Joe Hildebrand (jhildebr)
jhildebr at cisco.com
Mon Aug 6 10:44:24 PDT 2012
On 8/6/12 11:26 AM, "Dave Crocker" <dhc at dcrocker.net> wrote:
>On 8/3/2012 11:02 PM, Joe Hildebrand (jhildebr) wrote:
>> On 8/3/12 2:30 AM, ""Martin J. Dürst"" <duerst at it.aoyama.ac.jp> wrote:
>> Once you can have a brand that goes beyond the capabilities of the
>> lineprinter format, we might decide that the differences in brand are
>Consonant with the note I posted a few minutes ago, I'd like to raise a
>very large flag with the above:
> It is stating that there will be different information in
>different formats (presuming original formats remain in the repertoire.)
No, that's really not at all what I was proposing. I'm proposing that
individually-submitted documents look a little different from
consensus-based documents in the same format.
Now, there *will* be differences for a given document between the
different formats. For formats that support all of the required features,
those differences are *hopefully* minor, particularly if the different
formats are generated mechanically from the same source. If hand-editing
is involved, all bets are off.
The lineprinter format will always lose information, however.
The potential for these minor differences is one of the reasons why there
are folks that would like to have one format marked as canonical.
>Branding, and the like, well might be far more visually appealing and
>more functionally useful, in one form versus another, but we do not need
>to restrict its occurrence to only one.
>The fact that an RFC indicates its standards status in text in one form,
>and might indicate via text and font change and color in another makes
>the latter friendlier, more broadly accessible, and the like, but it
>doesn't change what 'information' is present.
The branding is information. It could give the reader some indication of
the intended quality of the technical content of the document, for example.
>Current txt-form RFCs distinguish the stream, for example, in line 1,
>upper left corner. And 'status' in line 4, left-hand-side.
Correct. The questions are:
a) is that enough?
b) should we do more even if it is?
>These aren't the most visually friendly and alternate rendering forms
>can be made more helpful for noting such distinctions. But, again, that
>doesn't mean one contains the information and another does not.
Agree. But keep in mind the question that I'm asking. If you *can* do
better, *should* we do better? I'm arguing "yes", but not strongly. For
example, I'd be open to an argument that the core brand of our output is
"RFC", not "IETF", so the gradations don't matter. That argument would
likely need to withstand some scrutiny by the IETF community; those that
understood the fine distinction might have strong opinions.
(note: this issue is roughly aligned with the issue that Ekr brought up in
the plenary last week)
More information about the rfc-interest