[rfc-i] Review of draft-reschke-xml2rfc-00

Julian Reschke julian.reschke at gmx.de
Sat Nov 9 01:49:11 PST 2013


On 2013-11-08 20:44, Elwyn Davies wrote:
> HI.
>
> I checked through the draft mainly using the knowledge that I gained in
> building
> https://www.ietf.org/proceedings/65/slides/xml2rfc-0.pdf
> and
> https://www.ietf.org/proceedings/65/slides/xml2rfc-1.txt
>
> A few comments.. clearly a lot more to do.
>
> s2.1: Note: No <...ref> elements in abstract.

Yup.

> s2.4: To the best of my knowledge area is not used anywhere.

But could provide valuable metadata.

> s2.5.1.2: I never worked out what 'name' did.  Maybe something to do
> with filenames.

Indeed.

> s2.5.1.3: Values 'abnf' - describe current effects.  Should also use
> 'mib', 'pib', 'xml' and a default.

The "current effects" depend on the processor. But yes, we want to 
describe common values.

> s2.5.1.5: The default for align used to be the same as the parent figure
> element.

Interesting; this needs a test case.

> s2.6.2: 'organization' used to be required but could be empty - required
> if attributes are empty in refs.

It's not required.

> 2.12: Worth noting *not* in 'artwork' (true for all other ...refs.

Yes, but I believe we should change this in the future vocabulary.

> 2.13: Will need to explain the arcane rules on what is in a date.

Yes.

> 2.13: Also in reference.

Nope.

> s2.17: Check: Are alt, height, width, src in both figure and artwork?

Per DTD yes. I have no idea why. We may want to check the published RFCs 
for usage.

> s2.20.1.2: [there is some inconsistency in documentation which uses
> sub-item instead of subitem.]

Where?

> s2.27.1:  Note that processors are supposed to preserve the order of the
> elements in the contents.

Are they?

> 2.34: Also mentioned in Contents.

?

> 2.38.1.1: Using the anchor in a <t> is problematic!

Setting it isn't problematic, but referring to it can be. That needs a 
whole paragraph of text...

> [And it has an index!!]
>
> Regards,
> Elwyn
>



More information about the rfc-interest mailing list