[rfc-i] List aesthetics and run-on numbering

Tony Hansen tony at att.com
Mon Apr 21 20:25:14 PDT 2014


On 4/21/14, 2:11 PM, Elwyn Davies wrote:
> On Mon, 2014-04-21 at 13:15 -0400, Paul Kyzivat wrote:
>> How about adding another attribute to a list, 'continuation-of' that
>> references the anchor of a prior list. The processor could then possibly
>> apply some heuristics. At a minimum it could continue numbering of
>> numbered lists. It might also fill in defaults for unspecified
>> attributes from the prior list. And it might try to adjust indentation
>> levels to match.
>>
>> 	Thanks,
>> 	Paul
> My initial reaction was this was a neat idea, but then I got to thinking
> about editing and maintenance of the source.  You'd have to have
> different anchors for each segment of the list and maintain the chain of
> anchors in order, I assume; that is not so attractive.  I think I'd
> rather go for a scheme where the segments of the extended list share a
> group name.  This has advantages in that it doesn't require any work if
> segments are shuffled in the source and if the first segment has the
> group name it cues the processor to expect more segments so it can do
> anything extra that might be needed like saving counters and calculating
> indent over multiple list segments.

The other aspect of groupings that we currently have with the v2 
processor is being able to specify the format of the grouped bullet, 
such as a prefix and suffix to the numbering. For example, a number of 
RFCs use a format of something like

     Req. %d -

When combined with the v2 equivalent of names, this produces "bullets" of

     Req. 1 -
     Req. 2 -
     Req. 3 -
     etc.

This is not possible with the current v3 definition.

How important is this capability? (I personally feel that it is very 
important for certain types of RFCs.)

     Tony Hansen


More information about the rfc-interest mailing list