Den 2009-05-11 13:57 skrev Pierre Ossman:
> There are some corner cases in the framebuffer update system that need
> to be explicitly mentioned.
> 
> Signed-off-by: Pierre Ossman <oss...@cendio.se>
> ---
> 
> Index: rfbproto.rst
> ===================================================================
> --- rfbproto.rst      (revision 3809)
> +++ rfbproto.rst      (working copy)
> @@ -637,6 +637,12 @@
>  2               ``U16``                         *height*
>  =============== ==================== ========== =======================
>  
> +A request for an area that partly falls outside the current framebuffer
> +must be cropped so that it fits within the framebuffer dimensions.
> +
> +Note that an empty area can still solicit a *FramebufferUpdate* even
> +though that update will only contain pseudo-encodings.
> +
>  KeyEvent
>  --------
>  
> @@ -1108,6 +1114,16 @@
>  for the format of the data for each encoding and `Pseudo-encodings`_
>  for the meaning of pseudo-encodings.
>  
> +Note that a framebuffer update marks a transition from one valid
> +framebuffer state to another. That means that a single update handles
> +all received *FramebufferUpdateRequest* up to the point where the
> +update is sent out.
> +
> +However, because there is no strong connection between a
> +*FramebufferUpdateRequest* and a subsequent *FramebufferUpdate*, a
> +client that has more than one *FramebufferUpdateRequest* pending at any
> +given time cannot be sure that it has received all framebuffer updates.
> +
>  SetColourMapEntries
>  -------------------
>  

Maybe add something about the fact that it is not safe to send
SetPixelFormat messages when you have outstanding
FramebufferUpdateRequests? (which is important in the above
context...)

Cheers,
Peter

------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables 
unlimited royalty-free distribution of the report engine 
for externally facing server and web deployment. 
http://p.sf.net/sfu/businessobjects
_______________________________________________
tigervnc-rfbproto mailing list
tigervnc-rfbproto@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/tigervnc-rfbproto

Reply via email to