[01:53] chronic problem: Caja is not adopting the system theme. Anyone know why this is? [01:53] sometimes it starts from boot, and sometimes it happens later. no idea how to fix [01:53] sometimes a relog can fix, sometimes a reboot. sometimes not! [02:28] Hi, can someone suggest me what should i use to open .rar files.. [02:36] rar? [02:37] "unrar" [02:37] is the command [02:37] To extract file.rar file into the current directory, enter: [02:37] $ unrar e file.rar [02:37] unrar e filename.rar [02:54] nomic: sorry didnt sawa before your chat, ok i will try unrar now [02:55] seems that i have to install it [02:57] nomic thx works like a charm [03:03] being "how to extract rar files" [03:54] no one can help me with broken themes on Caja? [05:43] helo experts.. [05:43] need one help/// [05:43] while running you tube or vlc ...screen flickers.. [05:44] does any one have solution for this.. [05:44] Try this command: marco --no-composite --replace [05:44] Does this fix the issue? [05:44] nope.. [05:45] You already tried it, in 7 seconds? [05:46] was running this fro very first day.. [05:47] use omxplayer [05:47] on the raspberry pi 3 [05:47] This is on an rpi3? [05:48] well, if. [05:50] mrbhardw: is this on an rpi3? [05:51] He really doesn't want to say :) [05:52] ? [05:52] . [05:52] rpi3 ? [05:53] nomic suggested that you use omxplayer if you're using a raspberry pi 3 [05:53] And I asked if you are indeed using one [05:53] issue is with youtube as well.. [05:54] What is the output of this command? lspci -nn -k | grep -A 2 VGA [05:55] lspci -nn -k | grep -A 2 VGA [05:55] 00:02.0 VGA compatible controller [0300]: Intel Corporation Broadwell-U Integrated Graphics [8086:1616] (rev 09) [05:55] Subsystem: Toshiba America Info Systems Broadwell-U Integrated Graphics [1179:0004] [05:55] Kernel driver in use: i915 [05:55] OK so you're using intel graphics, no raspberry pi [05:56] And the output of this command? gsettings get org.mate.session.required-components windowmanager [05:56] mrbhardw@MRBHARDW-IN:~$ gsettings get org.mate.session.required-components windowmanager [05:56] 'marco-no-composite' [05:56] mrbhardw@MRBHARDW-IN:~$ [05:57] Are you using a custom xorg.conf? [05:57] nope [05:57] So you didn't set "tearfree", i assume [05:57] yes [05:58] i did'nt [05:59] I would try to switch back to uxa, rather than using sna. The default got to be sna for a couple of years, but it wasn't very maintained, and the ubuntu xorg maintainers in 16.10 are going to default to uxa again. [05:59] Ask in #ubuntu-x for more details [06:00] ok [06:24] woa - my pine64 actually working [06:24] what sort of a miracle is that [06:30] Any advice on Flash for pine-64 on ubuntu-mate [06:30] don't use flash :) [06:31] so is youtube out of range then [06:31] its not [06:31] nowadays youtube uses html5 [06:31] good to hear - more? [06:31] flash is legacy [06:31] excellent [06:31] how about nz Lightbox [06:31] whats that [06:31] i haven't ran flash in at least 6 months now [06:31] its an online TV channel in NZ [06:32] Maybe kodi can help there [06:32] It has a lot of plugins that don't need flash [06:32] if they are using old technology you need to run flash in firefox in wine, if not you don't [06:32] I don't think he can run wine in an arm platform [06:32] hmm [06:33] ...virtualbox? [06:33] not sure I want to retool to kodi [06:33] It's a different cpu, it would need an emulator [06:33] arm is already 10 times slower, adding an emulator would make it 100 times slower, like 1 frame per hour [06:34] ARM is fine for android, but it's not very well supported on gnu linux yet [06:34] (meaning all the gpu acceleration etc, not just the cpu) [06:35] rat - then I might need to retool to android [06:35] but I am trying to move entirely to linux [06:35] i mean [06:35] i can run linux on a 10 year old powerpc g3 [06:35] nigel: consider retooling to this: http://www.gearbest.com/tv-box-mini-pc/pp_343636.html [06:35] :) [06:35] so you must be able to use it on your arm device [06:36] Normal x86 64bit processor :) [06:36] Akuli: does your powerpc use the gpu to boot? [06:36] i dont know [06:36] Does it rely on the gpu for all the intensive operations, like video decoding? [06:36] That's where linux support has issues, and omxplayer etc are needed [06:37] i had some trouble with getting x to work and it was really slow [06:37] so i decided to just get rid of x and use it the real way [06:37] It's about the same for arm platforms. They're best suited for mobile phones, not for desktops. [06:37] i mean... it worked [06:37] Sure, rpi2 here works too, but it's not really usable as a desktop pc [06:37] if that was my only computer i would get my stuff done with it [06:38] I'll check it out [06:38] 256MB of ram was the only real limit, but other than that it was usable [06:38] looks OK - but my Pine64 was $15 [06:39] nigel: add the ram, sdcard, hmdi cables , case etc [06:39] There are cheaper ones as well, from 20$ [06:39] (including all those that I mentioned) [06:39] But if you want to do normal desktop work, better get a x86 machine [06:41] sd card was lying around, RAM included, hdmi cable also lyting around, case is made of cardboard [06:41] keyboard - $1 from charity shop [06:41] And 200 hours to get it to work in 1/10 of the real pc capabilities ==> priceless [06:42] If you want a print server, sure, use it [06:42] If you want a desktop pc, save yourself the time [06:42] Actually I had issues making rpi2 a print server, some lexmark printer only had x86 binaries [06:42] haha - its a photo server and a TV server [06:43] desktop is my $25 repurposed Dell Latitude [06:43] 21 euros: http://www.gearbest.com/tv-box-mini-pc/pp_401687.html [06:43] way faster than the RPi [06:44] quad core, designed as a tv box, using android, has binary gpu drivers for acceleration [06:44] Me, I do want to be able to use flash, so I don't buy arm [06:45] i don't, but i don't have any arm devices anyway [06:45] just powerpc, i386 and amd64. [06:54] ? === donny is now known as nisco [08:32] hello [08:32] How can I check all the peripherals before installing? [08:33] Try it with a live cd [08:33] I'm on the live session [08:33] Plug in the peripherals that you want to test, and see if they're working [08:33] I just like to see that everything works [08:34] the usb is working [08:34] internet too [08:34] is there a special way to check all? [08:35] Not really, e.g. you can't have a reliable test for devices that don't really work. Do it manually. [08:36] At some point there was a "checkbox" effort, but it didn't get far... [08:36] what would be the most important things to check for? [08:36] I'm new to linux [08:36] I have win 10 [08:37] I'd like to do a dual boot [08:37] The devices that are outside of your pc, like printer, 3g sticks... [08:37] I read it windows can cause problems [08:37] !dual-boot [08:37] Dual boot instructions: x86/AMD64: https://help.ubuntu.com/community/DualBoot/Windows - Macs: https://help.ubuntu.com/community/MacBookPro https://help.ubuntu.com/community/YabootConfigurationForMacintoshPowerPCsDualBoot [08:38] i should disable secure boot [08:38] yes [08:38] windows 8 and 10 do cause a lot of problems when dualbooting [08:39] I would say "may cause if they're not handled properly" [08:39] Normally they don't cause any issues... [08:39] what would be your advice for me? [08:39] Follow the instructions... [08:39] :) [08:40] Yes they are clear, I just wanted to get more info for those that have done dual booting with win 10 [08:40] most importantly: if ubuntu doesn't detect your windows, cancel the installation and check whats wrong [08:41] you need to disable "fast boot" in windows power settings, which basically means it never shuts down fully [08:41] so you want it to do that [08:41] ok will do that [08:41] You can also reboot from windows, in order to boot from the live cd, instead of shutting down [08:41] reboot is always clean... [08:41] its behind something like "settings that are currently unavailable", you can probably find a nice tutorial from somewhere [08:41] alkisg, wow i didn't know that [08:42] It's because they hang a lot, and provide a way to clean up things :) [08:42] i guess i've just gotten used to operating systems that do what i want them to do :) [08:42] gparted will refuse to resize the windows ntfs volume if it's in use (hibernated etc) [08:42] because i've never actually had windows 8 or 10 on one of my computers [08:42] * alkisg had to install a lot of them recently, mainly for games for his kids... [08:43] I'll look for reboot from windows, that's new to me [08:43] I really like the aesthetics of ubuntu [08:45] I've searched on google and it only tells of rebooting windows itself === andreas_ is now known as Guest51241 [09:31] alkisg and Akuli, also others what do you think about the post i made on UM forum [09:31] https://ubuntu-mate.community/t/a-suggestion-how-to-powerup-ubuntu-mate-irc-support/8715/1 [09:33] nice idea, but then someone could just tell others to say thanks [09:34] and get a bunch of cookies without helping at all [09:34] logical: I didn't get the idea. For me, this channel should be specific to issues that belong to the ubuntu-mate code (desktop environment issues, seed issues etc), while the rest of the issues should be redirected to #ubuntu [09:34] i don't think that dividing these channels in half like that is good idea [09:34] "this does this, that does that" [09:35] E.g. suppose someone comes here and wants help installing phpmyadmin [09:35] different people need different kind of helping, some prefer a small channel and some need a big channel [09:35] Why would that be a topic of #ubuntu-mate instead of #ubuntu? [09:35] if that someone doesn't like big channels [09:35] emm alkisg you didnt get the idea xD [09:35] also why couldn't we help here with something not specific to mate? [09:35] logical: it's a "thank you" idea, while I think the main issue is the topics that we support... [09:35] i don't see any problems in that [09:36] akuli you just set the bot that the same person cant give cookies to the same guy for next 10 min [09:36] what if i just say "thank you"? [09:36] I believe that the person asking for phpmyadmin, would get more help from the 1000 persons in #ubuntu rather than the 100 persons here, because there would be less persons here actually using phpmyadmin [09:36] who's going to get the cookie? [09:37] alkisg, then that person can go to #ubuntu, which is currently how it goes [09:37] but i don't want to see people being forced to #ubuntu from here just because their problem is not mate-specific [09:37] Akuli: when people in #ubuntu ask offtopic questions, they redirect them to #ubuntu-offtopic [09:37] We don't have to be very strict [09:38] and we are not very strict right now [09:38] and i think thats great, lets keep it that way [09:38] Ofc the bot needs to be refined, i just put the idea, maybe if i say just thank you it can go to the last person that mentioned my name etc [09:38] But it's better if we inform the users where they'll get the best support [09:38] the reason for redirecting people away is that the channel is full all the time and there's not time for chitchat [09:38] which is obviously not a problem here [09:39] logical: about your idea, google for "irc bot karma plugin" [09:39] There are many karma plugins implemented [09:39] But I don't think I would feel better if I had e.g. 1000 thank you [09:40] It might help the new users a bit, so that they would trust someone more, if he had a lot of thank yous [09:40] Well the main point of my question is what do you think about the idea xD, it needs to be refined and prepared [09:41] I got the idea from the freecodecamp chat, since their chat have that ability it rly keeps the people chatting [09:41] I was thinking, what could be helpful is to have "shifts" for helpers. I'm not sure cookies are that good a motivation but what would be nice is if we find a way to maintain round-the-clock coverage of this place [09:43] but people who come here are volontering (i hope i wrote it nice xD) they are not paid members. I come at chat when i log in [09:43] i'm not able to come here regularly, and most other helpers aren't either [09:43] also i don't want a strict helpers-helpeds division [09:47] !karma [09:47] !thanks [09:47] You're welcome! But keep in mind I'm just a bot ;-) [09:47] ...we had a karma plugin activated in #ltsp... [09:49] We just needed to say !person+1 or something [09:51] lol [09:51] (Y) [09:52] karma seems to be highly motivating on stackexchange [09:52] compare the thorough helpfulness of replies on there, to traditional forums [09:53] (1) stackexchange has a "view profile" option, while irc doesn't, and (2) stackexchange is very focused on questions/answers, while forums are not that strict, and that actually motivates helpers looking to help rather than chat [09:55] their karma system isn't exactly motivating me though [09:55] true. and quantifiable rewards can demotivate participating "for the fun of it" [09:55] https://www.youtube.com/watch?v=u6XAPnuFjJc [09:55] im just looking at other peoples karmas like 10000something, mine is 11 [09:57] that too. different folks, and all. on any site I use that has a karma-like system, I notice every slight fluctuation, like a facebook friendlist junkie watches their friend count [10:15] hello [10:16] hello and bye [10:17] that happens in other channels, too. people leaving if they don't get a reply in one minute. what if the bot greeted people, could that help? [10:42] not bad idea olscumpy [10:55] can ubuntu mate do okay on a crappy old netbook that originally came with windows xp on it? or should I stick with the typical recommendation of lubuntu, xmonad, etc? I'm a little smitten with mate so far and the netbook is already running ubuntu 14 so it probably upgrades easy... [10:56] yes [10:56] i have ubuntu mate on my old laptop at the moment [10:57] and it works so nice [10:57] 2.0 dual core, came with 1 gb ram, some integrated graphic card 300+mb , i put 2x 2gb ram so i have 4, it works perfectly. I dont need anything better at the moment for my studies [10:57] 8y old laptop [10:58] cool. I only have mate on my good laptop currently [10:59] xfce is more resource heavy than mate (at least in my experience on this laptop) so i would suggest you to stick with mate, it is also very stable [10:59] only if your laptop is ultra old then.. :P [11:01] I would not suggest to go with Ubuntu for old pcs === pavlushka_ is now known as pavlushka [12:05] I like to workout! #chatisdead [12:12] busy :P [12:54] Hey I have on this pc win 7 and UM as dual boot. Now I want just to install UM, so I separated my disk like this 1 part is Win7 2nd Files (whenever i reinstall i keep important files in that part of hdd) and 3rd linux. Can i reinstall and still keep the "files part" as a ntfs. Do I have to manually delete windows and linux and create as one ext4 and swap partition? [12:55] doesn't really matter how you do it as long as you have an ext4 mounted at / and a swap [12:55] the swap should be about the same size as your ram, but it doesn't need to be more than 3GB [13:03] oops. my swap is 20 gb or so haha [13:06] right doesn't need to be that big :) [13:07] if you look at the system monitor you'll notice that its not being used much [13:08] wow. 20.8 MiB of 22.4 GiB [13:08] why does linux write it as mib? there's no i in that word [13:08] ok thx [13:09] olscumpy, its mibibytes [13:09] i mean [13:09] mebibytes [13:09] not megabytes [13:09] wild [13:09] 1000 of them instead of 1024 ? [13:09] something like that yes [13:10] i forgot which is which [13:10] it was always 1024 and "mega" when I used dos / windows [13:11] windows might be lying also [13:11] there's videos about this on youtube [13:11] there was an option during mate setup to make things 1000 or 1024 [13:11] i've never heard of that [13:11] maybe it was when I was setting up partitions [13:12] and was just for the one new partition I created [13:33] hello guys [13:33] is there a way i can install 4.8 kernel on ubuntu mate ? [13:34] are there any ppa's for that ? [13:34] why do you need it? [13:34] i run 16.04.1 [13:34] yes, but why do you need linux 4.2? [13:34] i mean 4.8 [13:34] problem is, i use docking station and intel pushed few bugfixes that i'm affected into their nightly build which will come mainline in 4.8 kernel... so far ubuntu mate is on 4.4 ... [13:35] https://bugs.freedesktop.org/show_bug.cgi?id=89055 | https://bugs.freedesktop.org/show_bug.cgi?id=92878 and another that i forgot link [13:35] Freedesktop bug 89055 in DRM/Intel "[SKL APL] WARN_ON(!wm_changed) hit in the WM code with multiple pipes enabled" [Major,Closed: fixed] [13:35] Freedesktop bug 92878 in DRM/Intel "eDP link clock recovery fails with *ERROR* too many full retries, give up" [Critical,Resolved: duplicate] [13:35] i guess you can build it from source if you really need to [13:36] that would be worst possibility :) [13:36] BlackPanx: this isn't related to mate specifically, it affects all ubuntu versions [13:36] i really would rather take ppa for it, so i can simply remove it if i hit some other problems [13:36] Normally newer kernels are available for lts releases, after each non-lts release is out [13:36] This is called "hardware enablement stack" [13:36] what does 16.10 come with? [13:37] So, when 16.10 is out, you'll be able to install the 16.10 kernel to 16.04 [13:37] I don't know [13:37] i thought it is already [13:37] But there's also the "mainline ppa" which has newer kernels that do not have the ubuntu patches [13:37] https://wiki.ubuntu.com/Kernel/MainlineBuilds [13:37] You can get 4.8 rc4 from there [13:38] But when the 16.10 kernel becomes available in 16.04, you should switch to that one instead [13:38] https://wiki.ubuntu.com/Kernel/LTSEnablementStack [13:39] So, normally, the 16.10 kernel will be available for 16.04 at the time when 16.04.2 is out [13:39] https://wiki.ubuntu.com/XenialXerus/ReleaseSchedule [13:39] alksig that is awesome [13:39] thank you so much for detailed information [13:39] use "tab" to autocomplete usernames in irc, to get them right :) [13:39] You're welcome! [13:41] Btw, google says 16.10 comes with 4.8 [13:45] awesome [14:06] alkisg, would you suggest me normal mainline kernel from ubuntu or drm-intel-nightly kernel ? [14:07] i'll just try both and see how it goes [14:10] important note: if you want to choose your old kernel you can do that on startup, hold down left shift, go to advanced options and there it is [14:11] thanks [14:11] was it always left shift ? [14:13] i think so [14:13] most of the time i don't use the right shift anyway so i'm not sure [14:13] or if you dualboot to windows you get to that grub menu without holding down anything :) [14:29] hello [14:29] ? [14:30] hello :) [14:31] Hi, I was wondering if I could get help in ubuntu-mate 16. When I go to the software center on the "welcome" page all I get is package can't be found [14:45] 大家好 === pavlushka is now known as Guest64188 === pavlushka_ is now known as Guest93617 === Guest93617 is now known as pavlushka [18:59] Hi folks, I'm hoping you'll permit me to be teeeeetering on the edge of INSANITY [19:00] There is a bug... a really annoying one - one that is so pervasive it existed on 14.04 and now after installing 16.04 (fresh) i see it still exists [19:00] every time I've ever talked about it - no one has any ideas [19:00] but what it does demonstrate is this... ONLY AMERICANS USE UBUNTU MATE [19:01] they must! because everyone else must get really REALLY annoyed at the fact their keyboard layout *ALWAYS* resets back to en_US upon reboot [19:01] despite every config file, every dialog box, every setting saying en_GB (or whatever other language exists) [19:02] now I know americans are a deeply conceited folk and mustn't think all us other folks exist - but we do!! and some testing of other keyboard layouts would be handy! [19:02] I don't have that issue [19:02] Well I have it on like 5 different machines [19:03] p.s. I'm normally a very calm little IRC talker - but I've so GOT to fix this and I'm completely at a loss what it might be - I need IDEAS people! [19:04] errmm... Under keyboard preferences what do you see in the layouts pane? [19:04] rahtgaz, british [19:04] well English (UK) [19:04] No English (US) entry? [19:04] no just that [19:04] you have to add a 2nd English (UK) option, then remove the first one - and you'll magically find it works [19:04] until you reboot [19:04] where it'll be en_US again [19:04] What's the keyboard model? generic 105-key Intl? [19:05] yeah [19:05] this is a fresh 16.04 install ... no ibus or anything like that installed [19:05] actually I've just booted up my dell lappy (also has 16.04 on it) and that doesn't have the issue [19:05] I have no issue here English UK so I imagine there is something wrong rather than a bug [19:07] Go to language support under system->Preferences->Personal [19:07] Do you have anything there that is English (United STates)? [19:08] rahtgaz, the only languages which are not in grey (i.e. they're in black) on the Language tab are 'English (United Kingdom)' and 'English' (in that order) [19:08] hmm.. [19:08] i'm asking my partner to get his acer laptop - that has 14.04 on it I think - if that doesn't have the bug... then it must be something to do with the Logitech K360 keyboards I use both here at home and at work [19:10] right ok - his lappy is fine also [19:10] lemme reboot with a normal ps2 keyboard - brb [19:13] ok so at least things have conspired to help narrow this down a little... as my old ubuntu 14.04 hdd went from one pc to another recently - and the other it went to *doesn't* have a logitech k360 keyboard - it seems that it (I never noticed) no longer has the issue [19:13] so the OS hasn't changed - just the hardware it booted up on did [19:13] plus I just tried a ps2 keyb on the machine which isn't working now - works fine [19:13] and both laptops work [19:14] so all this time I've been thinking its multiple machines (at least 3 I use have a logitech k360 keyboard) it must just be some kind of odd reaction linux has to a k360 keyboard [19:14] in that case try to change the keyboard model to one of the Logitech ones. Probably Logitech Generic Keyboard [19:15] rahtgaz, ok - I can... sure.. but why? why when my own bios doesn't have an issue, and windows doesn't, heck even dos booted up from a usb pen doesn't (for firmware updates) would linux need to be difficult? [19:15] there is something more going on here [19:16] The 'why' I can't answer. We are still rying to find a solution. Sometimes the oddest problems only find an explanation after they stop happening [19:17] ok well i picked logitech/generic which instantly fixed it - but that doesn't mean anything - so does re-adding english (uk)... just rebooting now... [19:17] no still broke [19:17] that was fast [19:17] but that at least proves changing any setting in 'Keyboard' reminds it that it is doing something wrong [19:17] ssd - and it's a fresh install with nothing on it [19:18] this is 14.04, correct? [19:18] no it is 16.04 [19:18] but the issue existed in 14.04 as well - presumably releases in between too - but i don't use them [19:18] yeah, sorry. I emant 16.04 [19:19] i've got 4 machines running 14.04 and 2 running 16.04 in total [19:19] It's just moving to a PS2 isn't helpgul. I wished you had another USB keyboard around [19:19] i do - want me to try? [19:19] that would be better. if it happens also on that keyboard we know it's ubuntu-mATE fault [19:19] ah you think it might be usb keybs in general? good idea [19:20] i'll connect - then reboot it [19:20] btw i'm on another pc - which is why i don't go offline [19:20] * rahtgaz nods [19:20] lol - love it [19:20] connecting the usb keyboard instantly fixed both of them [19:21] lemme just reboot and reproduce that to be sure [19:24] yeah [19:25] same? [19:25] so i boot up with just the k360 connected (the little logitech unifying receiver) and open pluma... pressing shift+2 gives me @ not " [19:25] if I connect the little keysonic usb keyb i have - the k360 still does the same [19:25] but if I type on the keysonic [19:25] then both keyboards act as they should [19:26] it's like something is triggered in the background [19:26] this is on generic 105 int keyb btw [19:26] Right. I'd fill a bug on Ubuntu-MATE tracker for the logitech k360. [19:27] it'd be very unlikely it'd just be that one keyb though right? [19:28] I'm not sure. I find this problem one of the oddest things I've come across. I just can't see What could possibly cause this. [19:28] i can film this in action if you'd like lol [19:28] Not the keyboard for sure. It can't advertize its language (to my knowledge). The system doesn't care and actually usually asks us during intals... [19:29] wait... [19:29] all i know is the moment you type even a single character on the keyb that you connect 2nd (after it's all booted up) *then* the system realises it has *ucked up and the layout is fine on both keybs [19:29] Actually there might be a bug here... do you remember if during install you may have announced it as a US keyboard? Or no chance it happened? [19:29] i picked UK [19:30] alright. File the bug. They will know better what's going on [19:30] see i always found it bizare i never got this in ubuntu 16.04 or 14.04 mate vm's [19:30] i.e. it's always been a physical keyb thing - but wasn't sure until now it was just this one keyb [19:31] i'm wonder if i get this if I just leave the non-wireless usb keyb plugged in on boot [19:31] lemme try that [19:31] (i.e swap them over) [19:33] ok no - lol it's k360 only it seems [19:33] which would explain why so few (i.e. none) other people have been so irritated by this [19:33] you notice it more if you dual boot lol [19:34] still file on ubuntu-MATE. I just can't see how a keyboard can instruct the system to change its layout [19:35] the system doesn't think its layout has changed though [19:35] the system (using all known tools/options/dialogs - i've been through them all before) believes it is already on en_GB [19:36] i'm gonna do some more testing at work tomorrow - they have other logitech wireless keybs that use the Unififynig Receiver usb dong;e [19:36] *dongle [19:36] perhaps it is something to do with that [19:38] rahtgaz, could it be this? https://wiki.archlinux.org/index.php/Logitech_Unifying_Receiver#Keyboard_layout_via_xorg.conf [19:38] which may explain why i never saw the bug in 12.04 [19:38] as that didn't have linux 3.2+ [19:39] or perhaps it did - hmm [19:40] Indeed. That explains it a lot better. A kernel module being behind the cause. I couldn't think on how hell would otherwise a keyboard layout change without you saying it so [19:40] rahtgaz, think I thought blacklist hid_logitech_dj and reboot - see if it works first time around [19:41] if I blacklist it - surely it'll just treat it as a normal USB hid [19:42] sorry. have to afk [19:42] ok - thanks for your help anyway [19:42] always good to have another ear [19:43] 12.04 had kernel 3.2 already [19:43] Check the output of `setxkbmap -query` before and after the issue [19:43] alkisg, layout is gb - it always is [19:44] well when you ask anything (like that command... yes I just re-ran it) [19:44] Have you verified this with setxkbmap -query? because there are many places where the layout is changed [19:44] yes! [19:44] i know of these many places - this is not the first time I've been in here chatting about this [19:45] the system is truely in every area/respect thinks it is programmed for GB and not for US [19:47] If you press alt+ctrl+f1, do you still get the issue outside of xorg? [19:47] I.e. in vt1? [19:48] alkisg, no i only get it in xorg [19:49] I think blacklisting the module won't do. Those nonsense receivers have their own protocol, and without a kernel module the keyboard will be missing or even messier [19:49] If the keyboard works fine in vt1, then it's a xorg configuration issue [19:49] Can you try this? sudo service lightdm stop; sudo xinit [19:49] This will show up an xterm; try it there, then type exit to exit. [19:49] rahtgaz, yeah blacklisting it just made it stop working entirely [19:50] This will allow you to see if it's an issue with mate/session/ibus/fcitx/whatever, or plain xorg config [19:50] ad if its xorg, you can try the Arch recipe [19:50] s/ad/and [19:52] rahtgaz, i'm not using that workaround - i already know about several workarond - i want to kill it... KILL IT (the bug I mean - actually find the cause) [19:53] well, they are telling you the cause is the module. And that makes sense in fact. But being that the case, there's little you can do, other then sifting through the kernel bug tracker [19:53] alkisg, well your command (type whole yes? i.e. 'sudo service lightdm stop; sudo xinit' ) just results in a black screen [19:53] Lantizia: lightdm stop, stops xorg [19:53] Then you login to vt1 [19:53] Then you type sudo xinit [19:53] ah ok that makes more sense [19:53] So you need an alt+ctrl+f1 between [19:54] yeah i get it [19:55] alkisg, yeah it's broken in the terminal you get too [19:55] i.e. the windowless/desktop-environment-less one [19:55] Lantizia: in that terminal, [19:55] yeah [19:55] first see `setxkbmap -query`, [19:56] then try `setxkbmap gb` or whatever your layout is, [19:56] and see if it's fixed with a plain setxkbmap [19:56] still claims gb (in that terminal) [19:56] And re-setting it, fixes it? [19:56] oh setting it works sure - but I knew that [19:56] So it's a plain xorg issue, because nothing else is loaded there [19:56] Or xinput etc [19:57] I.e. complately unrelated to mate or other desktop environments [19:57] yeah i never thought it was exclusively mate [19:57] it's just I don't use anything other than that [19:57] Just focusing on where to file the bug [19:58] What is the issue exactly? I haven't read all the chat. The internal keyboard has issues when you also have an external one connected? [19:58] lol [19:59] even if your system has been installed picking something other than us/english(us)/america/etc... sets say british/uk/english(int) or some other language [19:59] the keyb (in my case a k360) will always on boot up be in en_US [19:59] even though all utilities (like your -query command) and preference dialog boxes ... all SWEAR they are the language it is meant to be [20:00] So the issue happens with a single keyboard? [20:00] if you change any keyboard settings (e.g. what keyboard you use from generic - or re-add your language, anything) then it magically works [20:00] Or do 2 keyboards need to be connected for the issue to appear? [20:00] if you add a 2nd usb keyboard for example - the moment you start typing on that 2nd keyboard - BOTH keyboard have correct layout [20:00] the 2nd keyboard was just for testing [20:01] the issue is about have 1 normal logitech keyboard [20:01] but something about connecting a 2nd one and beginning to type - reminds the system that is using the wrong locale and resets it [20:01] only when you begin to type on that 2nd keyb though [20:02] So probably this bug? https://bugs.freedesktop.org/show_bug.cgi?id=49950 [20:02] Freedesktop bug 49950 in Server/Input/XKB "Logitech Unifying Receiver and wrong keyboard layout" [Normal,Reopened] [20:02] basically there are mutliple was of triggering the system to realise it isn't doing it right - the 2nd keyb option is just a more elaborate one [20:02] wow ok that Mathieu Bérard has it exactly! [20:02] especially the 2nd keyboard idea [20:03] Read comment 34 for a good explanation [20:07] alkisg, so in other words - they won't fix it (i understand the reasoning though - makes sense) [20:07] better to recode it so it's less crappy [20:07] Lantizia: they fixed it in libinput [20:08] They wouldn't fix it in xorg, they delegated it in evdev (won'tfix) and in libinput (fixed in late 2015) [20:08] alkisg, that bug report just says they 'want' to fix it in libinput - not that they have [20:08] See the second link that they give for libinput [20:08] https://bugs.freedesktop.org/show_bug.cgi?id=92896 [20:08] Freedesktop bug 92896 in Input/libinput "Split mixed devices into distinct pointer/keyboard devices" [Normal,Resolved: fixed] [20:08] 2nd link? [20:09] They give the links in comment 35 [20:09] "For evdev it's Bug 92897, for libinput it's Bug 92896. " [20:09] bug 47312 in kdebase (Ubuntu) "duplicate for #92897 Message partially hidden in "File Sharing"" [Medium,Fix released] https://launchpad.net/bugs/47312 [20:09] bug 85159 in GtkHTML "duplicate for #92896 evolution crashed with SIGSEGV in html_object_get_left_margin()" [Critical,Fix released] https://launchpad.net/bugs/85159 [20:09] oh i see sorry [20:09] so if its fixed - does that mean my packages are just old? [20:09] i mean if I backported some from 16.10 would that work? [20:09] Could be. I don't think 16.04 would have commits from late 2015. [20:10] You can try with a 16.10 live cd, sure [20:10] And when it's time, just upgrade to the 16.10 xorg stack === philroche_ is now known as philroche [20:10] nah I'd rather have the workaround than something that isn't LTS [20:10] i.e. apt install xserver-xorg-lts-yakkety [20:10] but it'd be nice to know that it is fixed - for 18.04 lol [20:11] 16.04 will officially get the 16.10 xorg stack [20:11] It's called hardware enablement stack [20:11] That will happen when 16.04.2 is out [20:11] officially as in I don't need to run antything to get it - it'll just be a normal system update? [20:11] https://wiki.ubuntu.com/Kernel/LTSEnablementStack [20:12] If you install 16.04.2, you'll get it [20:12] It doesn't get in 16.04.1 updates without a manual apt-get [20:12] Read the wiki page for details [20:15] alkisg, so i'm on 16.04.1 are you saying just a normal 'upgrade' will work or must it be a 'dist-upgrade' ? [20:15] Lantizia: do you really want me to type the wiki page here? :) [20:15] You'll need this: (11:10:46 μμ) alkisg: i.e. apt install xserver-xorg-lts-yakkety [20:15] It won't come to 16.04.1 with either "upgrade" nor with "dist-upgrade" [20:16] It will be there in the 16.04.2 live CDs and installations though [20:17] Btw, don't do normal "upgrades", only do "dist-upgrades" which properly upgrade your system [20:18] how poo [20:19] alkisg, is xorg the only thing that is held back between point releases? [20:20] because if there is more - that is extra poo [20:20] Err I think you misunderstood something but I'm not sure what [20:20] "dist-upgrade" doesn't mean "upgrade to the next release" [20:20] no I never said it did [20:20] It just means "install all packages available this this release" [20:21] what you have said (and the wiki doesn't clearly say) is that if you install 16.04.1 then you won't get packages which are in 16.04.2 automatically the usual system upgrade [20:21] you need to (this bit is poo) manually specify what you want [20:21] OK, now I got what you're asking [20:21] so even though something is now deemed "the default" for the major release "16.04" in general - you won't get the updates which are available [20:21] THAT is poo [20:21] My answer is: if I upgrade 16.04.1 to 16.04.2, then the difference I will have from a new 16.04.2 installation, will be the kernel and the xorg versions [20:22] Does that cover your question? [20:22] just kernel and xorg are held back then? is there anything I can set (perhaps in apt conf?) which disables this silly practice? [20:22] i.e. i don't care if my xorg and kernel upgrade to whatever it on the latest LTS ISO;s [20:22] It's not very silly. Kernel and xorg can cause serious regressions. [20:23] I maintain about 1000 school labs [20:23] then it would cause them for new users too! [20:23] i.e. fresh installs [20:23] In about 50 cases, I needed older or newer kernels and older or newer xorgs [20:23] So, in 5% of the cases, that auto upgrade you're mentioning, would cause boot issues [20:24] so if I want to throw caution to the wind (being a desktop user with nvidia drivers anyway - lots of automatic dkms fun) - can I disable this *in advance* of any updates? [20:24] i.e. so I just get them anyway without having to do anything? [20:25] There was a -hwe package available at some point, which depended on the last lts kernel/xorg [20:25] Let me see if they still support it... [20:25] that wiki article is mad - listing a huge apt-get line to manually scrounge for what should already be there as a part of normal updates [20:26] Nope it looks like the dropped the -hwe metapackage after 12.04 [20:26] It really shouldn't be there as part of normal updates [20:26] then it is isn't a part of 16.04 [20:26] You just haven't seen your share of kernel and xorg regressions yet :) [20:26] it's some kind of *forced* weird backport which you only get if your using a new ISO [20:27] Why, do you think it's bad that e.g. 14.04 supports 2 different kernels and xorgs? [20:27] if it *is* a packeg of 16.04 - it should upgrade.. if it *is-not* then it shouldn't - but then that shouldn't even be a package available without enabling another repo [20:27] How about "apt install php5"? Should that be automatically upgraded to "php6" and break all your scripts? [20:27] Or python2 to python3? [20:28] because it is inconsistent! i like to think two systems of the same major version - both with a recent upgrade task ran on them - WILL have the same package versions [20:28] It's OK to have different series supported in parallel [20:28] but these are weird special condtions which have been introduced [20:28] alkisg, err from what I can tell looking at 14.04 on that wiki page - these have the *same* package names [20:28] Unfortunately kernel major versions don't map with ubuntu releases [20:29] ah no they don't actually - my bad [20:29] they say xenial in them! [20:29] even more confusing [20:29] it's like a backport - but it's not a backport [20:29] It's possible to have 30 different kernels installed [20:29] it's a carefully choosey backport that only helps users who installed recently [20:30] There are no conflicts in the kernel package names [20:30] therefore one user of ubuntu mate 16.04 for 3 months might have a completely different bug to someone using it 3 days [20:30] even though the situations are identical [20:30] that's why it's bad! it's inconsistent! [20:30] Lantizia: before you start proposing new things, realize that dozens of clever programmers decided to implement it this way [20:31] Understand their reasons first, and *then* make suggestions :) [20:31] How would you propose to support the new xxx intel architecture, without giving regressions to all existing users out there? [20:31] have it as a choice on the ISO [20:31] Because unfortunately the kernel and xorg do cause serious regresssions on older hardware, when they start supporting new hardware [20:32] want 16.04 ? or want 16.04 with updated xorg/kernel from 16.10 ? [20:32] otherwise no one (no one non-technical) will know they have that [20:32] then if you want it the way 16.04 was originally released - you can... from the same install media [20:32] So now you still get fragmentation, and also a confusing question to the users [20:32] I don't think you solved anything that way [20:33] if you know 16.04 worked on your hardware 6 months ago... and now you have 16.04.1 - you *know* it'll still work if you don't pick a newer kernel/xorg on install [20:33] And, a much bigger 16.04.5 live cd [20:33] i.e. if its a class full of pc's of the same make [20:33] With 5 different options [20:33] 5? no just two [20:33] How about the 16.04.3 kernel? [20:33] You won't allow that one? [20:34] Not to mention wasting a whole lot of ram to dynamically add/remove packages while booting [20:34] it's 2 options.... if you boot 16.04.0 you get 1 option - just install.... if you choose 16.04.1 you get the choice of 16.04 or 16.04+new/kernel/xorg .... you get the same choice onj 16.04.2 and 16.04.3 (again just 2 choices) [20:34] ...as you can't have 2 different xorg versions installed in parallel, they conflict with one another [20:35] and those newer packages should be in a different repo! they're not 16.04! [20:35] Anyway, if you think you have a better idea there, you don't have to convince me, file a bug report and try to convince the maintainers [20:35] no one said have them installed in parallel! [20:35] Personally I think they handled it fine [20:35] You implied it when you asked for an option [20:35] That means that the live cd squashfs image would have one of them, [20:36] and then dynamically, upon boot, if you selected the other option, it would install all the .deb packages from the other xorg version, [20:36] wasting about 100 mb of your ram [20:36] the kernel/xorg used for the live cd session shouldn't be anything fancy anyway - it should be for maximum compatibility [20:36] you don't need 3d and other fancy crap to install an OS [20:37] i'm not saying text either - but bloody VGA will do [20:37] Now you're asking for yet another version of kernel/xorg to maintain and test [20:37] no i'm not! [20:37] :) [20:37] Anyway, I don't think it makes sense for us to talk about that [20:37] You're free to create a bug report and try to convince the maintainers [20:37] Me, I'm not convinced, but it doesn't matter at all [20:38] see it frustrates me how things like thunderbird get new versions automatically - even on LTS [20:38] introducting new bugs [20:38] same with firefox and libreoffice [20:38] this is only a ubuntu thing - not a debian thing! [20:38] That's because thunderbird doesn't maintain old releases with security updates etc [20:39] the security updates should get backported by ubuntu or debian then [20:39] like it is with any other package [20:39] They were doing that for years, and it was too difficult for them [20:39] so we suffer from firefox regressions don't we? :D [20:40] if only it was a choice :P [20:40] Yup, updates cause regressions, and choices are made between them [20:40] not by the user they're not [20:40] i'd understand if ubuntu is a rolling release - but it isn't [20:41] it's 1% a rolling release - this 1% isn't completely understood by all [20:41] I'm very glad with the decisions they made [20:41] kernel = no automatic new releases, firefox = automatic new releases [20:41] That's exactly what I would ask from them [20:42] they used to maintain a rep for firefox instead (the ubuntu team had a ppa) - i *liked* that [20:42] i knew then my firefox would stay the same (and perhaps other things like libreoffice/thunderbrid/etc) unless I opt-in to that PPA [20:42] there isn't even a way to opt-out of this - unless you freeze your packages [20:42] Browsers face a lot of threads; people should almost always have their latest versions [20:43] *threats [20:43] kernels face a lot of threats too :P [20:43] Not really, they don't have to revoke certificates [20:44] ah! there is a section for things like virus updates and ca-root-certs and things which are constantly changing *data* as opposed to actual binaries [20:44] so no - no deal! i don't buy that reason! [20:44] I think the reasons for maintaining firefox and kernel that way are mentioned in some UDS session [20:44] I certainly don't know them so I'm not trying to list them here, or convince you about them [20:45] I imagine a few things, I know of a few things, but not the whole picture. And from what I know, I feel that they did a great job. [20:45] If you don't, go ahead and read all their rationale. [20:46] their rationale is a) an easy life and b) to make LTS more usable for people who purposefully avoid all the 6monthly releases [20:47] the 6monthly releases being *dreadful* .. heck who ever heard of things being magically ready without bugs in april and october [20:47] i'd rather it be ready when it is ready - no 6monthly nonsense [20:47] Well if you already know both their rationale and why it's wrong etc, you have all the knowledge to file a good bug report and make your case there. Go for it. [20:47] On the other hand if you just assume you know... change your tactics. [20:49] alkisg, ok one last question... is 16.04.2 due in october when we have 16.10 ? (and thus linux-generic-lts-yakkity plus xorg-lts-yakkity)... [20:50] and if so - does that mean 16.04.3 will be in april next year when there is 17.04 ? [20:50] Lantizia: see this one: https://wiki.ubuntu.com/TrustyTahr/ReleaseSchedule [20:51] ...and add 2 years (because the 16.04 page isn't complete yet) [20:51] 17 February 19th Ubuntu 14.04.2 [20:52] so the point releases are purposefully february and august? [20:52] Yes [20:52] so no 16.04.2 until feb? [20:52] Yup [20:52] even though the kernel/xorg we'd get in that will be available next month? [20:52] I think the lts* packages arrive a few weeks before the point release, just to have enough time to test them [20:52] Yup [20:53] further poo I say [20:54] smelly too [21:05] I wanna copy my old /var from my previous install to a clean install, but all the uids and gids are different [21:05] is there any way to automatically fix these? [21:05] will dpkg-reconfigure do it? [21:06] allanViking: run this: dpkg -S /var [21:06] This will tell you that dozens of packages put files there [21:06] ...which means it's a bad idea to use your old one [21:06] Only transfer the dirs that no package ships [21:06] (subdirs) [21:07] I only copied the dirs I need and that are not systemic, the rest is from the clean install [21:07] these: [21:07] https://thepb.in/p/JZhpkmRvQQ1Ug [21:08] some of them dont have packages installed in the new system, (mysql, apache, postgre and stuff like that) [21:08] but I think I might need gconf for the new system to be configured the same way as my old [21:08] alkisg, I was going to put setxkbmap in my autostart (either in /etc/xdg/autostart or in my own ~/.config/autostart) - but would that fix it on the lightdm login too? i'm guessing not [21:09] allanViking: gconf isn't used anymore, I don't even have a /var/gconf directory here [21:09] weird, the fresh install had one [21:09] ubuntu mate 16.04.1 [21:09] Lantizia: # cat /etc/X11/Xkbmap [21:09] -layout us,gr -option grp:alt_shift_toggle,grp_led:scroll [21:09] Lantizia: put a similar line for your use case [21:10] allanViking: # dpkg -S /var/gconf [21:10] dpkg-query: no path found matching pattern /var/gconf [21:10] alkisg: im in a livecd atm, can't [21:11] but the only dirs that conflict in /var/lib are: [21:11] NetworkManager [21:11] upower [21:11] alkisg, i have a /etc/X11/xkb directory - is that what you mean? or just a new file in /etc/X11 ? [21:11] gconf [21:11] thats all [21:11] Lantizia: new file [21:11] im hoping when I install mysql, apache etc they will chown their /var/lib dirs correctly [21:13] allanViking: $ dpkg -L gconf2-common | grep /var/lib/gconf [21:13] /var/lib/gconf [21:13] /var/lib/gconf/defaults [21:13] /var/lib/gconf/debian.mandatory [21:13] /var/lib/gconf/debian.defaults [21:13] This means that gconf ships those files, it doesn't generate them [21:13] So if you put your old ones there, they will probably be replaced on updates [21:14] alkisg, that line to put into the file - can that be exported somehow from setxkbmap? (since it already thinks it is set correctly anyway) [21:14] i.e. dump what setxbkmap thinks is correct into Xkbmap file [21:14] For files generated on postinst, yes it's possible that they use chown, but in general you should fix those yourself and not rely too much on package postinsts [21:14] Lantizia: what's the output of setxkbmap -query? [21:14] alkisg, it isn't a one liner like that [21:14] I know [21:19] Anyway, good night, /me waves... [21:20] thanks alkisg [21:33] i love you ppl [21:33] <3 [21:39] alkisg: thx for your help, restarting to see if it helped :) [22:25] if I wanna backup all my mate settings (panel config especially), the is it enough to backup ~/.config ? [22:25] if I wanna backup all my mate settings (panel config especially), then is it enough to backup ~/.config ? [22:33] sorry for the flood, for some reason I thought shift + edit would fix the last row [22:40] 2 [22:41] testing from 1983 TRS-80 Model 100 [22:52] allanViking: Panel settings are afaik stored in GSettings, which by default uses the dconf backend, so backing up ~/.config/dconf/user is enough, although I suggest you consider using the dconf command-line tool to dump the contents of that db in a textual form: dconf dump / [22:52] You can direct that into a file: dconf dump / > dconf-settings.txt [22:53] thx [22:53] np [22:54] allanViking: the dconf tool can also help you re-create your db as well when you want to restore it later. [22:55] nice and small :) [22:55] my .config was 1.4GB [22:55] is there any way to grep out systemic packages from a dpkg package dump? [22:55] I want to install all my old apps, but I dont want to revert anything [22:56] I have the output of dpkg --set-selections