- From: Poul-Henning Kamp <phk@phk.freebsd.dk>
- Date: Wed, 23 Nov 2016 07:28:06 +0000
- To: Martin Thomson <martin.thomson@gmail.com>
- cc: Mark Nottingham <mnot@mnot.net>, HTTP Working Group <ietf-http-wg@w3.org>, Patrick McManus <mcmanus@ducksong.com>
-------- In message <CABkgnnXwOOmv3sJKrufs19ErUNA65iMm32aypRBH=z+F8mWGSQ@mail.gmail.com>, Martin Thomson writes: >On 22 November 2016 at 20:51, Poul-Henning Kamp <phk@phk.freebsd.dk> wrote: >Let's not talk about policy and instead concentrate on cost-benefit. >Mark has summarized the costs quite well, I think. I would go further >and point out that a recursive syntax with quoting cannot have >elements that are skipped by a parser, which I find results in simply >externalizing those costs. Have you ever implemented SOM, EOM and EOT according to spec in any of the programs you have written to deal with ASCII text ? No ? So can we agree that the cost of those three ASCII characters have been very close to zero for you ? Just because a standard contains an extension point, does not mean you have to implement that extension - unless you actively decides to use it. And I keep repeating: There should be no recursion in IETF defined HTTP headers, but we should leave the extension there for end-to-end usage. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 phk@FreeBSD.ORG | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence.
Received on Wednesday, 23 November 2016 07:28:36 UTC