Hi Alexis, (Resending to the list, too :-)
On Wed, May 14, 2025 at 1:27 PM Alexis Rossi <[email protected]> wrote: >> >> >> On Wed, May 14, 2025 at 3:59 AM Carsten Bormann <[email protected]> wrote: >> On 14. May 2025, at 11:09, Michael Richardson <[email protected]> wrote: >>> >>> We needed an abstract that was higher level than the markdown tables. >> >> I often generate tables from CSV files. >> E.g., >> >> https://github.com/cbor-wg/draft-ietf-cbor-cde >> >> (Files with “table” in the name.) >> >> In this case, this is a CSV, which could also be made available with the RFC. >> > An HTML table is generally accessible for screen readers, which know how to > tab between cells. Is the CSV for people who use the TXT file? Or maybe I'm > misunderstanding? This observation is less about screen readers, but more about the accessibility of large tables to average human readers (*). Running code on them can help unearth structure that simply can’t be expressed in the form of a table. Or, simply looking at the (unformatted) content of the table in Apple Numbers or Microsoft Excel... This is simply a continuation of the observation that we need to have a way to embed the source of an illustration in an RFC. >> In other cases, more happens in the generator program, and pointing to the >> repo might be a better way to make more complex generation schemes available. >> > If it's normative, I think it wouldn't be a good idea for an RFC to point at > a repo that might disappear (or change dramatically). In this particular case, the tables are for appendices with examples. But even if the tables are normative, having the source available that was used to generate them can be very useful, so absent a source embedding mechanism, the text currently says: >> Implementers that want to use these examples as test input may be interested >> in the file example-table-input.csv in the github repository >> cbor-wg/draft-ietf-cbor-cde.¶ Grüße, Carsten (*) It doesn’t help that the RFCXMLv3 support for tables is so anemic that the author has no chance to guide the table rendering in a way that the result is readable. Calling the information a renderer would need for that by the name of “style” is rather misleading. _______________________________________________ rfc-interest mailing list -- [email protected] To unsubscribe send an email to [email protected]
