[rfc-i] Notes on "submission format"

Brian E Carpenter brian.e.carpenter at gmail.com
Fri Sep 21 16:33:17 PDT 2012

On 21/09/2012 23:09, Paul Hoffman wrote:
> On Sep 21, 2012, at 2:40 PM, Dave Crocker <dhc at dcrocker.net> wrote:
>>>    Submission format = the format submitted to the RFC Editor by
>>>    authors.
>>>       *  might not be the same as the canonical formats (though it would
>>>          make the workflow somewhat simpler for the RFC Editor if it
>>>          were);
>>>       *  will be converted to another format for further processing and
>>>          publication if necessary
>>>       *  Currently: .txt (required), XML (optional), NROFF (optional)
>>> =====
>>> "Authors" do not submit to the RFC Editor: stream managers do.
>> Strictly speaking, reference to the actor doing the submission is not needed; as demonstrated here, it's even distracting.
>> So, neutral language would work better, such as:
>>   = the format submitted to the RFC Editor for publication


> I prefer to keep the "stream manager" in because some people in the earlier discussion conflated two different ideas: "the format I turn in Internet Drafts in my intended stream" and "the format the stream manager would turn in to the RFC Editor". It is plausible that if the submission format was X in the future, Stream Y might accept Internet Drafts in format X and Z, but would covert Z to X when they are ready to become RFCs.

In the real world, the author or document editor submits the file(s) to a tool
that puts them in the tracker database, and the RFC Editor pulls the file(s)
from there. The stream manager just sends an approval notice.

We're computer scientists, so we know the difference between call by name
and call by value, but it's an implementation detail. I think that
Dave's passive tense formulation is true for all possible implementations
and therefore better.


More information about the rfc-interest mailing list