[PATCH 5 of 5 postargs] http: support sending hgargs via POST body instead of in GET or headers

[Augie Fackler]

> On Mar 11, 2016, at 13:00, Gregory Szorc <gregory.szorc at gmail.com> wrote:
> 
> We will probably want to leave this behavior off by default in servers
> forever, because it makes the old "POSTs are only for writes"
> assumption wrong, which might break some of the simpler authentication
> configurations.
> 
> This patch and series look correct.
>  
> I do wish we had a larger discussion about future directions of the protocol. e.g. can we use the Upgrade header so we can employ WebSockets, HTTP/2 or a custom protocol that gets us bi-directional streaming, which would be very beneficial for things like more efficient discovery.

I admire your ambition, but I don't think websockets is going to be viable for Google's needs in the foreseeable future, so we'll continue to want plain-old-http to be nicely performant.

I'm also skeptical that http/2 offers much value over http/1.1 for our use cases in Mercurial, but I'd be happy to be wrong.

> That being said, this is behind an experimental flag and I'm OK with that. Why stop progress when we have no better alternatives.

I'll want to consider making this just a [server] flag some day, and have the client support it by default if servers are willing.