I've never fully read an RFC before, so I've always had this bias towards software or libraries that implement the RFCs. "Why don't they just follow what's laid out in the RFC", "Code to the standard" and other uninformed remarks like that.

Well, this last week I've had to read an RFC to implement a code library for a project at work. It's been a rather eye opening experience too.

First of all, from what I can tell, this RFC is not overly complicated. Compared to long RFCs like SMTP and just about anything email, this one is rather tame.

Second, RFCs can be extremely vague or extremely general at some points. This can be problematic when you're trying to write code. Vague instructions make coding very difficult.

Third, alternatively, RFCs can be extremely specific. Say what? Well, they are usually long documents, and they need to be specific enough to express the idea they were written for. Specific instructions are very helpful when coding. They help to reduce problems of ambiguity and help tear down roadblocks. They make decisions easier as well.

This particular RFC, thankfully, was more specific than it was vague. It makes a strong attempt to speak in UML-ish language and terms. This makes things much easier on me since I can directly translate those terms and language to code.

The progress has been steady so far. There have been several points of confusion when doing the implementation since some of the terms I've come across are totally foreign to me (like what is IBM SNA? Everyone I work with knows what it is except me...but then again I'm rather young).

I expect it won't take a terrible amount of time to finish.