[02:35] is there anyway i can remove the mounted internal hdd icons from my desktop? [02:43] I'm sorry xubuntu09w but that is a question that needs to be taken to the support channel, #xubuntu, instead of here [02:46] * pleia2 pulling down dailys for an exciting saturday night [03:02] pleia2: Will said dailies leave a smoking mess, though? [03:02] skellat: I don't know, will they? :) [03:02] I'm running 14.10 on my laptop so, as it is, nothing has gone kaboom yet [03:02] that's good to hear [16:33] Oh joy, validation: https://paste.unit193.net/?63de00817bedcfa5#HlGwS2YBARmbCJWtOq3AdMvKUDMsiSS6anV74To2By4= [20:50] are indicators in trusty gtk3 or gtk2? [20:51] gtk3 [20:52] so http://developer.ubuntu.com/api/devel/ubuntu-13.10/python/AppIndicator3-0.1.html? [20:53] yes [20:53] okay [20:56] thanks! [21:01] ali1234: do you know where it searches for the icons? [21:01] no idea [21:16] Pwnna, icons are provided by the indicator via libindicator [21:17] andrzejr: if i have a script, for the indicator, can i put the icons in the same directory? [21:17] http://askubuntu.com/questions/289279/what-are-the-design-guidelines-for-appindicator-icons [21:17] i see that [21:18] sorry, I don't know much about it. [21:19] the plugin itself only displays what it gets from libindicator. [21:19] no idea what happens before [21:47] guys, any idea why xubuntu 14.04 is freezing for up to several seconds on "heavy" IO load? [21:49] ("heavy" as in decompressing a large tarball or running cron) [21:50] Fresh installation but for the first time the 64bit version and with full disk encryption enabled. [22:13] andrzejr: it's because ubuntu uses the deadline io scheduler, which is garbage for anything other than servers [22:15] see for example: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/770258 [22:15] Ubuntu bug 770258 in linux (Ubuntu) "Extremely high latency when writing to USB drives." [Undecided,Expired] [22:18] the problem with deadline is that once it misses the deadline it causes a failure cascade [22:32] ali1234, I didn't have this problem on earlier versions [22:32] well, i did :( [22:33] which were 32bit without encryption. Maybe that wasn't stressing the disk enough. [22:33] maybe. full disk encryption is going to mean extra stress [22:34] and i haven't used 32 bit since about 2008 [22:35] what scheduler do you recommend? cfq or noop? [22:35] depends on what type of disk you have [22:36] (afair I have seen this problem with cfq as well) [22:36] 5400rpm laptop hard drive [22:36] i've heard very good things about bfq [22:36] unfortunately it's not in kernel [22:37] I will try noop for a while [22:37] noop literally means "don't attempt to opimize at all" [22:38] I'm ok with not optimal performance, as long as it isn't disastrous [22:38] cfq might be better on a spinning disk [22:39] nobody has ever been able to adequately explain why the freezes even happen [22:39] but they have been a problem for a long time [22:40] another way it can happen is if you hit swap [22:41] eg if you accidentally open a huge image file and the viewer attempts to allocate several gigabytes of memory, this will freeze the entire system [22:41] assuming you have enough swap [22:42] if you press ctrl-alt-f1 it will take 15 minutes to switch to the console [22:43] I've been experimenting with that (lower swappiness etc). I don't think this is the issue. [22:43] swappiness doesn't make any difference [22:43] or rather it only affects what the system does when idle [22:45] the thing is, this problems occurs on fairly idle system. It's enough to run the upgrade tool and the disk activity brings the system to crawl [22:46] well, that is probably because of the encryption [22:47] noop scheduler does not make much difference [22:48] not sure if that's an unrelated problem but general system performance is pretty bad at times. Even playing some flash videos can result in 1fps slideshow. [22:50] it's i5 machine with 4GB and NVidia graphics card. Currently with nouveau drivers but I've seen same behavior with NVidia drivers [22:51] well, all i can say is you're lucky it runs at all with nouveau [23:03] ttyl, going to reboot the system with noautogroup to see if it makes a difference [23:16] ali1234: Have you been tracking the deadline/cfq debate on ubuntu-release@lists.ubuntu.com that K has stirred up? [23:16] no [23:17] hahaha i love how he accidentally wrote "deadlock" instead of "deadline" [23:18] The Kubuntu folks are trying to force a switch from deadline to cfq for K via a udev rule in a unique way that could screw with other flavors [23:18] we should support them, deadline is crap [23:18] Trying to pass it through via their flavor's -default-settings [23:19] cfq should be the default for all of ubuntu [23:19] ali1234: Well, that's a case the crack testing squad needs to make as slangasek and others have been dead set on keeping deadline [23:19] so far the system booted with noautogroup works much better but I will leave it running for several days. I have a feeling the performance issues may be related to suspend/resume cycles [23:19] well feel free to post my bug report from TWO YEARS AGO :) [23:20] popey also has experience of this issue [23:20] I'm mostly trying to stay out of the fireworks [23:20] Multiple lists are being copied and I'm only seeing what is on the -release list [23:21] i only subscribe to ubuntu-devel and xubuntu-devel [23:21] and the former seems to be "ubuntu-phone" now [23:21] so i don't actually bother to read it very often [23:23] I could have sworn the phone people had a separate list [23:23] they do [23:24] hmm #1219548 [23:25] bug 1219548 [23:25] bug 1219548 in linux (Ubuntu) "sched_autogroup_enabled breaks niceness" [Wishlist,Triaged] https://launchpad.net/bugs/1219548 [23:25] what is noautogroup? [23:27] kernel option that disables process grouping, which apparently breaks ionice [23:27] seems consistent with what I'm experiencing [23:27] Which was what the K folks were worried about [23:27] ionice & their package BALOO [23:27] If they can't ship baloo they're likely not shipping K 14.10 [23:27] Or so it has been threatened [23:28] my bug report is actually from three and a half years ago, when ubuntu was still using cfq [23:28] i still get the freezes though [23:28] so clearly switching to deadline didn't help [23:35] And I just heard a breaker trip on a UPS so I best investigate... [23:36] anyone got a link to the udev rule? i'd like to do some long-term testing, ie change scheduler permanently [23:37] arch wiki to the rescue https://wiki.archlinux.org/index.php/Solid_State_Drives#Using_udev_for_one_device_or_HDD.2FSSD_mixed_environment [23:45] * bluesabre is amused [23:45] "menulibre doesn't work... I've tried using sudo..." etc... [23:45] probably means their menus are owned by root now [23:45] yeah [23:46] And this doesn't bode well: http://www.iuculano.it/linux/apt-get-purge-chromium/ [23:47] why? [23:49] The DD referenced as not playing nice is one of my predecessors as Ohio Leader [23:49] i don't understand the significance of that [23:51] I break the line for Ohio Leader [23:51] One of my predecessors, Paul Tagliamonte, is an FTP Master Debian-side and doesn't do much Ubuntu-side [23:52] what is Ohio Leader? [23:52] Ubuntu Ohio LoCo [23:52] and what do you mean by "break the line"? [23:53] Meaning that I'm the first one who isn't a Debian Developer in years who has stuck with *buntu despite all the changes that have been going on [23:53] Though if Michael isn't playing nice upstream packaging chromium we really shouldn't sync it from Debian [23:54] And take however large of an Ubuntu delta we have to [23:56] i don't get it... he downgraded chromium in experimental, which causes it to crash? [23:57] And apparently big, unexplained commits [23:57] i don't like it when people write "change some stuff" as the commit message, but it's not exactly uncommon [23:58] so much random talk here [23:59] I've just installed the numix theme.. and windows 10 looks now fabulous