Cory, On Jun 18, 2014, at 6:20 AM, Cory Benfield <cory@lukasa.co.uk> wrote: > On 18 June 2014 10:52, Poul-Henning Kamp <phk@phk.freebsd.dk> wrote: >> and I would mandata that a request always provide the following >> fields in this exact order, as the first fields in HEADER(...): >> >> :scheme >> :method >> :authority >> :path >> :query >> >> and that responses always have the :status as the first field. >> >> My finger in the air estimate that this reduces the processing >> requirement for the "is this a trivial case" decision by about >> a factor of ten. > > This is an interesting proposal, I'll be interested to see the > response it gets from the list. This isn't problematic for hyper to > do, though having to sort the headers is a bit of a pain. I wonder if > it defeats the 'streaming' nature of HPACK, however, if we mandate > some kind of ordering from the higher layers. Given that these fields are what forms the initial request line in HTTP/1.1, I don't think it will be an issue for streaming... I'm also +1 on this, and FWIW IPP mandates something similar in its message encoding. _________________________________________________________ Michael Sweet, Senior Printing System Engineer, PWG Chair
Received on Wednesday, 18 June 2014 12:10:54 UTC