Re: [Qemu-devel] [RFC 00/11]: QMP feature negotiation support

[Luiz Capitulino <lcapitulino@redhat.com>]

On Mon, 25 Jan 2010 15:33:40 +0100
Markus Armbruster <armbru@redhat.com> wrote:

> Anthony Liguori <anthony@codemonkey.ws> writes:
> 
> > On 01/21/2010 03:09 PM, Luiz Capitulino wrote:
> >> """
> >> {"QMP": {"capabilities": ["async messages"]}}
> >>
> >> { "execute": "query-qmp-mode" }
> >> {"return": {"mode": "handshake"}}
> >>
> >> { "execute": "change", "arguments": { "device": "vnc", "target": "password", "arg": "1234" } }
> >> {"error": {"class": "QMPInvalidModeCommad", "desc": "The issued command is invalid in this mode", "data": {}}}
> >>
> >> { "execute": "async_msg_enable", "arguments": { "name": "STOP" } }
> >> {"return": {}}
> >>    
> >
> > Maybe:
> >
> > enable-capability "async messages"
> > disable-capability "async messages"
> >
> > I think that's a bit more obvious and it means that a client doesn't
> > have to maintain a mapping of features -> enable functions.  It's also
> > strange to use an enable command to disable something.
> 
> Agree on both counts.  But why two commands?  Why not simply "capability
> NAME VALUE"?  Works even for non-boolean capabilities.  I'm not
> predicting we'll need such capabilities.

 I slightly prefer two commands because that's probably how I'd
write it in a program (ie. two functions), also enabling/disabling
a group of features is a bit easier too, as we can use an array:

capability_enable [ "foo", "bar" ]

 Now, only one command is not terrible difficult, but we would
have to accept an array of objects, like:

[ { "name": "foo", "enabled": true }, { "name": "bar", "enabled": true } ]