RFB Protocol related question

[Sood, Sanjeev]

James,

I understand that viewers can request any format they wish. I am
specifically trying to understand the rationale behind asking for a
color depth greater than that specified by the client. 

The section 4 of the spec does state that if client can cope with
different formats, it may choose one which is easier for servers to
produce. 

Is that what RealVNC client viewer is following as it didn't ask for a
color depth higher than what server had specified? 

I will appreciate it if you can confirm this. Thanks,
-Sanjeev

-----Original Message-----
From: James Weatherall [mailto:jnw "at" realvnc.com] 
Sent: Wednesday, September 21, 2005 10:38 AM
To: Sood, Sanjeev; vnc-list "at" realvnc.com
Subject: RE: RFB Protocol related question

Sanjeev,

As I stated in my previous mail, no, your assumption is incorrect.
Viewers
can request any format they wish, and some do take advantage of this.

Please see the protocol specification for details on how viewers specify
the
format they require.

Regards,

Wez @ RealVNC Ltd.
 

> -----Original Message-----
> From: Sood, Sanjeev [mailto:sanjeev.sood "at" intel.com] 
> Sent: 21 September 2005 17:55
> To: James Weatherall; vnc-list "at" realvnc.com
> Subject: RE: RFB Protocol related question
> 
> Hello James,
> 
> Thanks for your response. So what you are saying is that viewers can
> request color depth greater than that of the server. In that 
> case, does
> server typically pad the extra bits/bytes to its pixel data (e.g.
> server's color depth is set to 16bits & client is asking for 24bits)?
> What is client expecting when it is asking for higher color depth than
> can be supported by a server and what is its benefit?
> 
> While running RealVNC client against RealVNC server, I observed (using
> the packet capture) that if server's color depth was set to 16bits
> (using -depth option), client started w/ 8bits and increased the color
> depth to 16 during the session, but not beyond that. Then I 
> changed the
> color level to Full in the client options and saw client 
> directly start
> w/ 16bits and not ask for any higher color depth. 
> 
> When I had set the VNC server to 24bits color depth, the 
> client started
> with 8bits and then increased it to 24bits color depth. 
> 
> So is my assumption correct here that atleast in RealVNC client
> implementation, it doesn't ask for color depth greater than the max
> supported by the server? 
> 
> Thanks again,
> -Sanjeev
> 
> -----Original Message-----
> From: James Weatherall [mailto:jnw "at" realvnc.com] 
> Sent: Wednesday, September 21, 2005 2:28 AM
> To: Sood, Sanjeev; vnc-list "at" realvnc.com
> Subject: RE: RFB Protocol related question
> 
> Sanjeev,
> 
> Yes, viewers can request pixel formats that are "larger" than the
> server's
> pixel format.  i.e. a viewer on a 32bpp desktop can request 32bpp even
> though the server is only doing 16bpp.
> 
> Regards,
> 
> Wez @ RealVNC Ltd.
>  
> 
> > -----Original Message-----
> > From: vnc-list-admin "at" realvnc.com 
> > [mailto:vnc-list-admin "at" realvnc.com] On Behalf Of Sood, Sanjeev
> > Sent: 21 September 2005 02:46
> > To: vnc-list "at" realvnc.com
> > Subject: RFB Protocol related question
> > 
> > We are implementing VNC server on an embedded platform and 
> > are using the
> > latest RFB protocol spec (version 3.8, Dated 8 July 2005). I have an
> > implementation related question which I am not sure of from 
> the spec:
> > 
> > *	Spec allows client to change the pixel format and requires
> > servers to support the client requested format. Being an embedded
> > device, our RFB server can only support up to 16bit color. 
> My question
> > is: can a VNC client ask for a color depth greater than 
> that specified
> > by the server in its server initialization message?  Or do 
> clients do
> > not exceed the max color that was specified by a server? Would our
> > implementation violate the protocol if we close the 
> connection in such
> > an instance? In my test I observed that RealVNC client 
> doesn't exceed
> > the server's max color depth, but would like to verify that.
> > 
> > 
> > 
> > I will appreciate any clarification you can provide on this.
> > 
> > 
> > 
> > Thank you very much,
> > 
> > -Sanjeev
> > _______________________________________________
> > VNC-List mailing list
> > VNC-List "at" realvnc.com
> > To remove yourself from the list visit:
> > http://www.realvnc.com/mailman/listinfo/vnc-list