ID-nits [was Re: suggestions (voting)]

[Fred Baker]

Leif Johansson wrote:

> Why do we need the ID-nits document, or rather: why does it need
> to be so long? Obviously part of the problem is that people are
> unable to write correctly formatted text (hence the word nits). It
> would be interesting to know just how much time is spent by all
> involved doing clean-up usually left to software. I.e is this a
> real problem or not?


As one of the authors of the document, I will tell you why I wrote it. 
You can decide for yourself whether it is a real problem, or whether the 
document is a good solution.

The IESG, during my tenure, received many drafts that were expected by 
their authors and working groups to become instant RFCs. Very few of 
these passed must without being edited, and it seemed to me that the 
reasons were very repetitive. I, frankly, got tired of seeing drafts 
sent back to working groups to ask them questions related to security 
("have you considered man-in-the-middle attacks?" etc.), or asking them 
to consider including documents they referred to in the bibliography, 
etc. So Thomas Narten and I sat down and tried to write down the 
repetitive issues in one place, with a view to asking people to address 
the obvious issues before sending the document to the IESG.

If you go read the page, I think you will find that some of the issues 
can be fairly substantive, and some are well described as "nits".

The problem that is real is that the IESG is routinely asked to publish 
documents that don't pass muster. I'm perfectly willing to hear that the 
id-nits page is the wrong solution, but I would like to hear a better 
solution suggested at the same time.