- From: Mark Nottingham <mnot@mnot.net>
- Date: Fri, 27 Aug 2021 12:13:55 +1000
- To: HTTP Working Group <ietf-http-wg@w3.org>
<https://github.com/httpwg/http-extensions/issues/1606> Right now, parsing of targeted Cache-Control is specified like this: > Targeted fields MAY be parsed as a Dictionary Structured Field [RFC8941], and implementations are encouraged to use a parser for that format in the interests of robustness, interoperability and security. [...] However, implementers MAY reuse a Cache-Control parser for simplicity. The idea here is that while we'd like the interop benefits of SF, we recognise that some -- perhaps many -- implementers will have a strong motivation to reuse their Cache-Control parsers. This seems like it might be the worst of both worlds, in that we now have two possible implementation strategies. Given that there are several SF implementations now,[1] I wonder whether should reconsider this. E.g., 1) Strengthen the first MAY to a SHOULD, making the field sf-first, or even 2) Require SF, don't allow reuse of CC parsers Thoughts? Cheers, 1. partial list at <https://github.com/httpwg/wiki/wiki/Structured-Fields> -- Mark Nottingham https://www.mnot.net/
Received on Friday, 27 August 2021 02:14:14 UTC