Opened 11 years ago
Last modified 11 years ago
#255 assigned defect
xpra slow in desktop session (application session is okay)
Reported by: | fgnievinski | Owned by: | fgnievinski |
---|---|---|---|
Priority: | major | Milestone: | |
Component: | Xpra | Keywords: | |
Cc: |
Description
tested with both PNG and h264 encoders.
maybe it's a side-effect of the screen-size?
maybe xephyr is the culprit?
Change History (5)
comment:1 Changed 11 years ago by
Resolution: | → invalid |
---|---|
Status: | new → closed |
comment:2 Changed 11 years ago by
I looked at the above mentioned xpra ticket but failed to understand its implications.
Does it mean that only xpra app sessions are supposed to be fast and that xpra desktop sessions are known to be currently slow?
Furthermore, does it matter than I'm not shadowing a physical display, rather having a virtual display setup through xephyr?
comment:3 Changed 11 years ago by
Resolution: | invalid |
---|---|
Status: | closed → reopened |
Sorry, my bad. I assumed you were using a desktop shadow session, not a desktop session using Xephyr. (I didn't read the bug description properly)
Which type of desktop session is this? (gnome?, lxde?) Does it have lots of little animated icons?
What type of connection do you have? Have you tried plain rgb
? Changing the speed option with h264
?
comment:4 Changed 11 years ago by
it's just a static background. it's highly unusable due to interaction latency. I've tried to reduce the screen size and it didn't help. also all encoders show similar problem. in contrast, all encoders are fine in a non-desktop application-only session, given this same connection. maybe it's another xephyr bug? I'll check if downgrading it helps this too. any other ideas would be appreciated. e.g., what logs would be relevant? thanks.
comment:5 Changed 11 years ago by
Owner: | set to fgnievinski |
---|---|
Status: | reopened → assigned |
There is a lot of information here: https://www.xpra.org/trac/wiki/ReportingBugs, incidentally this is also the bug tracker for xpra, where this bug probably belongs (I don't think winswitch does anything different - though that is worth checking too)
No, it's just inefficient because it uses screen polling, see http://xpra.org/trac/ticket/390
Note: this is winswitch bug tracker. (closing)