The IT Crowd Was Right – What I learned by reading a lot of RFCs

Due to a change of job I’ve recently had to teach myself about networking.

RFCs had always been a bit of a mystery to me but since they came up over and over again when reading about network concepts I thought I’d familiarise myself with them as a whole.

With a quick:

apt-get install rfc-doc

an organised set of RFCs was downloaded and categorised into various folders under /usr/share/doc/RFC. I looked closely at these three:

best-current-practice
for-your-information
standard

and skimmed the rest:

draft-standard, experimental, historic, informational, links, old, proposed-standard, queue, unclassified
Here’s some of the many things I learned by looking through them and reading a good proportion of the active ones:

– There’s a ‘Service Location Protocol’ specification (RFC2608) which anticipates the need for scalable service discovery. Which begs the question: why are we all reinventing the wheel now? There are implementations already written and available (slpd, slptool). Beats me.

– There’s a very handy glossary of internet terms which is still useful (RFC1983), even though written in 1996, is still useful.

– They’re very well written. Really basic things like NFS (RFC1094) and UTF-8 (RFC3629) are explained in a clear and straightforward way.

– There’s nothing quite like dropping the phrase ‘if you read the current RFC on the subject…’ into a meeting.

– The IT Crowd was right – the ‘elders of the internet’ really do exist (RFC1462 – What is the Internet).

Who Governs the Internet?

In many ways the Internet is like a church […] It appoints a council of elders, which has responsibility for the technical management and direction of the Internet.

I’m off to Big Ben to commune with Stephen Hawking.