Re: Server Push used for long polling

On Thu, Jan 15, 2015 at 11:30 AM, Greg Wilkins <gregw@intalio.com> wrote:

>
> On 15 January 2015 at 20:23, Wenbo Zhu <wenboz@google.com> wrote:
>
>> Wouldn't you need an open (i.e. pre-opened) websocket stream to do this?
>> IMO, spontaneous server-push (i.e. without a client-initiated stream) is a
>> unique feature of PP, which is orthogonal to the problem of streaming data
>> from the server to client.
>
>
> Wenbo,
>
> I'm not saying that PP is not needed.   It looks well designed for purpose
> and that is pushing associated resources to a HTTP request. PP is one of
> the key features in HTTP2 that make it worthwhile deploying now despite all
> it's warts.
>
> What I don't think it is good for is replacing long polling as a way of
> providing a server to client messaging transport.
>
Agreed. Also note that long-polling is not necessarily the most efficient
way for providing server to client messaging, i.e. the RTT overhead is a
problem of long-polling, not HTTP/2.



> That is what websockets has been designed for and we should look to how
> websockets can be well carried over HTTP2 rather than how we can trick HTTP
> Push into being a messaging transport for non HTTP messages.
>
> cheers
>
>
>
> --
> Greg Wilkins <gregw@intalio.com>  @  Webtide - *an Intalio subsidiary*
> http://eclipse.org/jetty HTTP, SPDY, Websocket server and client that
> scales
> http://www.webtide.com  advice and support for jetty and cometd.
>