[Spice-devel] spice-html5

[Алексей]

Hello,

I was using Chrome of up-to-date-at-that-time-version and i used my own page with some nodejs-based design. The main difference from default spice-html5 page was that parent of spiceDisplay window had zero offset from browser window and in my page spiceDisplay window is in another window which is overlapping over some other elements and offseted from global window. So the total offset is summed from parent1.offest+parentofparent1.offset+...
It is the problem that my patch fixed.
About the motion problem - i figured out it is not html5 problem, because i have the same problem in virt-manager window.
I found this bug described on bugtracker and marked as fixed: https://bugzilla.redhat.com/show_bug.cgi?id=867885
But in fact the bug is not in spice-gtk (or even spice-glib), because i removed them, rebooted, and virt-manager works and still have problems with mouse motion.
My guess is that the bug is somewhere in spice-server, but at the moment i didn't find in which file/function.
Hope my explanation made things a bit more clear.

Alex.

Пятница, 12 апреля 2013, 10:47 -05:00 от Jeremy White <jwhite at codeweavers.com>:
>Hi Alex,
>
>On 04/08/2013 07:38 AM, Алексей wrote:
>> Hello!
>> 
>> I'm very interested in your spice html5 project, so i tried to use it and made a virtual windows 8 machine and another fedora 18 machine. But, unfortunately, i found out that mouse movement are wrong a bit, it seems like they're using some amplifier about 1.2x to move and on fedora there was no movement at all!
>> Looking inside the source helped me fix the no-move problem of fedora: the window didn't know it could have more then one parent, so i made little reccuring function that fix the problem, but i couldn't fix the amplifiying mouse movement problem.
>> I think the problem is about positioning: in fedora spice-html5 uses "SpiceMsgcMousePosition" function and in windows it uses "SpiceMsgcMouseMotion" function, which is probably the main problem. Also, i found out that the look-a-like method does have spice-gtk, and motion method accept dx and dy instead of absolute values.
>> So, would you be so kind to find the missing detail between dx and dy and motion\position functions?
>> Sorry for my bad english and disturbance, after all.
>> P.S. Attached the fix of absolute positioning bug.
>
>Thank you for the email, and for testing!
>
>Would you mind if we moved this conversation to the public mailing list?
> That way, it can benefit others as well, and we can benefit from their
>comments as well.
>
>If you don't mind, if you could reply and include
>spice-devel at lists.freedesktop.org in the response, I would appreciate it.
>
>With that said, I am curious: what browser were you using?  And also,
>did you use a different html page such that the canvas had a different
>set of parents?  Can you explain how that was different?
>
>Thanks,
>
>Jeremy

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/spice-devel/attachments/20130412/aee9d144/attachment.html>