First up this week: The Red Screen of Death. “Hold hard!” we hear you mutter, “surely you mean the Blue Screen of Death?” Nope, red it is, at least currently where Longhorn is concerned.
In his blog, Michael Kaplan, a “Technical Lead from Globalization Infrastructure, Fonts, and Tools at Microsoft,” notes that he was messing around with a Longhorn beta and after doing something evil to the registry and booting under Virtual PC, he got the fabulously ungrammatical and totally unhelpful message, “Info: An error occurred transferring execution” (yes, it really was spelled that way) on a screen with a red background!
Kaplan notes, “I am not sure I would class the change as an improvement.” Never was a truer word spoken. We hope that Mr. Kaplan is flying beneath the executive radar at Microsoft, otherwise he may well be going down in flames.
According to Microsoft droids, this display will not be included in the release version of Longhorn, but as reader Jack Miller, who brought this amusing diversion to our attention, points out: “Early versions of DOS had better error messages . . . but it seems Microsoft is going the wrong way. I can only conclude this is a planned journey to make users more dependent on Microsoft and increase their stronghold on the market. . . . That’s why I’m rooting for Linux.”
But that isn’t all readers have been writing in about. Morely Dotes was the first to write in after our recent brief discussion of an anti-spam product called Qurb.
Dotes, which is not his real name for reasons that are unlikely to be cleared up here, disagrees with the challenge/response methodology Qurb uses: “I don’t suppose it ever crossed your mind that [challenge/response] systems reduce your spam problem by offloading it onto unwilling strangers, did it? People who use [challenge/response] systems are, inevitably, spammers themselves.”
What Morely means is that challenging a message from a spammer who uses a false address will result in the challenge request going to an innocent bystander, the real owner of the abused address.
Dotes, who runs an e-mail hosting company, feels that such an irritation should not be allowed to go unpunished: “As an e-mail admin, I have a two-pronged response to challenges sent to my users: 1) I reply to the challenge, so that future spam using the same forged ‘from’ address will be delivered to the lazy sod who demands that my users sort his spam for him. 2) I block the IP address of the sender permanently at the firewall, so that future contact from that lazy sod is turned away by the Border Patrol.”
Dotes finished “. . . since you . . . have a track record of admiring ‘solutions’ to the spam problem that only make it worse, I’m having this e-mail archived for the public record.” There Mr. Dotes, we did the public record thing for you.
Of course, Dotes wasn’t the only person to write in arguing against challenge/response systems. Check out the Gibbs forum for a well-argued letter on the topic.
We raised these issues with the chaps at Qurb and they pointed out that the number of challenges from challenge/response systems will be trivial compared with the number of bounces that the owners of forged return addresses receive.
The problems with challenge/response systems raise some interesting questions. First, even when the incoming mail is filtered so that obvious spam is removed and obviously incorrect challenges are not sent, is the use of challenge/response ethical and/or reasonable? We think it is both, but we’d like your thoughts.
Second, do we need a standard for challenge/response so that its use would be generally acceptable? This is easy: If the standard challenge was to include an X-header (for example, “X-Challenge:”) followed by the original message ID, then when a challenge recipient gets a challenge with an ID he didn’t create, he would just delete it sight unseen. Voilà! A small increase in Internet traffic, a large increase in the value of e-mail.
Third, and specifically with respect to Dotes’ behavior regarding challenges: Is Dotes’ way of dealing with challenges commonplace and is it ethical? We suspect that it just exacerbates the overall spam problem without extracting any reasonable or effective justice.
Cry “Havoc!” and let loose the dogs of code at gearhead@gibbs.com. And, ahem, there’s always Gearblog.
