Rupert Goodwins' Diary

Monday 15/03/2004
It's the most challenging day of the high-tech news gathering week. The mighty machinery of the West Coast has been at the beach for the weekend, phones sit silently on desks and there's only a light dusting of press releases in the inbox for inspiration. So this hardy searcher after truth hits the Web, rattling along dusty lanes leading to university research departments, the better class of blog, technical forums.

And what's this? "Researchers in North Carolina have developed a data transfer protocol for the Internet that achieves speeds roughly 6,000 times that of DSL"? Heavens! I get quite excited for a Monday, and don't quite engage the little grey cells. A real news story. The usual claims are there in the press release – download MP3s in the blink of an eye, tremendous implications for supercomputing and science, blah blah blah.

But how does it work? Can't write the story without that...

Ten minutes of hard labour untangling the terminally confused press release later, and the truth is depressingly clear. The new protocol is just that – a protocol. It goes 6,000 times faster than DSL when you run it on a gigabit fibre that's, um, 6,000 times faster than DSL to start with. The real story is that Internet protocols aren't that good at working over very, very fast links – and this invention is somewhat better. Which is fine if you're writing news for IP Week and TCP Review but not really the meaty stuff you, the beloved reader, expect as your right. No good.

It's early in the week, but I can swear. And I do. Visions of gigabit DSL vanish, to be replaced by a mental picture of a press release writer swinging gently in the breeze while enraged researchers and frustrated journalists prod at the corpse with electrified pokers.

It's the only language these people understand.