say that I mind all that much. I think I actually read faster on the Kindle, but I’m not sure. I do know that I’m almost done with Hitchhiker’s Guide to the Galaxy and I’ve only been reading it for abut three hours.

So far the convention has been pretty awesome. On Monday I took a mentored CCIE routing and switching practice lab, which was a great experience. I learned a bit about some areas where I’m weak, specifically IPv6 and multicast configuration. I’ve got a good grasp of IPv6 concepts, but there’s so seldom an opportunity to actually configure and troubleshoot IPv6 that it’s going to be a significant hurdle when I take the CCIE Lab exam in a few months. I also got a good look at the format of the exam and the sorts of ‘gotcha’ scenarios Cisco puts on the lab exams. Normally I’d have had to pay $1500 to fail the test to get that kind of experience.

This morning I attended a session on deploying scalable OSPF in a service provider network. The Open Shortest Path First routing protocol is the most common routing protocol in the world, because it’s easy to configure, every vendor supports it, and every engineer learns it forward and backward. It’s a lot like IPv4 (or VHS) in that it’s not the best tool for the job, but everyone knows how to use it and everyone supports it.  It was a good lecture but I spend all day hip deep in OSPF at work, and I didn’t really learn anything new. I’m not sure what I expected. . . I think I was hoping that there was some secret OSPF juju that I don’t know about.

Of course, as I mentioned before, I took the CCIE written exam a couple of hours ago. I failed the exam by 14 points out of 1000. In other words, I failed by one difficult question or two easy questions out of 100. It’s the worst way to fail, in my opinion. The earliest I can retake the exam is Monday, so I’ll study over the weekend and hopefully pass then.

This afternoon I have a class on IPv6 security and tomorrow I’ve got a couple of sessions on multicast deployment and MPLS. These are my weakest subjects, so hopefully they’ll also do some good both on the certification front and in my job, where we’re currently designing a multicast video deployment.

I’ve got to run to the IPv6 session now.

In 12 hours I’ll be on a flight to San Francisco in order to attend the annual Cisco Live conference (formerly Cisco Networkers) until Thursday. Should be interesting, I’ll be attending seminars on IPv6 security, MPLS traffic engineering^[1], multicast routing, and a CCIE Routing and Switching lab tutorial. On Tuesday afternoon I’ll be taking the CCIE R&S written exam.

I’m taking my Eee in addition to my work laptop, so I’ll try to blog frequently, there should be plenty of network coverage. . .

1. I couldn’t find any good articles to explain MPLS traffic engineering simply. Basically, it’s a way to implement end-to-end QoS and policy-based path selection using label switching. In other words, it’s faster and less CPU intensive than true packet-switched QoS. However, it’s only useful in high-bandwidth, large-scale networks. The benefits are negligible in smaller networks. [↩]