Re: [rfc-i] getting SVG of RFC diagrams

[Paul Kyzivat <[e-mail]>]

On 11/23/23 11:47 AM, Carsten Bormann wrote:
> On 2023-11-23, at 17:39, Paul Kyzivat <[e-mail]> wrote:
>>
>> On 11/23/23 10:01 AM, Marc Petit-Huguenin wrote:
>>
>>> XPointer can be used as a URI fragment[1].  Firefox does not implement it[2] but I found an obsolete extension that did[3], so maybe such extension is still possible.
>>
>> I would like there to be a sufficient naming convention so that someone that wants a piece of the rfc can construct the URL based on what is visible in the rendering of the RFC.
> 
> I don’t think this is attainable in general.
> 
>> That may be easy for figures that are labeled, but might be a problem for some source code such as ABNF that often is interleaved with text, without any label.
> 
> Indeed, artwork (including SVG) often is not embedded in a figure, so it doesn’t have a natural “Figure 4711” label.
> 
> (It also turns the numbered labels are rather brittle during development, so giving the artwork a name= attribute for external referencing generally is good practice.  Slugified names can stand in for that, but also are more brittle than given names.)

I'd settle for the URL being discoverable only from *some* of the 
renderings. E.g., by hovering over the item in the html rendering. That 
could either show the URL, or a tag that can be used to construct the 
URL using a standard convention.

(One of my goals is to be able to fetch ABNF from a document - to 
support validation/compilation of ABNF from document Y that depends on / 
extends the ABNF from document X. IOW #include from a document fragment.)

	Thanks,
	Paul