Postel’s Principle is a Bad Idea

Post by barrkel » Tue, Feb 05, 2013, 8:44 AM UTC

tedunangst wrote: post #39

barrkel wrote: post #26

Earlier quoted context omitted.

A nontechnical user, given a choice between two environments, one of which nags them pedantically over technical details, and another which displays the gist of entered content but perhaps with sometimes screwy formatting which one would win? Word processors won out over text processors for the nontechnical user partially for this reason. Postel's law applied to HTML let nontechnical users get things done with less i…

Writing correct HTML is not significantly harder than writing crappy HTML. Comparing it to latex vs word isn't a good analogy IMO. And wouldn't nontechnical users be using higher level HTML editors anyway? At least some bad HTML comes from lazy developers who should know better, and could have done better with a stricter tool.

I'm was talking about the 90s.

Post by Arnt » Tue, Feb 05, 2013, 11:20 AM UTC

im3w1l wrote: post #59

Arnt wrote: post #53

Earlier quoted context omitted.

I've written ten RFCs of varying quality. It's terribly difficult to write something that a) gives a good overview of the subject, b) explains the choices that had to be made, c) spells out every detail, and d) remains short enough that implementers actually read all of it. All of mine fail in some way. I've heard the OSI documents failed too. Quoting one implementer, whose code did not accept non-ASCII passwords: "O…

Is there any way to request an official clarification for borderline cases the RFC-author didn't think about when typing it up?

You can submit an erratum and an author will comment and often clarify, so formally speaking the answer to your question is yes. But other implementers don't generally read the errata, so you have to expect that your interoperation peers haven't read the clarification.

Once you understand the problem, the clarification, and that your interop peers do not, I bet your implementation's handling of the issue will be conservative in what it sends and liberal in what it receives.

Re: Postel’s Principle is a Bad Idea

Re: Postel’s Principle is a Bad Idea