Re: Header field name representation, was: Rechartering HTTPbis

[Julian Reschke <[e-mail]>]

On 2012-01-26 11:37, Willy Tarreau wrote:
> On Thu, Jan 26, 2012 at 11:13:28AM +0100, Julian Reschke wrote:
>> On 2012-01-26 10:35, Willy Tarreau wrote:
>>> ...
>>> I find it pretty cumbersome to force everyone to support zlib, especially
>>> in environments where it provides no benefit (small requests/responses)
>>> and only adds CPU usage and latency. It's especially true on intermediary
>>> components which would have to decompress everything to be able to perform
>>> trivial actions such as decide what server to forward to. Using either pure
>>> binary header names or short forms would already be quite efficient.
>>> ...
>>
>> What's a binary header name?
>
> Oh I'm realizing I wrote that ! I was meaning the use of enums instead of
> headers for the common ones. For instance, we could have bytes 0x80 to 0xFF
> directly mapped to most common headers and be able to represent 128 different
> headers with a single byte, and have the other chars for the other ones.
> We could even push the principle further and have the Connection header
> apply the same rules (eg: use high bytes to reference well-known headers
> or tokens).

Ack, just clarifying.

An alternative would be to reserve two-letter header field names, and 
assign those to those headers where it makes sense...