W3C home > Mailing lists > Public > ietf-http-wg@w3.org > April to June 2019

Re: Empty lists in Structured Headers (#781)

From: Roy T. Fielding <fielding@gbiv.com>
Date: Thu, 2 May 2019 09:25:39 -0700
Cc: ietf-http-wg@w3.org
Message-Id: <627257EE-FE78-40A6-AA91-9E488C53A8FC@gbiv.com>
To: Martin Thomson <mt@lowentropy.net>
> On May 2, 2019, at 2:55 AM, Martin Thomson <mt@lowentropy.net> wrote:
> On Thu, May 2, 2019, at 15:18, Mark Nottingham wrote:
>> <https://github.com/httpwg/http-extensions/issues/781>
>> PHK and I have discussed this, and I think we agree that this issue 
>> should be closed without any change to the specification. 
> I agree.  An empty value is a footgun and that's a big part of why we're doing this spec.

In what way is an empty field value a footgun?

The reason why we started doing this spec was to allow field values to encapsulate other
field values in a regular way.  That is, given any arbitrary field definition foo, one could define

    More-Foo: #container{ foo }

and not trip over the syntax in horrible ways. That's why it started as JSON.

Then the spec morphed into a mechanism for self-descriptive binary content and more
efficient date formats that could be mapped back to HTTP/1 in less efficient ways. IOW,
something that might be forward-useful. Unfortunately, that seems to have been abandoned.

Now it is just a set of arbitrary decisions that have nothing to do with HTTP and cannot be
defended without falling back on a straw poll on how tired the editors are of receiving
further comments.

This is not how the IETF is supposed to work on standards track proposals.

Received on Thursday, 2 May 2019 16:26:20 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:15:34 UTC