I'll be frank - I couldn't think of a good idea for an article today. There are a couple of interesting links in the news, of course, which I could share with you. And we will get to those in a second, but…

…truth is, I wanted to get a little introspective about things.

This blog is based on Movable Type v.3, and while we can talk about what I should have done, switching to a different system, such as WordPress really didn't make a whole lot of sense.

The one problem that MT had was that we were getting deluged with spam. Hundreds of spam messages a day.

Now, there is a setting that is supposed to catch junk posts. However, this was worse than useless, as it didn't catch a great deal of the junk messages - and it was classifying some good messages as junk mail. In fact, I think it might have been classifying most good messages as junk mail, which may be one of the reasons that we didn't get many comments on this blog for - oh, the first 16 months or so.

A few weeks ago, we moved to a CAPTCHA based system, using ReCAPTCHA. This has been working well - we've gotten more comments, more frequently, and spam is almost entirely gone. Yeah, I know CAPTCHAs are a pain, but it's the only solution we can think of at this time.

Still, 16 months of bad comment moderation may have discouraged regular readers from becoming regular commenters. What I'd like to ask is that, if you've tried to comment in the past but got discouraged, try it now. And if it still doesn't work, for whatever reason, feel free to send me an e-mail to my work address, brian.boyko@netqos.com. I really could use your suggestions for stories to investigate or issues in technology to talk about.

That said, here are those interesting links:

New York Times: Hopes for Wireless Cities Fade as Internet Providers Pull Out:

EarthLink announced on Feb. 7 that "the operations of the municipal Wi-Fi assets were no longer consistent with the company's strategic direction." Philadelphia officials say they are not sure when or if the promised network will now be completed.

For Cesar DeLaRosa, 15, however, the concern is more specific. He said he was worried about his science project on global warming.

"If we don't have Internet, that means I've got to take the bus to the public library after dark, and around here, that's not always real safe," Cesar said, seated in front of his family's new computer in a gritty section of Hunting Park in North Philadelphia. His family is among the 1,000 or so low-income households that now have free or discounted Wi-Fi access through the city's project, and many of them worry about losing access that they cannot otherwise afford. Philadelphia officials say service will not be disconnected.

"We expect EarthLink to live up to its contract," said Terry Phillis, the city's chief information officer.

The problem was that EarthLink's plans required more routers than they initially predicted, which makes me wonder if those predictions were tested on smaller scales first. However, there is no problem with the technology - it performs as advertised. The problem is that there's no real clear way to make a profit from that technology - which, to me, makes it an ideal service that the municipality should provide, rather than outsourcing it to a private company.

George Ou: Fixing the unfairness of TCP congestion control:

George Ou at ZDNet claims that "swarming" is causing a significant bandwidth problem, and goes to great lengths to explain why, in a four page article.

Simply by opening up 10 to 100 TCP streams, P2P applications can grab 10 to 100 times more bandwidth than a traditional single-stream application under a congested Internet link. Since all networks have a bottleneck somewhere, a small percentage of Internet users utilizing P2P can hog the vast majority of resources at the expense of other users. The following diagram illustrates the multi-stream exploit in action where User A hogs more and more bandwidth over User B by opening more and more TCP streams.

But as I read it, it seemed a bit dubious to me. After all, if my multiple TCP-stream connection on my home computer allowed me to have multiple bandwidth links, wouldn't that mean that a download on BitTorrent of a Linux distribution operating at max capacity would be faster than a single TCP stream and FTP connection to a server? In practice, I've found that both speeds are roughly equal - except when there's a lot of demand on the server side; like the first few days after a new Ubuntu version comes out. Then the multiple TCP streams come in handy because they are coming from multiple TCP connections to different locations. But it's impossible for the multiple TCP connections to take up more bandwidth than had been allocated by the service provider under their QoS policies.

Where there is some validation to this is when the pipe gets completely congested to the point that the available bandwidth per user is less than the bandwidth allocation provided by the ISP to the individual users. In other words, it only occurs when the provider has under-provisioned for the network demand and is delivering less than promised to begin with.

Ou suggests an update to the TCP/IP stack that prevents this problem, but for ISPs, the solution is simpler. Either add more bandwidth so that you can deliver the service promised, or promise less if you can't deliver.

PC World: Tech Workers favor McCain, Obama:

Not getting into politics, but this little fact from the article is interesting:

The survey, of 600 self-identified IT workers, found that 27 percent have used the Internet to contribute to a political campaign. By comparison, less than 0.3 percent of U.S. residents have contributed more than US$200 to a U.S. political campaign during the 2008 election cycle.

Which implies that the techies, who by definition are likely to be Internet savvy, are highly politically motivated and therefore very interested in events from the 2008 presidential race.

Hmm, did you experience a bump in recreational network traffic around the time of Obama's speech on race?