I’ve been meaning to check out OpenDNS for some time now and finally got around to it, in part thanks to Paul and WordPress developer Matt Mullenweg’s blog post. I’m only playing with it in my dorm right now (I set it up as my static DNS servers in my router and DHCP does the rest). It’s billed as a fast DNS service with typo-redirection and anti-phishing capabilities. It is fast, Tech’s DNS servers have been noticeably slow this semester when performing uncached lookups (they’re still fast when the lookup is cached though) and when I visited several sites that are notorious for being uncached lookups it zipped right along. The TLD typo (.comm -> com) works well and quickly (I will admit to making such mistakes from time to time). I haven’t done much with their phishing capabilities. Phishing hasn’t been an issue for me, but obviously it is for some people since we have billions of dollars disappearing as a result of identity theft, some of which occurs due to phishing.
The other service the OpenDNS people rolled out today is their PhishTank service. It’s essentially a user-driven phishing URL submission service that relies on the community to verify phishing URLs. It’s got a great interface and what looks to have a good API. They’re working on plugins for SpamAssassin and Thunderbird, which will make this service useful beyond its use in OpenDNS (the plugins are what really interest me). There’s talk of using it with antispam plugins for blogging comment spam, à la Akismet. One thing they’ll have to do is make it incredibly simple to submit phishing URLs with their plugins (One click submission in your email client is almost a must), otherwise this service will quickly become useless (user driven services are useless if they’re not easy to use and no one ever submits content).
There are already services that do URL blacklists (Spamhaus does phishing and spam), the issue that has arisen from time to time is that these services aren’t very democratic and from time to time there have been abuses (a lot of people use these services (or their ISPs use them) to filter junkmail and as such they hold a lot of sway over what people ultimately see in their mailbox). That’s not to say that a democratic system is immune to such abuses, but it should be easier to rectify issues that come up when others can verify the validity of your claims without waiting for a select few to “get around to it.” I’ll be interested in the statistics on false positives, percent of phishing emails caught (emails that trigger PhishTank / actual phishing emails), and such when the project has gone on long enough to get good information on its real usefulness.
