Re: Who implemented which version of the compression draft?

On 8/08/2013 10:11 a.m., Brian Raymor (MS OPEN TECH) wrote:
>
> MS OpenTech (Katana) implemented -01.
>
> Fred, here’s the link to related emails on the list - 
> http://lists.w3.org/Archives/Public/ietf-http-wg/2013JulSep/0440.html 
> and earlier 
> http://lists.w3.org/Archives/Public/ietf-http-wg/2013JulSep/0256.html
>
> Which suggests that it could be helpful to merge the compression draft 
> into the http/2 draft? - https://github.com/http2/http2-spec/issues/224
>

But not if we are still at odds about even whether this particular 
compression is the true way forward. There is still a lot of memory 
hungry statefulness and processing overheads being placed on middleware 
simply to multiplex the streams between client and server connections. I 
have mentioned the mandatory compression/decompression as being the 
worst hurdle already.

I have been thinking we should have been adding a compression indicator 
to the declared version string. At least during the experimentation 
period so we can identify http2-draft-04-z00 from http2-draft-04-z01 
from http2-draft-04-b03 from http2-draft-04-b04 or whatever.

Amos

> …Brian
>
> Brian Raymor
>
> Senior Program Manager
>
> Microsoft Open Technologies, Inc.
>
> A subsidiary of Microsoft Corporation
>
> *From:*Fred Akalin [mailto:akalin@google.com]
> *Sent:* Wednesday, August 7, 2013 11:00 AM
>
> The spec text I mentioned is discussed in 
> https://github.com/MSOpenTech/http2-katana/issues/19 .
>
> On Wed, Aug 7, 2013 at 7:54 PM, Jain, Shakesh <shjain@akamai.com 
> <mailto:shjain@akamai.com>> wrote:
>
>         Actually, that particular issue may just be a difference in
>         wording.
>
>         I'm still curious, though, about who implemented what version.
>
>     At Akamai we implemented draft –00 (with at least one bug to fix
>     in our implementation). I could not make it to interop so can you
>     please share what part of the spec you are talking about?
>
>     Shakesh
>
>         On Wed, Aug 7, 2013 at 7:35 PM, Fred Akalin <akalin@google.com
>         <mailto:akalin@google.com>> wrote:
>
>             Hey guys,
>
>             There are two versions of the compression draft: -00 and
>             -01. httpbis-draft-04 refers to -00, but I vaguely
>             remember an e-mail going out previously saying that we
>             should work off -01. However, I forgot about that and
>             worked off -00.
>
>             Out of curiosity, who implemented -00 and who implemented
>             -01? I'm asking because there's at least one important
>             difference (see
>             https://github.com/MSOpenTech/http2-katana/issues/19 ) and
>             I'd like to know which one to refer to to get the widest
>             interoperability.
>
>             -- Fred
>

Received on Thursday, 8 August 2013 03:39:39 UTC