[FFe][MIR] Please promote xcompmgr to main for LTSP
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
xcompmgr (Ubuntu) |
Fix Released
|
Undecided
|
Unassigned |
Bug Description
Recent experiment in upstream LTSP showed a significant drop of bandwidth usage and increase in performance when using compositing.
The bandwidth benefit of this change may be of as much as 50%, so very significant for LTSP users on crowded application servers.
Sadly compiz doesn't always work but xcompmgr in our tests always did.
As such, we'd like xcompmgr promoted to main so that it can be used by the next bugfix release of LTSP.
The package will end up on ubuntu alternate and edubuntu but it only weights 24K so I doubt it'll be a big problem.
LTSP would only use "xcompmgr -a" so not any of the fancy shadowing code. I tried to reproduce each of the bugs at:
https:/
And couldn't see any of these in the target environment.
Same thing with the bugs at http://
I tried moving windows around quite a lot, drag&drop, resize, minimize/maximize, locking the screen, all without a single glitch in our target environment.
The package wasn't updated in the past 2 years, however it's I think feature-complete and simple enough to be pretty much risk free.
summary: |
- [MIR] Please promote xcompmgr to main for LTSP + [FFe][MIR] Please promote xcompmgr to main for LTSP |
No updates for 2 years is a concern, but if its feature complete and does the job, I've been told that shouldn't be a blocker.
challenge will be getting MIR resources to look at it in time. It will depend on the timing of when we can get it in.
Agree it would be good to get that level of performance improvement in. Talked to jdstrand, and he said security audit wouldn't be necessary.
Approved if we can get the MIR done, and change made by April 4.