Re: SPDY Header Frames

On 15/07/2012 10:05 a.m., HAYASHI, Tatsuya wrote:
> Dear Willy,
> Response inline.
> On Sun, Jul 15, 2012 at 4:22 AM, Willy Tarreau <> wrote:
>>>> On a personal taste, I find it fast too. 4 months to provide proposals
>>>> to replace the 15-year old HTTP/1, and 4 others to review them is short
>>>> in my opinion. Roy did not even have the time to publish the Waka spec
>>>> which could have brought a lot of fuel to the discussion !
>>> +1.
>>> I think that it is an important point.
>>> Should we make what is replaced with HTTP1?
>> I'm not sure what you mean.
> Sorry.
> "HTTP/2.0" are the marks of a big change for me.
> Probably it is the same also for many Web programmers.
> For example, big change excited like "Waka".
> I had imagined "Version Number" such.
> So, My "replaced" was putting it.
> In this short time, I think that it is difficult.

"Big change", but in what way? ....

If the next protocol were to be called HTTP/1.2 we would be forced to:
- use plain text lines limited toASCII charset ending in CRLF
- use full-name headers in human readable format
- support all the human-readable date formats, mime type formats, 
options names etc.
- use strict one request message followed by one response body stateless 

In short, we would be forced to NOT fix any of the real performance 
problems visible in HTTP/1.

"2.0" is a BIG change for the incoming stream parsers and outgoing byte 
packers. Flow control, input validation, messaging and potential future 
extensions are also affected in a big way. With just binary framing 
change, no new features, no new headers, no new anything else but framing.

Hopefully the result will be a BIG change towards simplicity. But only 
time will tell about that.


Received on Sunday, 15 July 2012 06:38:36 UTC