**** BEGIN LOGGING AT Sun Mar 24 02:59:58 2013
Mar 24 12:10:11 <s6581> hi!
Mar 24 12:16:23 <s6581> anyone here not idling?
Mar 24 15:04:02 <Tofe> Morning !
Mar 24 15:16:40 <s6581> hi
Mar 24 15:35:18 <ka6sox> morning Tofe
Mar 24 22:31:56 <destinal> hmm.. our channels are too quiet
Mar 24 22:32:41 <PatrickC|Away> well, that can be a good thing
Mar 24 22:32:47 <PatrickC|Away> it could mean that everyone is busy coding :)
Mar 24 22:33:11 <ka6sox> if only...
Mar 25 00:02:31 <lucaat> Hi
Mar 25 00:02:43 <lucaat> is the gnex port still in the works?
Mar 25 00:03:21 <ka6sox> yes
Mar 25 00:08:55 <lucaat> looks good already from the videos, but the fonts seem way too small imo
Mar 25 00:09:09 <lucaat> as if it were for a tablet 7"+ display
Mar 25 00:09:39 <lucaat> very much enjoyed the whole webos deal on the touchpad
Mar 25 00:11:08 <ka6sox> you mean LunaCE?
Mar 25 00:11:13 <ka6sox> or something else?
Mar 25 00:13:33 <lucaat> I don't know what that is, enduser here sorry
Mar 25 00:13:42 <lucaat> I mean from the videos on the twitter account
Mar 25 00:13:44 <lucaat> on the gnex
Mar 25 00:16:07 <EricBlade> hey guys. i'm poking at the code that sets up the oom killer parameters for luna processes, and i'm noticing that while it appears to be working (when the kernel OOM kills something, it reports the correct values), i can't seem to tell that from the outside world (catting the oom_score_adj file from a terminal reports "0" instead of the "-1000" we wrote to it) .. which makes it really difficult to convince someone wh
Mar 25 00:16:42 <EricBlade> anyone have any idea why if we're writing -1000 to /proc/x/oom_score_adj, cating it shows 0?
Mar 25 00:17:36 <ka6sox> what kernel are you using
Mar 25 00:18:00 <ka6sox> we mucked with the settings for that to make it less suceptable to OOM errors.
Mar 25 00:18:09 <ka6sox> (and too many cards errors)
Mar 25 00:19:10 <ka6sox> EricBlade, are you writing that as a priv user or a normal user?
Mar 25 00:20:56 <EricBlade> ka6sox: hmm. i guess in desktop, i'm a normal user. which might explain why it's not working. i thought that a process could write successfully to it's own /proc/self though
Mar 25 00:21:33 <ka6sox> sysfs entries also have priv settings like normal files :)
Mar 25 00:22:17 <EricBlade> ofcourse, i'm not getting any errors trying to write to it
Mar 25 00:22:35 <ka6sox> blissfully ignoring you.
Mar 25 00:22:44 <EricBlade> so i'm not even sure if it is working on desktop, but desktop is where i need to demonstrate that it does. which might? not be possible?
Mar 25 00:23:41 <EricBlade> i thought for any given process, it would have write to /proc/self though
Mar 25 00:23:43 <ka6sox> ls -l /proc/x/ and let me know the results?
Mar 25 00:24:21 <ka6sox> (not running the desktop version)
Mar 25 00:24:26 <EricBlade> the files are there, oom_adj and oom_score_adj (desktop on 11.04 ubuntu so kernel .. uh.. 2.6.38)
Mar 25 00:25:15 <EricBlade> and i don't have any kind of emulator built at the moment, nor a Open device
Mar 25 00:26:14 <ka6sox> not on my version kemosabe...
Mar 25 00:26:57 <EricBlade> ... and they are both writeable only by root .. still i swear that at some point in history, anything could write to it's own /proc/self .. grr.
Mar 25 00:27:19 <ka6sox> of course I'm on 3.2.0 with 12.04
Mar 25 00:28:07 <EricBlade> i was a little surprised that the removal of oom_adj didn't happen in 3.4
Mar 25 00:28:49 <EricBlade> in any case, i'll just assume that that is the problem then, and explain to the person who doesn't believe me that it's already being done correctly, that you can't tell on desktop because sysmgr and friends aren't root on desktop
Mar 25 00:29:18 <EricBlade> :D
**** ENDING LOGGING AT Mon Mar 25 02:59:59 2013