John Pescatore

Back in the day, I spent 11 years working for GTE. For those of you who don’t remember back in the day, GTE was a telephone company back in the regulated days when phone companies had to provide reliable dial tones and were measured on the availability of their services. Back in the day, telephone people sneered at the Internet and its lack of availability guarantees or service level agreements. Of course, they also sneered at how unreliable cable TV was – and absolutely guffawed at the idea that the Internet or cable TV could every carry reliable data services, let alone voice services.  

What the telephone people couldn’t accept was the idea that “ooh, error 404 – I better try again” might become acceptable. But it did – today’s Internet users accept pretty crappy voice and data service because it is cheap. It is basically the “buffet mentality” – the food isn’t very good but it is all you can eat!!

So, given that every Internet user has gotten used to error 404 when web browsing, and never once blames the IT shop, I’d like to see some innovative use of other HTTP error codes.  If a web server was smart enough to have some doubts about the legitimacy of a connection, why not send something like Error 408 (Request Timeout) and see if the user tries again? Something like that could help differentiate bots or spiders from humans at least – or at least until the bad guys caught on and then maybe we would switch to Error 418 (You are Ugly and Your Mother Dresses You Funny) until they figured out that one, then we would switch to Error 420 (Meet Me Behind the High School), etc.

I’m sort of over-simplifying here but the idea is to take advantage of how accepting users are of crappy Internet service as long as it does not appear to be the fault of IT.  Automated code isn’t so tolerant – do you think a bot every actually tried to click on the duck flying by in a banner ad?

4 Comments »

Category: Uncategorized     Tags:

Share