[00:01] elisa87: check what process has the port: "sudo ss -lpt sport = :http" [00:01] Ben64: I changed 80 to and still get the same error [00:01] ~$ sudo ss -lpt sport = :http [00:01] State Recv-Q Send-Q Local Address:Port Peer Address:Port [00:01] LISTEN 0 128 :::http :::* users:(("docker-proxy",15779,4)) [00:01] Ben64: TJ- [00:01] elisa87: there you go then, docker [00:02] is it safe to kill -9 docker on EC2? [00:02] I have not been using it [00:03] I was here asking about this before and still need help. I made a mess of my direcotory icons a while back and it's really causing a problem. I've read the articles and there doesn't seem to be a complet and up-to-date answer that I've seen. I've posted on the forum and not gotten a soln. I don't want to reinstall the system to fix it. Is there anyone willing to field this quesiton and help me get back to the default direcotry icons on my system. [00:03] ? [00:03] ^ I run 16.04 LTS [00:03] buggaboosanchez, made a mess of how [00:03] so it still fails but without that error $ sudo service apache2 start [00:03] * Starting web server apache2 Action 'start' failed. [00:04] Ben64: TJ- [00:04] elisa87: best to disable the docker service [00:04] and use a paste service, paste the full command, output, and apache log maybe [00:04] [Mon Sep 18 00:03:43.218199 2017] [ssl:emerg] [pid 30647] AH02312: Fatal error initialising mod_ssl, exiting. TJ- Ben64 [00:05] elisa87: you need to check the apache log file(s) under /var/log/ to find out what went wrong - probably you hav bad syntax in the apache config files [00:05] that is what I did [00:05] [Mon Sep 18 00:03:43.218159 2017] [ssl:emerg] [pid 30647] AH02240: Server should be SSL-aware but has no certificate configured [Hint: SSLCertificateFile] ((null):0) it is weird I get this error because I thought I setup the SSL in 000 file [00:06] elisa87: sounds like the mod_ssl config is incorrect [00:06] well there you go, you need a certificate [00:06] I have a certificate [00:06] it says you don't [00:06] elisa87: but apparently apache doesn't see it [00:07] I have set the path to my new certificate here sudo vi /etc/apache2/sites-enabled/000-default.conf Ben64 TJ- [00:07] bazhang: A few mos ago I thought it was a good idea to customize my directory icons by goint into the properties of each and every direcotry ( one-by-one ) , clicking the little icon in properties to change the icon and then navigatiing to some uniqe, custom icon I had stored. Now there are several different directory icons in use throughout the system and it's hell trying to work with. [00:07] this is what I have SSLCertificateFile /etc/apache2/ssl/4757dc1c480db857.crt [00:07] you can keep saying you did it right but the error says you don't [00:07] Got this from godaddy [00:07] elisa87: you really need to talk to the apache/http channel they're the experts on this [00:08] also have rgis SSLCertificateKeyFile /etc/apache2/ssl/buhcomp_com.key [00:08] bazhang: I know, I'm a retard, what can I say? I did it and it's done. I need a way out. [00:09] buggaboosanchez, reinstall ? [00:09] good lesson not no mess with icons [00:09] buggaboosanchez: those changes are stored in either 'hidden' .directory files in each directory, or under the user config at ~/.local/ or ~/.config/ [00:10] here is more information https://pastebin.com/PGpNBYEP Ben64 TJ- [00:11] buggaboosanchez: you could try finding them all: "find $HOME -type f -name '.directory' [00:11] oerheks: I really, really, really wish there were some other way. Innevitably, everyone I say this to will refuste to believe it, but the truth is I don't have nor can I afford a way to back up my personal data in order to do so ( and that probably isn't going to change any time soon ). <--- I don't see how I can reinstall [00:12] buggaboosanchez: are the dirs in question only in your homedir? [00:12] Hello, I have been struggling to run 1080p movies h264 encoded on Ubuntu 16.04, the movie runs fine but the frames are kind of "blurry" whenever there is camera movement. I'm using VLC and I have GTX970 GPU and i5 4690k CPU any help would be apprecaited [00:13] TJ-: Thanks. If I had the actual default image available I would be willing to got back through each and evcery directory property and set it to the correct image ( like I did when I screwed evcerything up - but fixing it this time ). [00:13] ^ Does anyone have the default image ? [00:13] buggaboosanchez: if you can find a file that does have an icon= type of setting pointing to one of your customisations we can then craft a shell script to search and replace those entries with the default [00:13] Moe_A, that could be a property of how h264 is actually compressed and your ability to see finite details; how serious is it? [00:14] buggaboosanchez: it should be in your filesys already, not sure where it will be located for Unity tho [00:14] jer, it hurts the eyes after few minutes of watching a movie.. the same movie runs fine on windows 10 though' [00:15] buggaboosanchez: for example did you customise the Desktop folder icon? if so, show us the $HOME/Desktop/.directory file contents [00:15] TJ-: I tried find $HOME -type f -name '.directory but must have gotten it wrong. It's ok. (2) what do you mean by 'type of setting' above? [00:16] buggaboosanchez: there should be a closing quote mark there [00:16] Moe_A, have you tried to reinstall the nvidia drivers, or change the version you are running? [00:17] buggaboosanchez: " find $HOME -type f -name '.directory' " [00:17] buggaboosanchez: I'm assuming the directories you changed are all in your home directory, not system-wide? [00:17] TJ-: I also have a simlink between my user dir and Desktop ( so my entire home dir appears on the desktop ). I'm not sure if a 'Desktop' icon even exists anymore ( or maybe is not seen if it does ). [00:17] TJ-: should that not be -type d ? [00:18] jer, yes I did reinstall but didn't try a different version. When I'm gaming the graphics are normal though [00:18] venzen: no [00:19] find $HOME -type f -name '.directory' yeilds : /home//DEVELOP--A--CONTENT--AUTHORING--PROGRAM/SourceCode/scribus-1.4.6/scribus/.directory [00:19] buggaboosanchez: that sounds like you'll be creating an endless loop of directories! $HOME > Desktop => $HOME > Desktop ... [00:19] buggaboosanchez: OK, so that won't be it then, not in the .directory files [00:20] buggaboosanchez: i guess it'll be written into some config under $HOME/.local/ or $HOME/.config/ then [00:20] TJ-: That part has been working fine. It's the icon image(s) [00:20] I will look [00:20] buggaboosanchez: what's the name of one of those custom icons, exactly? I can craft a search for you [00:21] TJ-: I just recalled how this got so out of control. It looked great right after I changed all the icons but I didn't realize that new dirs that I created after that would not have the custom image but the default instead. That's how I ended up with multiple dir images on the system [00:22] TJ-: let me see if I can get a name === leion1 is now known as leion [00:28] An icon used on one, single dir is "camera.png" Another one is ... I can't seem to find the other one. Then the reast are a third one. [00:28] buggaboosanchez: " grep -rn 'camera\.png' $HOME/.local/ $HOME/.config/" [00:28] Idk how to find out the name bc all I know how to do is nav through the properties and that doesn't tell you [00:28] buggaboosanchez: if that search gets a hit you've at least found out where the settings are stored, and can start to figure out how to change/revert [00:28] ok [00:29] how can i install hexchat? [00:29] i don’t know it exactly name is haxcht or something? [00:30] goIIum: it's called hexchat [00:30] install it with apt, as like: sudo apt install hexchat [00:31] thanks :) [00:31] it's by default on the Ubuntu repos [00:31] np :P [00:31] default [00:31] :) [00:32] The program 'hexchat' is currently not installed. You can install it by typing: ASpacy [00:32] ? [00:32] TJ-: Not sure I should publish the output of that grep. There are 3 results. All 3 are binary files. 2 have to do with zietgiest/sqlite somthing and the top entry is: /home/myuname/.local/share/gvfs-metadata/home matches [00:32] what I said, install it via apt, run this command: sudo apt install hexchat [00:33] ASpacy: thanks [00:33] ASpacy: anyway ubuntu user use hexchat to log in irc? [00:34] what do you mean? [00:34] hexchat is most popular irc app? [00:34] there is more popular irc app than hexchat? [00:35] buggaboosanchez: Ahhh, well, we're making progress at least [00:35] TJ-: cool [00:35] ASpacy: ? [00:35] thank you for helping btw [00:35] goIIum, there is what you like, there is NO best [00:35] buggaboosanchez: this answer suggests you can just delete all those files: see https://superuser.com/questions/100545/files-in-local-share-gvfs-metadata-in-ubuntu#144238 [00:35] goIIum: HexChat is one of the most used clients, although it depends mainly on your likes [00:36] thanks ASpacy :) [00:36] ok bazhang [00:37] ASpacy: yeah i installed it, but you know networks name? === leion1 is now known as leion [00:38] Ubuntu servers(freenode) is correct? ASpacy ? [00:38] goIIum: it includes a list by default, for example this one - freenode/Ubuntu servers (freenode) [00:38] yup [00:38] thanks :) [00:38] that one is for here [00:38] np! [00:38] buggaboosanchez: this seems to confirm that is where the "emblem" (icon) data is: https://stackoverflow.com/questions/10874702/gnome-where-does-nautilus-store-emblem-data-and-how [00:41] TJ-: Sorry, my computer froze ( bc of "ubuntu browser" ) and I had no choice but to reboot. I lost the link you gave. Can yo paste it again please? [00:42] buggaboosanchez: this answer suggests you can just delete all those files: see https://superuser.com/questions/100545/files-in-local-share-gvfs-metadata-in-ubuntu#144238 [00:42] buggaboosanchez: this seems to confirm that is where the "emblem" (icon) data is: https://stackoverflow.com/questions/10874702/gnome-where-does-nautilus-store-emblem-data-and-how [00:43] brew cask install font-inconsolata [00:43] :) [00:43] TJ-: I'll read that. Ty [00:51] TJ-: And you think that deleting those file ( and restarting? ) will solve my problem? [00:52] there are about a dozen files there some of them start with "uuid" and end with some kind of hash or something ( letters numbers ) [00:52] ^ I thought things with a uuid were a block device or something that you shouldn't delete? [00:53] buggaboosanchez: UUID is a hash and can be generated and used as identifier for anything, including block device or as a file name [00:54] unique universal identifier [00:54] I'm wondering what sequence of events would transpire if I did rm -R that dirs content/ ? [00:54] Would it be regenerated after a reboot? [00:55] not sure, what does the SO answer say in the link TJ- provided? === leion1 is now known as leion [00:55] hi, are users created during install added to the sudo group? [00:56] venzen: I'm not sure it applies bc the op's question seems to have a different goal than mine. Is there any way for me to make the connection that it applies to my sitch? If I bork this system I have no way to recover [00:56] I'd lose all my personal data [00:56] If something went wrong [00:57] Hi, I have a question about uncompressed tar archives and deduplication. If I have backup software that checks files for duplicate chunks and only uploads the diff to the server, will having all the files in an uncompressed tar archive still allow deduplication? Or will the tar files look completely different even if only one byte on the files in it have changed? [00:58] buggaboosanchez: well you can backup that specific dir to somewhere else in your homedir, then delete its contents. if it doesn't work you can always copy its contents back into place either as your own user or, worst case, as the root user [01:00] venzen: It sounds fairly certian that deleting that dir will, at least, not be catastrophic then. ( I know TJ-said as much but I'm nervoise enough about the sitch to want a confirmation ). Ty and sorry Tj [01:00] It's important [01:00] buggaboosanchez: I'd do something like "mkdir $HOME/gvfs-metadata; rsync -delete -av $HOME/.local/share/gvfs-metadata/ $HOME/gvfs-metadata/" to move the files to a backup location [01:00] buggaboosanchez: unlikely to be catastrophic, make a backup just to be sage [01:00] *safe [01:01] TJ-: Thank you. Here goes nothing :0 [01:01] I willl backup ty [01:01] buggaboosanchez: then log-out, log-in to refresh [01:01] buggaboosanchez: be sure to delete the contents and not the dir itself :) [01:01] I'll come back let you know how it worked [01:01] venzen: ok [01:01] see you in a min [01:07] Here goes the reboot [01:11] lol [01:11] hey ahoneybun [01:17] TJ-: venzen: Nothing catastrophic happened ( you know ) - yay! So now all the emblems are the same across the system but they aren't the default either ( they aren't horrible, but they aren't great either ). Here is what it's like now ... Emblem is: https://imgur.com/a/5lPQs And the content of the gvfs-metadata dir is: https://pastebin.com/axW2bSJb [01:17] I'm curious ... Is it possible to use this method to set the emblem of my choice but to have ti system wide as well as new dirs created have it when created? === deww is now known as eww [01:18] buggaboosanchez: OK, so now create a new user, log-in to it, then copy over the contents of it's /home//.local/share/gvfs-metadata/ to /home//.local/share/gvfs-metadata and the original emblems and other stuff should be restored [01:18] buggaboosanchez: no idea, you'd need to consult the source-code [01:20] TJ-: Ok, sounds great. One follow up questoin. If I do that ( copy contents from newuser ) will I also have to do a chown and / or chmod to set everything to ? [01:21] Would I have a ownership issue I mean? [01:24] buggaboosanchez: You'd need to do it as root, using 'sudo ...' so yes, you'd need to re-won the files after with "sudo chown -R $USER:$USER $HOME/.local/share/gvfs-metadata" [01:24] s/re-won/re-own/ [01:25] Ok. I thought I remember how a user with admin privs could assume the role of another user and issue commands on thier behalf [01:25] I don't recall the command(s) - we did it in school years ago ( I think ) [01:26] buggaboosanchez: so you'd do "sudo rsync -av /home//.local/share/gvfs-metadata/ $HOME/.local/share/gvfs-metadata/" then "sudo chown -R $USER:$USER $HOME/.local/share/gvfs-metadata" [01:26] ahh [01:27] TJ-: I really appreciate it man [01:27] ty [01:27] buggaboosanchez: in those commands you can pretty much copy-paste them, the only thing to change is to whatever you chose to use. The $HOME and $USER will be replaced by the correct values [01:27] right on [01:27] buggaboosanchez: I'm assuming you execute these commands whilst logged in as your current user, not the of course [01:31] TJ-: got it [01:31] thanks [01:40] TJ-: Running that command I get: 'rsync: change_dir "/home/mynewuser/.local/share/gvfs-metadata" failed: No such file or directory (2)' , total size is 0 speedup is 0.00 and rsync error: some files/attrs were not transferred (see previous errors) (code 23) at main.c(1183) [sender=3.1.1] in my output.. sudo ls newuser reveals only 4 files and no dirs in that 'mynewuser' directory. What did I do wrong? === wgrant__ is now known as wgrant [01:41] buggaboosanchez: did you log-in as that new user yet, to have the configs created by the gvfs daemon [01:42] buggaboosanchez: you need to log-in as , do something with the file manager just in case, check that the .local/share/gvfs-metadata directory now exists, then log-out and back into your regular account [01:42] swift110: heyo [01:43] TJ-: I did not know that. Thanks [01:43] buggaboosanchez: it's almost 3am here so I have to go... hope you get it all sorted out [01:46] TJ-: Have a peaceful sleep man [01:46] peace === brainwash_ is now known as brainwash === Mrokii_ is now known as Mrokii [02:36] what's the difference between 'apt upgrade' and 'apt full-upgrade' and 'apt dist-upgrade' ? [02:36] full-upgrade and dist-upgrade do the same [02:36] !full-upgrade [02:36] !dist-upgrade [02:36] A dist-upgrade will install new dependencies for packages already installed and may remove packages if they are no longer needed. This will not bring you to a new release of Ubuntu, see !upgrade if that is your intention. [02:37] hmm [02:37] !upgrade [02:37] For upgrading, see the instructions at https://help.ubuntu.com/community/UpgradeNotes - see also http://www.ubuntu.com/desktop/get-ubuntu/upgrade [02:37] full-upggrade gives newer kernel and such, if available === g2 is now known as g2[ATL] [03:09] how do I see what version of ubuntu I'm running from the command line? [03:12] lsb_release -a [03:12] lsb_release -a [03:12] thanks :) [03:14] Xenial software updater automatically updated my firefox; I need to go back to a previous version of ff. How can I go about that? [03:23] klf: why would you need to go to backwards? if it got updated it's for a good reason [03:25] Ben64, In the case of FF, I'd disagree. FF is transitioning to e10s, and dropping all the XPCOM stuff, as a result many many extensions are affected. [03:25] for good reasons [03:30] if their market share depends heavily on extensions, and the upgrades they make aren't supported by 3rd party developers then it's bad. [03:31] we're digressing. if anyone else wants to answer my uestion, that'll be great. [03:32] the answer is you shouldn't and can't, unless you want to delete your profile, purge firefox and install from elsewhere [03:34] Has anyone else been having storage volumes being, sort of, conflated in 16.04? I've been getting various disk storage warnings and pop-ups, some of which seem to indicate that baobab, df, etc are sometimes summing volume size in a non-intuitive way [03:36] I have a couple of external volumes, that seemed to get glommed onto my internal root / directory in a way that 16.04 loves to complain about [03:36] don't get me wrong, this is sometimes me. I have way too much going on on a 128G drive, sometimes though... [03:48] does anybody know how to uninstall a program from wine? [03:53] gt8ost4l: there should be a menu entry for that, but if its not obvious in Unity see: https://askubuntu.com/questions/101064/uninstall-a-program-installed-with-wine [03:58] vensen:i already tried that [03:58] it still doesnt go show up [03:58] the latest firefox i mean [04:03] nevermind i installed the wrong version [04:39] whats up ppkl [04:40] ppl [04:45] doing a clean install of 16.04 LTS, NVIDIA binary drivers are not installing through the Additional Drivers GUI, I did click apply changes, nothing happening, progress bar is stuck at the begining... [04:45] BRB [04:47] shit mang did you try software updates [04:47] the program [04:49] My calculator app (and some others such as gedit) open with artifacts from the window behind them in their main text area. [04:49] i think its under additional drivers [04:49] I could post a link to a screencap, would that be ok? [04:49] (to demonstrate the issue) [04:49] go for it [04:49] https://imagebin.ca/v/3am9Qjkd1jE6 [04:49] did u design that program === rkta_ is now known as rkta [04:50] or you just coding in the background [04:50] the background one in vim in my terminal? [04:50] Just showing that the window behind is showing through the calculator. [04:50] ok [04:51] It's not a transparency; if I move the calc around, it keeps the same image in its result window. [04:51] or result pane, whatever it's called. ;) [04:51] so what are you trying to achieve [04:52] Trying to achive being able to use the calculator or gedit again without background artifacts appearing. [04:52] Hey guys [04:52] damn maybe try ps aux [04:53] and then kill [04:53] Is there a special way to add an applet in KDE? [04:53] I'm trying to add the aplet for redshift [04:53] or system monitor depending on what desktop environment you are using [04:53] Already installed it [04:53] Just not sure how to actually enable it [04:54] I'm not explaining clearly, I know. If I open gedit, or the calc, the background gets captured as a portion of the application's main window area. This has been happening for awhile -- not related to the process tree. ;) [04:54] damn im not really sure [04:54] I thought the screen cap would demonstrate but I guess not so much. [04:55] I could probably find an answer somewhere if I had an idea what terms to search for, but it's rather difficult to conjure the right search terms to find anything like this. [04:56] yea im not sure how to resolve your problem [04:56] So no ideas on my question? (bump) [04:57] Nevermind, got it. [04:57] alright. I've been living with it for awhile. I can check back here during daylight hours (localtime) to see if a different set of individuals here might have some thoughts on it. [05:00] Odd though. I can even minimize the calc app, remove the window that was behind it (the terminal in this case), raise the calc again, and that same background residual image is still there. [05:03] strange error you got therer === g2[ATL] is now known as g2 [06:13] hi, after open laptop ubuntu is fast to display the desktop, but reconnect to wlan takes about 20 to 30 seconds and during this time, the mouse pointer is not responding, any hints? (17.04) === juboxi is now known as jubo2 [06:52] Hi guys - is there any reason why a Ubuntu 16.04 machine running Sendmail would have issues sending to a new TLD like .energy? [06:53] sorry, sending /from/ that TLD [06:53] we have an identically setup CentOS 7 box with the same app trying to send mail, and it works fine - in staging [06:55] gnomethrower: no hints in the logs? [06:56] ducasse: Not that I initially found, but let me take a better look now before we continue :) [07:01] ducasse: I think I've found the culprit [07:01] one sec, sanitizing [07:05] Sep 18 06:59:41 foobar-web01 sm-mta[32361]: 321123321: ruleset=check_rcpt, arg1=, relay=localhost [127.0.0.1], reject=553 5.1.8 ... Domain of sender address testemail@client.energy does not exist === jelly-home is now known as jelly [07:07] can dns on the host find the domain? [07:10] gnomethrower: ^^ [07:18] ducasse: Sorry to disappear [07:18] but I think we found the issue [07:18] most tools claim that the client domain does not exist [07:18] and indeed it is registered but has no records besides NS [07:19] gnomethrower: right, check the records on your dns server [07:19] and that it responds at all [07:20] ducasse: at this stage I think what has happened [07:20] is that the old version of sendmail on the CentOS 7 box is (necessarily) very old compared to Ubuntu's [07:21] and that at some point sendmail added a check that your from email exists and is on a domain that exists [07:21] so the fix should be simple; just add some records and make sure, as you said, that it responds correctly [07:21] but that explains why this "works on staging" issue is occurring [07:23] either that, or someone put in a hosts entry or something on the centos box [07:29] ducasse: That's a point - we could use that to test this [07:30] (temporarily of course!) [07:32] gnomethrower: depends on what sendmail looks for, i really can't answer that. [07:33] ducasse: either way thanks for your help :) [07:33] no problem, hope you'll get it working :) === r0Oter is now known as r00ter [07:39] ducasse: Spot on. Adding client.energy to /etc/hosts pointing to 127.0.0.1 bypassed the check just fine and mail now works [07:39] We're going to make them add a proper A record etc ASAP, but for now that solves the mystery of the failed email forms ;) [07:39] gnomethrower: brilliant, then you just have to fix dns ;) [07:39] Thank you so much for the suggestion, it let us properly verify what the issue waas [07:39] which is sometimes rare! [07:40] you're welcome [07:40] yep [08:03] hi [08:07] ubuntu suddenly says it wants to update "screen" from trusty-backports [08:07] this contradicts the article about backports saying "For this reason, the Backports Team recommends configuring the package manager to only install backported packages when they are explicitly requested, which is the default for all Ubuntu releases after and including Ubuntu 11.04 (Natty Narwhal)." [08:07] (I'm using 14.04) [08:09] then you must have backports enabled [08:10] obviously, but as said, it should not try to install packages from there at random [08:11] http://paste.ubuntu.com/25563703/ here's "apt-cache policy screen" [08:13] right, you have them enabled though [08:16] Ben64 correct, but that doesn't help [08:16] i will not disable backports because i need a couple of packages from there [08:16] not sure what you want then [08:17] but i dont want autoupdate to switch screen to trusty-backports [08:19] so either reset it back to default or install the newer screen [08:20] not sure why the big deal over a newer version of screen [08:20] i'm lucky to have caught this issue before installing the update and i'm concurrently working on looking for other affected packages [08:20] Ben64 i suggest you read https://help.ubuntu.com/community/UbuntuBackports before going on [08:20] i already did [08:20] especially sections 2 and 3 [08:20] if you read it you'd know how to set it back to default [08:20] about security and stability === leion1 is now known as leion [08:24] Ben64 i noticed this because i have no key for trusty-backports. when muon updater asked me to acknowledge updating an unauthenticated screen, i got suspicious [08:24] i'm pretty sure this is new behaviour [08:24] nope [08:25] i dont quite know how this works... what is the default behaviour if i dont have any /etc/apt/preferences or /etc/apt/preferences.d/* [08:26] that was the state of affairs when i started analyzing [08:26] added pinning with priority 100 after reading that article [08:27] my tplink tl-wn821n v5 usb wifi stick can find available networks but router doesnt accept password even though password is correct (it works in windows) what to do? [08:29] "apt-show-versions -i; apt-show-versions -b | grep trusty-backports" doesn't show anything, which is a relief to me and further evidence that the default was recently changed. [08:30] changed by you or something you did [08:32] i must have been sleepwalking [08:38] Ben64 see, if you can't even tell me what the default behaviour is, in the absence of any preferences file, I'll consider it more likely that ubuntu changed something that you are unaware of than that I did something that I'm not aware of. I've had backports enabled for like a year now and never had this happen. [08:39] again, no. you posted what the default behavior is, it's in the link you posted a couple times [08:39] http://de.archive.ubuntu.com/ubuntu/pool/main/s/screen/ and if you look at the last modified date here, screen 4.2.1 hasnt updated in 3 years [08:39] the default didn't change, you did [08:39] Ben64 what did i change? [08:39] you'd be more likely to know that [08:40] actually, you're right, i did change the value from the default. half an hour ago before coming here to tell you that ubuntu's default had changed [08:41] i assume ubuntu's default is supposed to be 100 [08:41] apt-cache policy showed 500. [08:42] in the past, muon never updated any of my packages to the version from trusty-backports automatically [08:42] the package it wants to update is ancient [08:43] i now created a config file in preferences.d which pins it to 100 [08:44] that remedies the situation [08:44] but i am worried about the defaults nonetheless [08:44] again, the defaults haven't changed [09:13] Ben64 ok let me take you through this one final time: [09:14] https://help.ubuntu.com/community/UbuntuBackports#Configuring_Backports_for_Manual_Install the note in this section says "Note: This is only necessary on Ubuntu releases before (but not including) Ubuntu 11.04 (Natty Narwhal); releases after and including Ubuntu 11.04 are configured in this mode by default." [09:14] i never touched that directory, so it is in its default state [09:16] you're assuming that is the only way to change it [09:16] that is the only way mentioned in the article [09:17] = default backports enabled, do you want to roll back? [09:17] the default hasn't changed, yet it is changed on your system. the preferences file wasn't modified on your system. the only logical conclusion is that it had changed elsewhere [09:18] oerheks nothing to rollback, fortunately [09:18] indeed, all you can do is a fresh install, choose not to update during install, and disable backports after first boot. [09:19] TomyWork: Best thing to do is pull the source-code *repo* for apt, find where the default priority is hard-coded/defaulted, then look at the commit history for that file [09:19] i wonder why you "TomyWork i will not disable backports because i need a couple of packages from there" [09:19] i should probably pull an installation iso, fire up a VM and try this from a fresh install for myself [09:19] oerheks: backports packages aren't supposed to be pulled in without a specific manual command by the admin [09:19] oerheks well i dont want *all* packages from backports, just a few :) [09:22] TJ-, but backports are enabled by default.. no admin needed [09:22] oerheks enabled, but not pulled from [09:22] oerheks: incorrect, the policy is that when backports repo is available, the *default* pinning should prevent those packages automatically replacing existing packages unless the admin commands the install [09:23] ... huh? [09:23] oerheks: like so https://hastebin.com/gakaleqola.rb [09:23] oerheks: otherwise the system could be automatically upgrading packages the admin does not wish to be touched, since 'backports' repo is for special cases where no other course short of a manual source-code build is possible [09:23] i see this different, you will automaticly get backported packages, as security updates [09:24] Ben64 that's what it says for you? any preferences files? [09:24] TJ-, backports <> proposed.. [09:24] TomyWork: nope [09:24] oerheks: backports aren't security upgrades [09:24] ls -alh /etc/preferences* [09:24] then that is very strange [09:24] why do you think i kept saying it hasn't changed [09:25] cause you hadn't backed it up with anything, of course [09:25] ha [09:25] oerheks: no. -proposed is the feed for -updates. -backports is for packages from later releases that provide additional features that really ought to be in the earlier release [09:25] i don't go around making things up [09:25] how would i know? :) [09:25] you could have asked if you were that unsure [09:25] well i instead tried to back up my position [09:26] by assumptions [09:26] no, not really [09:26] basically by the same output you gave [09:26] On ubuntu16.04 i am quite regularly running fsck from live boot to recover and login back into the machine.. how do i find out the issue causing this [09:26] except for minor details like a 5 instead of a 1 [09:26] all you were going on was your experience on one system [09:26] and you checked multiple systems? [09:26] i have a ton of 14.04 systems all with the same result [09:27] ah [09:27] well i dont have more than one at my disposal [09:28] TJ-, hmm indeed, backports contain no security fixes, i was wrong about this, for years... [09:28] :-( [09:28] 'selective newer versions' [09:29] well i have no idea what's going on... i havent updated screen since at least 2016-09-29 [09:29] and i dont know what, besides the "preferences" files affects pinning [09:29] hethkar, 'regularly running fsck ' .. sounds like a faulty disk [09:30] check disks > s.m.a.r.t. [09:32] hethkar: can you describe what actually happens? you've given us no details. [09:33] TomyWork: I'm going through the apt source and history for the version in trusty-updates/ecurity. Can you tell me the "apt-cache policy apt" installed version please? [09:34] Installed: 1.0.1ubuntu2.17 [09:35] either from trusty-updates or from trusty-security [09:35] TomyWork: source repos on bzr are a mess; they show the -security version behind -updates! [09:35] odd [09:36] ducasse: the user cannot boot into the ubuntu 16.04 OS whenever that happens it goes to the initramfs screen. So i use usb live boot and run sudo fsck -y on the devices and only then it will boot into the OS.. [09:36] TomyWork: and the latest from -updates as 2.10 ! [09:36] This is the smartdetails for the disk http://dpaste.com/05QG556 [09:37] TJ- hmm, that's the official bzr? :) [09:39] Lag: 639 ms [09:39] TomyWork: yes, the branches linked from launchpad. The source-code side has always beena mess, not knowing where to find the correct repo, because dev's/packagers can do it how ever they want - no consistency. I've lost track of the number of times I've been chasing a clue on a bug only to discover I'm using an obselete source code repo. Sometimes the only thing to do is pull the actual source package which [09:39] obviously doesn't contain the commit history [09:39] err, ignore that [09:40] TJ- so it's not just me? :) [09:41] TomyWork: and the version in the git repos for trusty-proposed is 2.15 ... getting closer! [09:41] TomyWork: oh no, it's one of those Monday mornings :D [09:41] if you're lucky, some maintainers accidentally leave their git or bazaar directory in the source package :) [09:41]