Month: August 2007

Seems I was trying the gods of air travel when I wrote: For example, it might seem surprising that an airline like Luxair (which does pretty much exclusively point-to-point traffic between Luxembourg and the rest of Europe, plus some holiday destinations, and should therefore have very little lost or delayed luggage) ends up on place 19 (losing 17.7 pieces of luggage) for the second quarter, and so on.While I was on a direct flight with them this morning, my luggage was not — or at least not on the same one. And that after early check-in, in Vienna…Later: Seems like I shouldn’t blame on Vienna what can be explained by sloppiness here. The suitcase “appeared” two or three hours later at the airport, and was then delivered. The tag on it suggests that it didn’t see any planes except for the flight from Vienna to Luxembourg that I was on, too.

Over at IGP Blog, Milton Mueller has posted a WHOIS timeline, which indeed makes for a nice presentation of the institutional drama that ICANN has seen over the years, and that I spent more time on that I’d ever want to be reminded of.The timeline gets a bit fuzzy on the facts, however, when it comes to the 2003 and 2004 events.

Greylisting is the idea to reject incoming messages with an SMTP error code that indicates failure, unless the source of the message has sent e-mail to the given recipient before (or some other heuristic for “we’ve seen that source before”). The observation underlying this scheme is that e-mail has traditionally been a store-and-forward medium able to deal extremely well with all kinds of temporary glitches. Ordinary mail servers will just queue up a message when they get “greylisted”, and try again after a while (and again), at which point the message will be accepted. The SMTP implementations used by spammers, however, seem to commonly just fail when they encounter any kind of SMTP error.

The statistics show that greylisting (I’m using postgrey) has been removing a significant part of my incoming spam burden; the green curve represents the spam and other junk detected on my personal mail server, the blue curve represents the messages that get through (still including a significant amount of spam, most of which is caught by a bogofilter installation on my laptop).Of course, greylisting has one significant weakness: It will only work as long as it makes an insignificant dent into spammers’ output (as opposed to just some sites’ input). As soon as the technique becomes widespread enough to become noticeable for spammers’ returns, we’ll start to see SMTP implementations for spammers that implement some kind of retry mechanism. Until that happens, however, greylisting is a truly useful tool.

The Creative Commons licenses come to Luxembourg: There will be a launch event on 15 October, at the Public Research Center Henri Tudor on Kirchberg.Speakers will include Patrick Peiffer (the driving force behind Luxcommons, and also involved with the CC-licensed Luxembourgensia collection at the National Library), John Buckmann (Magnatune founder), and Laurent Kratz (Jamendo founder).And for the Web 2.0 crowd: It’s on upcoming as well.

If you wonder why your baggage got lost this time, then the Association of European Airlines is a good source, with its statistics about punctuality, lost luggage, and a timeline of events. In their consumer report for the second quarter (found on flyertalk) , the number of times that Heathrow’s baggage system simply broke down — independently of any well-known external difficulties such as terrorist attacks — is striking, and scary.Unsurprisingly, British Airways ends up on the last place (23 out of 23) in the baggage handling discipline, “delaying” 28 pieces of checked baggage per 1,000 passengers. To put that number into perspective and see just how far an outlier it is, Swiss was on place 10 (losing 10 pieces per 1,000 passengers), Lufthansa on place 16 (losing 16 pieces per 1,000 passengers — the European average), Air France on place 18 (losing 16.3 pieces), and KLM on place 21 (losing 17.8 pieces). In the first quarter, BA had performed slightly less miserable (place 23 out of then 24, and ahead of TAP which is place 22 in the second quarter), losing “just” 24.7 pieces per 1,000 passengers, without any reported baggage system failures at Heathrow.Note that, when comparing lost baggage numbers, some care is needed: missing baggage reports are taken by the last airline in a chain of connections. Yet, I’m guessing that numbers for the likes of BA, LH, AF and their recent acquisitions are roughly measuring the same kind of experiment, given that all these airlines have lots of connections and large networks.On the other hand, numbers for some other, smaller airlines are likely to be distorted heavily, and don’t lend themselves to an easy comparison. For example, it might seem surprising that an airline like Luxair (which does pretty much exclusively point-to-point traffic between Luxembourg and the rest of Europe, plus some holiday destinations, and should therefore have very little lost or delayed luggage) ends up on place 19 (losing 17.7 pieces of luggage) for the second quarter — but then again, they provide feeder services for Lufthansa in Frankfurt and Munich, for BA (and others) in Heathrow, and for Air France at Charles de Gaulle.