HTTP-Headers Status Diagram

thoughtpad.net

59 points·luccastera··6 comments

If ever there was an existence proof of why flowcharts disappeared from engineering discourse, this is it. Seriously, which would be easier to understand: this monster or the 80 lines of code that it would take to actually implement it?

Web programmers everywhere should put this on their cubicle wall, to convince people how difficult their job is. (I saw a seven-layer-networking-cake diagram which served a similar purpose).

Web programmers can actually completely ignore the lion's share of the complexity represented in this diagram.

This concerns only web server implementors... and even then, a lot of web servers simply don't need to support the full set of operations.

Web programmers who follow your advice usually sound like idiots who don't know much about what they're doing. Lots of them like to reimplement wheels like HTTP caching, HTTP authentication and content negotiation, even error codes (!). I have grown tired of stupid shit web "programmers" without knowledge of HTTP are capable of generating. Especially when paid by an hour.

Relax... First, I had HEAD, PUT and DELETE in mind as stuff you usually don't need to support explicitly as a web developer...

I admit I went a bit overboard with saying "can completely ignore the lion's share" but I obviously didn't mean you should go out of your way to skip the absolute basics that caching, content negotiation, error codes etc. are.

I wasn't "advising" against learning HTTP, I just think lots of web developers get away with not learning much about it (I'm NOT saying that's a good thing.)

edit: Damn, the more I reread my previous post the stupider it sounds :|

Honestly, that's way more than I ever wanted to know about HTTP header status.

But certainly useful.

Reply on news.ycombinator.com