Voice chatting programs have been around for many, many years. CoolTalk, a program bundled with Netscape Navigator, was one of the first. Others followed, including Net2Phone, all of today’s popular IM networks, NetMeeting, and others. All of these solutions have one thing in common, however – they don’t really work all that well.

Thanks to a friend, I think I’ve found the future of the telecom industry – Skype (it rhymes with “type”). Like its competitors, Skype is a voice-over-IP application. It has text-based chatting as well. What sets Skype apart, however, is it’s peer-to-peer architecture.

P2P, for those of you who don’t know, is a way of transmitting data from user to user, rather than downloading it from a central server. It has a number of important benefits, including the ability to more effectively utilize bandwidth. As a network grows to service more and more users, its capacity grows simultaneously, ensuring there are always enough resources to handle the load. This reduces costs for everyone and usually increases performance significantly.

Skype, a project by the folks that created the [in]famous file-trading network KaZaA, uses P2P algorithms to maximize bandwidth and reduce latency. This, combined with a spectacular compression algorithm, allowed me to talk with my significant other, through a restrictive firewall, at better-than-telephone quality. Neat!

If you have loved ones beyond your local calling area, I would recommend giving Skype a try. Don’t worry, it’s not illegal any more than KaZaA is (the important thing is how you use software, not the mere fact that you use it).

Nobody disputes that the U.S. needs a law to stop unsolicited commercial e-mail before 2004. Not even the spammers themselves.

While I’m glad our representatives finally got a law to the President’s desk, the Controlling the Assault of Non-Solicited Pornography and Marketing (CAN-SPAM) law won’t do much good in stopping the onslaught.

The law, designed mainly to provide recipients a way to opt-out, pre-empts the anti-spam laws of 35 states. Many of these laws require spam to be opt-in, a crucial step to stopping one of the main sources of victim addresses – harvesting from the web and newsgroups. As of January 1^st, however, all bets are off. Instead of empowering users, the law blocks individuals from suing the spammer, instead reserving the privilege for states and ISPs.

The opt-out nature of the law also gives “one free shot” to spammers, as reported by CAUCE (no permalink). The law is extremely lenient on “truthful spam” – spam that doesn’t misrepresent facts. This has the effect of legitimatizing the spam industry, an industry that has been characterized by many illegal acts in pursuit of its profits.

The primary problem with a law like the CAN-SPAM is that it ignores the fact that most spammers use tactics that are illegal now, today. Most spam, even from spammers working in the U.S., is bounced from country to country, rendering it virtually untraceable. Most spam uses subject lines designed specifically to evade anti-spam filters. If you tell a spammer to remove you from their mailing list, it gets you added to 50 more. All the passage of this would accomplish is ensuring that more spam will hit your inbox in 2004.

If my testimony isn’t enough, here’s some from Dr. Jason Catlett of Junkbusters Corporation, about a very similar bill Congress debated on in 2001.

Since the President has indicated that he will sign this, there is very little hope for a fix for the problems – for now. However, let me leave you with a word of warning – if the FTC decides to establish the Do-Not-Spam list, do not add your address. Doing so would be like hanging a neon sign on your inbox.

Many people have tried to map the Internet – a daunting task, to say the least. In theory, a little more than 4 billion computers can be connected to the Internet (a limit we are fast approaching for a variety of reasons, by the way), so as you can imagine, this isn’t exactly the simplest activity to do. In the past, nobody’s had a very good idea of how many computers are actually online, because there’s no central authority computers need to report to.

Enter the Opte Project. The site has completed one of the rare (if only) maps of every class C network on the Internet (in other words, a good chunk of it). Most impressively, the entire map is built via a variety of open-source packages – including at least three packages in LAMP – Apache, MySQL, and PHP. Perhaps the most impressive statistic of all is the time it takes, from start to finish, to render a map – 268 hours, or 11 days. This could be even faster if they decide to do mapping via a cluster or grid.

I briefly spoke to Barrett Lyon, the leader of the project, via IRC.

Waileia: What are the points [on the map]?
Barrett Lyon: Every point is a computer or a network device, or something with an IP…There are over 13 million of them, [and] inside each, 256 IPs that could be live or not…If we were to draw them on the map, it would make a mess. 🙂
W: So, the edges could represent things like NAT boxes, internal LANs, and so forth?
BL: Yes, and the connected dots [and] lines are all routers.

While it’s been taken offline due, in part, to the slashdotting, there’s also a page where you can find yourself on the map.

Original link via Slashdot. Image used with permission of the Opte Project.