[rfc-i] Making the XML available
Joel Halpern Direct
jmh.direct at joelhalpern.com
Tue Mar 5 14:05:59 PST 2019
As far as I know, (3) (Making the XML as avaiable as any other form) is
indeed the plan.
On 3/5/19 3:50 PM, Kent Watsen wrote:
> Hi Joel,
> Yes, (1) and (2) are from RFC 7990, but I've seen nothing yet regarding (3).
> Of course, if (3) is in motion already (e.g., as part of the v3 rollout), then
> this is just what I'm asking for. Is this the case?
>> On Mar 5, 2019, at 3:39 PM, Joel M. Halpern <jmh at joelhalpern.com> wrote:
>> We are in the process of shifting to make the XML
>> 1) The normative form
>> 2) used to generate all other forms
>> 3) Available directly
>> As such, why would we mess with trying to adjust / improve the current process when we hope to get away from it ASAP.
>> On 3/5/19 3:36 PM, Kent Watsen wrote:
>>> The request is if the IETF could also make the final (post Auth48) XML for RFCs available?
>>> The motivation is twofold:
>>> 1. To better support extracting content from RFCs.
>>> * current approach is to extract content (i.e. source code
>>> inclusions, such as YANG modules) from the plain-text, but this
>>> is not ideal and, not for lack of trying, does not result in
>>> perfect extractions (indents, outdents, extra lines, etc.).
>>> 2. To better support kick-starting a "biz" draft.
>>> * current approach is for authors to use the last posted revision
>>> of the draft, that which the IESG approved, but this version
>>> does not include any changes made by the copy editors.
>>> It is understood that, today, the XML is used to generate the `nroff` format, and that final adjustments are made there. It is assumed that the adjustments being made there are minor (e.g., pagination, flow, etc.) as, from experience, XML is used throughout the Auth48 process.
>>> The request is to make this last XML version available.
>>> Since this XML is not the final version, there may be hesitation to publish it as another official "output" format. To address this inconsistency, could language surrounding the link (e.g., a click-thru popup) be used to disclose it?
>>> Alternatively, maybe no link is required, so long as it's well-known how to construct the URL to access it. For example:
>>> Given that:
>>> * Text ----> https://tools.ietf.org/rfc/rfcXXXX.txt
>>> * PDFs ----> https://tools.ietf.org/pdf/rfcXXXX.pdf
>>> * HTML ----> https://tools.ietf.org/html/rfcXXXX
>>> So, maybe:
>>> * XML -----> https://tools.ietf.org/xml/rfcXXXX.xml
>>> * or -----> https://tools.ietf.org/xml/rfcXXXX.xml
>>> If only a well-known URL were made available, extraction-tools and bis-authors (per 1 and 2 above) would have what they need, without exposing the XML to the general public.
>>> PS: I'm unsure if this is the best place to ask this question, so please bounce this message to another list if needed.
>>> rfc-interest mailing list
>>> rfc-interest at rfc-editor.org
More information about the rfc-interest