07-16-2014, 03:24 AM
You are setting the permissions manually and so this does not affect You.
Originally Posted by mike4
If that would be the case, the Pedals would not work all the time.
For the second time I ask You: did You try with Linux 3.2 (or whatever kernel version worked last)?
07-16-2014, 08:16 AM
Just tried with my script but my pedals are dead again. They worked before lunch. Need to boot windows....baahh
michael@michael-ubuntu:~$ uname -a
Linux michael-ubuntu 3.13.0-30-generic #55-Ubuntu SMP Fri Jul 4 21:40:53 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux
07-16-2014, 10:01 AM
You can not claim to have no time to properly report a bug to the appropriate projects, when a) the bug is about a gaming device, play a bit less, report a bit more, and b) it seems that you have plenty of time to sput about that bug on Phoronix.
Originally Posted by Mat2
It will not happen for me because:
- I know that Phoronix is not the appropriate place to report bugs.
- I know that bugs that don't get reported usually won't get fixed.
- I know that proposals like "When this isn't be fixed I am going back to Windows" are nothing more than poor attempts to force someone to work on the issue, providing evidence of too much self importance.
07-16-2014, 02:30 PM
Bug reports are needed, but so is bug discussion.
Phoronix is not the place to REPORT bugs, but it is a good place in my experience to DISCUSS bugs that have also been reported.
Originally Posted by Vim_User
Last spring I had to deal with the DRI3/SDL issue, when SDL based games slowed down at the X server switch to a DRI3 version. It was discussed both in these forums and the appropriate bug report originally filed against Xorg. A git bisect was beyond me as I am using PPA packages and not compiling X, so I rolled back packages, tested updates as they came out, and waited, still able to use the known good version. Performance of new versions improved little by little until a new version of libsdl appeared at the Trusty-Utopic transition time. A week later the slowdown returned, so I rolled back all packages and reinstalled one at a time until I found the offender, which was libsdl. I call that a "dpkg bisect" and it identified libsdl as the offender. The existing bug report was copied as a bug report against libsdl. It's closed now as the next update reverted an accidently left out change and fixed the bug for good.
If I had only discussed it here, libsdl's devs would not have heard about it unless they observed it themselves. If I had not discussed it here, I would not have found out about the hyper-Z issue also affecting performance and would still be down 20% or so. That was something I was easily able to fix myself, but would have had no idea was a problem without these forums.
If you want to be able to "bisect" a day's updates, keep all your old packages, and keep snapshots of your OS. When you get an unknown clunker, roll back, then reupdate one package at a time (or sets of packages known to be unrelated) until you find the problem. My experience is that this is essential to running a rolling release or Ubuntu alphas in working machines.
07-18-2014, 07:23 AM
My pedals are still dead since yesterday. Now I don't think that it's even possible to create such a bug. Probably the autor got money from M$, Apple or alike to
keep input devices buggy for the next 10 years.-
07-18-2014, 08:14 AM
Yeah, that's most likely what happened.
Originally Posted by mike4
I suggest you take your complaints to the appropriate venue though, as it is getting really annoying to read about your pedals all the time in tese unrelated threads.
That said, you would probably be better off moving back to windows, as these people are clearly in the pockets of M$ and Apple, so I'm sure they would not fix your bug.
09-09-2014, 01:42 PM
Originally Posted by tpruzina
The argument was that they share a lot of code which then had be be maintained in two different places.
09-09-2014, 05:54 PM
Don't get your hopes up as there's many ways this could turn out but:
Originally Posted by pgoetz
Read the whole thread for context, but, in brief, there's a problem with systemd requiring certain services to be run that aren't needed in a containerized runtime.