Re: LC on draft-mankin-pub-req-08.txt

Reading this, a few items caught my eye.

The POSTEDIT requirements seem to be worded as if it is desirable to 
minimize the changes that the document editor makes, or even the 
changes the document editor can make.  The general tone of "don't 
mess with the words we have carefully honed" is 
understandable.  However, in practice many of the words have not been 
carefully honed.  And they need to be.  For example, there is a 
document I just reviewed to which my personal reaction is "this needs 
massive editing."  It is not technically wrong.  But the language use 
makes it hard for the reader to understand what is intended.  I would 
sincerely hope that if it is approved as-is by the IESG that the RFC 
Editor would edit the document.
In general the editor has little or no way to tell which words are 
"carefully crafted."  I would hate to have a presumption that all the 
words a sacrosanct.
I realize that the text calls out the special case of "don't touch a 
letter of this", and even acknowledges that it is a rare case.  But 
the wording of the earlier text is not in line with 
that.  Specifically, POSTEDIT-4 reads "The IETF Technical editor 
should refrain from changes to improve readability that may change 
technical and consensus wording."  This appears to be a directive 
that prohibits almost all changes, since in a formal sense all the 
words in an WG and IETF LC approved document are "consensus 
wording."  That leads to what I consider a bad situation where we 
have essentially prohibited the editor from editing.
On a related note, POSTEDIT-3 seems to be inadvertently worded too 
strongly.  It prohibits changes which "introduce a substantial review 
load but only provides incremental increase in the clarity of the 
specification."  However, by definition any change at all, even a 
significant change that transforms a document from unintelligible to 
highly readable, is always an "incremental increase in the clarity of 
the specification."
With regard to the metrics, I think that it would be helpful to 
separate the notion of having metrics from the specific values.  I 
would suggest moving the specific values to an appendix, with a 
notation that these values are advisory and based on IETF perception 
at the time of writing.  I don't want to lose the numbers, but I 
think that they have a different status as requirements than the 
point that these time frames should be tracked, and should have well 
understood targets.  Separating this also allows for negotiation of 
cost-benefit tradeoffs without violating "requirements."

As a minor matter, figure one is trying to make a useful statement, 
but one of the headings caused me to have to spend more time staring 
at the figure, rather than making things clearer.  In the row labeled 
"Actors", WGLC and IETF LC appear.  Those are states, not 
actors.  Also, the action listed (Formal Reviewing) does not, as far 
as I know, currently occur during those phases.  The formal reviewing 
occurs after IETF LC ends, during IESG deliberations.
Yours,
Joel M. Halpern


_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf