[rfc-i] Referring to specific paragraphs, Re: draft-rfc-image-files-03

Iljitsch van Beijnum iljitsch at muada.com
Wed Apr 11 03:20:04 PDT 2012

On 11 Apr 2012, at 12:07 , Julian Reschke wrote:

> If we insist on line numbers, than we can essentially stop the format discussion, as having line numbers implies fixed formatting.

As I was saying, the need for this type of precision referencing means the author of the referred to text did a bad job.

If you really need to be super precise you can always quote what you're referring to.

>> Coming up with all kinds of clever stuff is exactly what makes XML2RFC so hard to use. Yes, it's very clever that the tool "knows" that it can form "F. Author" from<author initials="F" surname="Author" fullname="First Author">, but this clever scheme doesn't allow for the fact that my last name is capitalized as "Van Beijnum", "I. van Beijnum" and "Iljitsch van Beijnum" so the RFC Editor had to go in and manually correct for this.

> I don't think this has anything to do with the thing we were just discussing :-)

I think it is very relevant to the larger discussion. Because it shows how a format can be designed to do something, but then end up doing it badly. What XML2RFC should really have done is have a full name, a last name with initials and a last name without initials. This would have been an 80% solution. With that, XML2RFC wouldn't be in the situation where it really has to support all the different ways to handle names that are in use around the world (a 100% solution), then (obviously) fail at that (an 80% solution pretending to be a 100% solution) and thus requiring workarounds.

Also interesting in this regard are the discussions about what a publication date means in RSS. I believe ultimately they punted on this and said "it's what the publisher feels like putting in the publication date field".

We can only enforce exactness when we can have exact correctness. Exact incorrectness is a big problem. In other words: less is more.

More information about the rfc-interest mailing list