libera/#devuan/ Thursday, 2019-01-17

gnarfaceSpringlebutt: check the output of dmesg for complaints about missing firmware files at boot time.  some atheros devices were recently hit by a bug where they forgot to put all the firmware in the package apparently (still waiting for upstream to fix this afaik but you can also find the firmware files yourself.... somewhere out there.  not sure where, sorry)00:05
gnarfacese7en: if you have lots of free space on the disk, resizing partitions to steal space from another less crowded one is an option.  the only other thing that comes to mind involves you buying a second harddrive00:06
se7enI can't easily do that, I have encrypted LLVM00:07
gnarfacewhich one?  resizing or mounting additional drives?00:07
gnarfacedid you mean LVM, not LLVM?  i thought the whole point of LVM was to make this sort of thing easier...00:08
gnarfacei've never touched it myself though, sorry00:08
se7enLVM00:09
gnarfacenormally it would be an easy thing to just boot a livecd and use gparted to shift partition sizes around00:10
se7enThe problem is that an ENCRYPTED LVM is currently unsupported by many graphical partition managers00:10
gnarfacewell, that's believable.  i can't tell you off the top of my head if there are exceptions.00:11
gnarfaceare you able to just get a second drive and mount /usr on that?00:11
se7enWhat?00:12
se7enI have a single drive00:12
se7enYou misunderstand the problem00:13
gnarfacei think i understand the problem.  what i'm telling you is that the easiest solution is to throw money at it.00:13
se7enMy drive fills up too fast and the problem effects apt the most00:13
se7enIt is a 500GB drive00:13
gnarfacehow much space is on the /usr partition?00:14
se7enit's labled as 6539M by mc00:14
se7enFree00:14
gnarfacehow much total?00:14
gnarfacefree and used combined?00:14
gnarfacejust for /usr00:14
gnarfaceunless /usr is not on it's own partition...00:15
gnarfacewell, either way actually, but we need to know that too00:15
gnarfacewait, are you saying that it gives you an error that /usr is full when there's 6539M free on it?  that's ... a different issue00:16
gnarfaceapt might complain it is running out of space if the partition /var/cache/apt is on fills up, too00:17
gnarfacethe details would be in the error text i think00:17
se7engnarface: but it does00:17
se7enit does run out of space00:17
se7enThe partition is 6539M but it obly has 239M free00:17
gnarfaceok00:18
se7enAnd it fills up fast00:18
gnarface6.5GB isn't enough for the kitchen sink00:18
se7enYeah, I know00:18
gnarfaceyou want to be thinking more in the range of 20-60GB before you can be careless about how much you install, and that's only if /usr is on it's own partition00:18
se7enI used the guided paritioner on the Jessie installer00:19
se7enI don't think /usr is00:19
se7enLemme check here00:19
gnarfaceguided partitioning defaults might have put /usr alone, i'm not sure.00:19
se7enI know that /tmp and /var are both individual partitions00:19
gnarfaceok, that's good.  / and /usr should ideally be separate too00:20
se7enYes, /tmp, /var and /home are all their own part00:20
gnarface and /home should also be separate00:20
se7enBut /usr isn't00:20
gnarfacehmm.  /usr is shared with /?00:20
gnarfaceand also /usr/local, i assume?00:21
se7enYes00:21
se7enThat is right00:21
se7en/usr/local is same part as /usr, but not a unique partition00:21
gnarfaceyea, 6.5GB isn't gonna be enough, sorry.00:21
gnarfacewhere'd it put all the space then?  /home?00:22
se7enWell, how do I rectify this00:22
se7enThat is 425G00:22
se7enSo yeah00:22
se7en500GB dribe00:22
gnarfacethat's a setup that would make sense if you were reserving space for a Steam install.  next time you might want to take a little more manual control over it.00:23
se7enAnd I have 1.02MB unallocated to00:23
se7entoo00:23
gnarfaceeh, that's harmless00:23
se7enWell, what can I do to fix this00:23
se7enI'd rather just quick-fix it in gparted than in fdisk, but gparted has no support00:23
gnarfacegoogle suggests to me that there IS a methodology for doing it.  i can't give you a simple one-liner though, this isn't something i've done before with LVM and LUKS00:24
gnarfacethe good news is you probably don't have to reinstall, if you can get help from someone who has done it before00:24
gnarfacese7en: i would recommend you hang out and try to talk to someone who is more familiar with LVM and LUKS, but you should probably check some distro wikis for howtos too.  i see this topic on the arch linux wiki specifically coming up in the top of my google searches.  maybe there's an equivalent page on the debian wiki somewhere, or mabye the process form ubuntu will be similar enough to not break things...00:27
se7enok00:27
gnarface*process from ubuntu i meant to say00:27
gnarfacese7en: i take it 'pvresize' can't do it?00:29
gnarfaceor maybe 'lvresize' ?00:29
gnarfacehmm.  scanning the arch wiki it looks like you'd use a combination of cryptsetup, lvreduce/lvresize or the like, and whatever filesystem tools go with your filesystem (probably e2fsprogs if you don't know)00:32
gnarfacelooks complicated.  be careful00:32
gnarfacei'm kinda surprised gparted doesn't have a feature for this yet00:35
gnarfacemaybe it's just a matter of time00:35
gnarfacehas anyone here actually done this before?  resized an encrypted LVM setup?00:38
systemdletemy terminal sessions start in /, not my home directory.   I don't recall changing that, but somehow that's what is happening.01:00
systemdleteWhat needs to be corrected?01:00
DocScrutinizer05https://sintonen.fi/advisories/scp-client-multiple-vulnerabilities.txt01:00
gnarfacesystemdlete: verify the path to the user's home directory is correct in /etc/passwd first of all01:01
gnarfacesystemdlete: (then make sure the user actually has permission to access it)01:01
systemdletelooks like I need "cd" on the last line of .bashrc01:02
systemdletethat fixes it, but still... I don't know why I'd remove that01:02
systemdletedon't recall doing that01:02
gnarfacemy copy doesn't have that01:02
systemdletehuh.01:02
gnarfacesomething is definitely wrong01:02
systemdleteagree.  This is Ascii, btw01:02
gnarfaceyou're not using su or sudo to launch this user session are you?  you might have simply forgot the "-" flag?01:03
systemdleteno, this is only default, when I open a term window01:03
systemdletelet me try this from a console...01:03
gnarfacehmm. that's weird then.  customize any system-level bash scripts lately?01:03
gnarfacelike /etc/profile or /etc/bash* anything?01:04
DocScrutinizer05lol, .bash_aliases ?01:04
DocScrutinizer05https://sintonen.fi/advisories/scp-client-multiple-vulnerabilities.txt01:04
systemdletefrom console, this does not happen.  It appears to log me into my home directory01:06
DocScrutinizer05>> 1. The attacker controlled server or Man-in-the-Middle(*) attack drops .bash_aliases    file to victim's home directory when the victim performs scp operation from the    server.01:06
systemdletexfce4-terminal01:06
gnarfacesystemdlete: also weird that it would be able to find your ~/.bashrc if it couldn't find your home directory... could it be your window manager doing something at startup to move the user out of $HOME?01:06
systemdleteidk, gnarface.  It's possible.01:06
systemdleteDocScrutinizer05:  Are you talking to me?  I am not using scp01:07
systemdletegnarface:  that's what I was thinking.01:08
DocScrutinizer05systemdlete: not really, I just wondered if I met a coincidence on topic01:08
systemdleteok, Doc01:08
DocScrutinizer05though checking your ~/.bash* files woun't hurt I guess01:09
gnarfaceyea you shouldn't even have a ~/.bash_aliases file present by default afaik01:09
systemdleteon ascii, why does an update include libsystemd0?  Is that a dummy -- I hope so!01:09
DocScrutinizer05gnarface: ack01:10
gnarfacesystemdlete: libsystemd is just the runtime libraries.  they're assumed to be vestigial without systemd itself present and running.01:10
systemdletemaybe we could rename it Libsystemd_shim or the like?01:11
systemdleteor, better yet, how about libsystemd_shit01:11
gnarfacei think that renaming it would defeat the purpose of leaving it there to begin sith.  also there's already something named systemd shim or something like that01:11
systemdlete(a 't' instead of an 'm')01:11
gnarface*to begin with01:12
gnarfacethat's happening to me too01:12
gnarfaceheh01:12
systemdleteit's spreading, gnarface.01:12
systemdletefatkeying, I mean01:12
gnarfacesystemdlete: so, the issue is that just changing package dependencies would require rebuilding that package.  so to remove libsystemd0 as a dependency from every package it's already tied to, that would be actually a lot of work.  so basically they just banned systemd itself, and libsystemd0 is being simply ignored until it proves to be a problem on it's own.01:13
systemdleteanyone here using xfce besides me?01:13
gnarfacenot me, but actually it's very popular01:13
systemdleteso something wrong with MY configuration somewhere01:14
gnarfaceare you running anything weird in the xfce startup/launch whatever thingy?01:14
systemdlete"weird?"  Well, at least not anything I wrote myself, heheheheh01:14
gnarfacehmmm01:14
gnarfacethat sounds ominous...01:15
systemdletebut let me double-check...01:15
gnarfacetechnically this could be caused by changes to ~/.xinitrc or ~/.Xsession too i think01:15
gnarfacebut details on that might be specific to the graphical login manager you've chosen01:15
gnarfacenot everything treats those files the same01:16
gnarfacei don't know enough about elogind/policykit stuff to know for sure it's NOT a bug in one of those, or just a mismatch between the one you installed and whatever your chosen login manager is happy with, but i would assume if the issue was there, you'd be having problems even logging in or starting the gui01:17
systemdletehttp://pasteall.org/pic/9f467dba9d57967fcd4d0ea8a1e04a5101:18
systemdleteoh.  OK, login manager, let me check...01:19
systemdletewell, first off.  I have no .xinitrc or .Xsession01:20
gnarfacethat's fine01:20
gnarfacethey're optional01:21
gnarfacethe resident xfce experts are probably all asleep or working right now01:22
gnarfacehmmm01:22
gnarfacevboxclient01:22
gnarfaceis this a virtualized guest install?  or are you running on bare metal?01:22
systemdleteVM01:22
systemdletewhich file contains the choice of login manager?  I have forgotten again...01:23
systemdleteok01:23
systemdleteit is SLIM01:23
gnarfaceok, might be nothing, but just statistically speaking, the fact this is in a VM is a likely culprit ... purely statistically speaking mind you.  i got no specific suspicions on how this could be related01:23
systemdlete(it's in /etc/X11/default_login_manager)01:23
gnarfaceok01:24
systemdleteI highly doubt it, though I will not rule it out completely)01:24
systemdleteVbox tries not to interfere with the guest environment01:24
systemdleteexcept for handling the kb and mouse, and even there, it really just passes it off to the guest01:25
gnarfacerelease notes (https://files.devuan.org/devuan_ascii/Release_notes.txt) say that slim goes with consolekit, not elogind.  at this point you should check to make sure you have consolekit installed and do not have both installed01:25
systemdleteit does do some support for the video01:25
gnarfacebut again... i thought the symptoms of the slim/policykit/consolekit/whatever mismatch were total failure to start or log in... this seems ... simpler than that01:26
systemdletehuh.  whaddayknow, they are both inhstalled01:26
systemdleteI worry that if I remove elogind, though, maybe I won't get the system back01:26
systemdletebut!01:26
systemdleteAH!01:26
gnarfacewell it might not even be the problem01:26
systemdleteI could checkpoint the VM and roll back if there IS a failure...01:26
gnarfacethat would be a good idea01:26
systemdleteshould I try?01:27
gnarfacei think so, but i know for a fact this has been discussed somewhere on the forums.  for good measure you might want to find that thread first01:27
systemdleteis it worth the effort?01:27
systemdleteok01:27
gnarfacewell01:27
gnarfaceif you try to remove elogind, does it want to remove any other packages too?01:27
gnarfaceif not, i'd say it's safe.01:27
gnarfacethe mess comes with when it tries to gut your system on the way out and you didn't keep notes about what you had installed01:28
gnarfacerolling back to an earlier snapshot of the VM should obviate that problem though01:28
gnarfaceand if you haven't read the release notes, you really should read them, at least the part under "### Session management and policykit backends" that mentions the conflict01:29
gnarfacesystemdlete: oh, i just thought of something else too.  does it happen with every non-root user, or just that one?01:30
systemdleteLooking at them now.  Here is what it says:  "Two of such session management systems are available in Devuan 2.001:30
systemdleteASCII, namely:01:30
systemdlete  - consolekit01:30
systemdlete  - elogind01:30
systemdleteThese session managers are mutually exclusive, only one of them can be01:30
systemdleteinstalled and active at a time to avoid unwanted interference."01:30
systemdlete(sorry, will paste in future)01:30
golinuxsystemdlete: https://dev1galaxy.org/viewtopic.php?id=259701:30
gnarfacethanks golinux01:30
systemdletethanks golinux01:31
gnarfacesystemdlete: it's the "unwanted interference" part i'm not clear on because i don't use a graphical login.  i just run startx every time01:31
golinuxDiscussion was upthread.01:31
systemdleteFrom the link golinux has provided: "option would be to remove the hook in all the packages that call it.  Unless we have many more devs to remove these, we're going to have to live with it.  We actually did a pretty good job in jessie but the list is growing with each new release."01:32
systemdleteThis sounds a lot like the genesis of any ordinary disease or pathogen01:32
systemdletebut we already know all that...01:32
gnarfaceyea :(01:32
golinux:D01:33
systemdleteso I shall now attempt to remove elogind, and I'll see you guys in a few minutes.  I am running hexchat in Ascii.01:33
systemdleteanything else before I "blink out?"01:33
golinuxAlways read the Release Notes.  ;)01:34
gnarfacesystemdlete: if it wants to remove other packages with elogind, keep a list01:34
systemdleteI'll snapshot it and show it to you guys before I reboot01:34
systemdletehow's that sound?01:34
systemdleteok then... here goes01:36
systemdleteif there is a disaster, I'll hexchat from the host01:36
systemdletesnapshot complete... removing elogind now...01:37
gnarfacegood luck01:38
systemdletehttps://pastebin.com/cw4cmx7z (while trying to remove elogind)01:38
systemdletejust elogind -related packages01:38
systemdletethat's safe, right?01:39
gnarfacelooks fine to me.  like i said, haven't tested this myself personally though.01:39
systemdleteso Here I log out.  See you in a few moments.01:39
DocScrutinizer05>><systemdlete> or, better yet, how about libsystemd_shit<<   if anything then libsystemd_dummy01:39
gnarfacei think it'll work01:39
gnarfacehey, you made it back.  that's a good sign01:42
systemdleteI survived the reboot!  Cuz here I am!01:42
systemdletenow, to test the shell in terminal...\01:42
systemdleteoooh.  bad news.  Still logs me into /01:42
gnarfacehmmm01:42
gnarfacedoes it happen with every terminal?01:42
gnarfaceand does it happen even with terminals launched from other terminals with no command-line options?01:43
gnarfacei'm wondering if something weird snuck into the menu shortcut for the terminal you're using01:43
systemdletenow here's something interesting:  If I open a tab from a tab where the cwd is "/", it opens the new tab in "/"01:43
systemdleteBUT01:43
systemdleteif I open a tab and cd first, then open a new tab, it opens it in /home/systemdlete01:43
fsmithredthat's normal01:44
systemdlete("cd first" means I get to my home)01:44
systemdleteright.  But do you see what I'm saying?01:44
fsmithredtry creating a new user and see if it happens.01:44
systemdleteok...01:44
fsmithredwhat desktop?01:44
gnarfaceyea, i second that.  actually i mentioned it up above too.01:44
gnarfacefsmithred: he's using xfce in vbox01:45
fsmithredok01:45
gnarfacethere's no known vbox bugs that do anything like this, are there?01:45
fsmithredI've run into this in xfce, a few years ago.01:45
systemdletenow, logging in from a console gives me my home directory when I log in01:45
fsmithredI think there's a setting for it somewhere01:45
gnarfacehmmm01:45
systemdletefsmithred:  Well, do please share if you can remember!01:46
fsmithredso it works correctly for the new user?01:46
systemdlete:D01:46
systemdleteno, but I can login at a console prompt and get my home without having to cd01:46
systemdleteI haven't tried new user yet01:46
systemdletehold on...01:46
gnarfacefsmithred: (he was mentioning to catch you up that we already know this is only happening in xfce)01:46
gnarfacei know the emacs file loading cwd starts at whatever directory you launched emacs from.  i wonder if xfce is doing something like that?01:47
gnarfacepresumably the user launching slim would be launching it from /01:48
gnarfaceit could actually be considered a feature, no a bug01:48
gnarfacethough that wouldn't explain the mystery of how it got enabled in this case01:48
fsmithredhow is slim launched in this case?01:49
gnarfacei don't use slim.  isn't it an init script?01:49
gnarfacesystemdlete: stock slim install, right?01:49
gnarfacefsmithred: he inferred it was a fairly stock ascii install01:50
systemdleteafaik.  It has been several months since I created this Ascii VM01:50
gnarfaceor, implied?  inferred? i guess i might use those words wrong01:50
fsmithredhe could imply, you could infer01:50
systemdleteHaven't mucked with much except keyboard layout option.01:50
systemdleteuh...01:50
gnarfacethanks.  I inferred it was pretty much a stock ascii install01:50
systemdletey'know what.  Not that this should have much to do with it, but...01:51
gnarfacesomething to do with locales?01:51
systemdleteno.01:51
fsmithredecho $HOME01:51
systemdleteI am using a pretty sophisticated multi-desktop desktop, so to speak01:51
systemdleteI actually have 7 virtual desktops defined in xfce, using its built-in desktop switcher01:52
gnarfacehmm.01:52
gnarfacei wouldn't expect that to be related but i don't know for sure01:52
systemdleteand I sorta customized firefox, thunderbird and a few other launchers to use unique "profiles" (as defined by mozilla, etc) in each of these desktops01:53
systemdleteBut I don't fire those up at startup; I run them manually, as I need them.01:53
gnarfacehmmm01:53
systemdleteRightnow, I just have the terminal and hexchat open.01:53
systemdleteI've run nothing else since I rebooted after removing elogind01:53
systemdleteI doubt (like nearly 100% certainty) that it has anything to do with the login directory issue\01:54
systemdletebut I thought I should tell you this, as you asked if I had done any customization of the xfce environment.01:54
systemdleteClearly, I have01:54
systemdletehexchat, btw, uses the standard launch config.  I did not change that.01:55
systemdleteOMG.01:55
gnarfacehmmm01:55
fsmithredan unexpected user name in hexchat?01:56
* systemdlete looks around to see if there are any safe places to hide from flying debris, rotten tomatoes, etc. as he starts to tell these guys what he did...01:56
gnarfacegoogle gives me an ubuntu bug from 2010 related just to launching gnome-terminal from a keyboard shortcut...01:56
gnarfaceheh, oh boy.  what's the answer to the mystery systemdlete?  we'd all sleep better knowing01:56
systemdleteUhm... I kinda changed my login name when I ported my home directories to ascii from jessie01:56
systemdleteand, uh...01:56
fsmithredlmao01:57
gnarfaceoh! and you forgot to change the directory to match it?01:57
gnarfaceor just the line in /etc/passwd ?01:57
systemdletewhat I forgot to do was update the launcher for xfce terminal.  It has a "Working Directory" setting for my old home directory (old user name)01:57
systemdleteso...01:57
fsmithredif refractainstaller is installed, you could run change-username.sh01:57
systemdletegimme a minute01:57
gnarfaceah.  yea the ubuntu bug mentions setting that.  well, that's a relief01:57
fsmithredand it'll do all the things needed (I hope)01:57
systemdleteah-yay-yuy01:58
systemdletethat fixed it, yep01:58
golinuxsystemdlete: I also have 7 desktops and not a problem but I' still on jessie.01:58
systemdleteand, the worst part is that I wish I hadn't changed it.01:58
gnarfacealright, just a regular snafu01:58
gnarfacehappens to the best of us01:59
systemdleteIt's just hat it makes things easier when reporting bugs; I don't need to worry about obfuscation, etc01:59
systemdleteso happy to hear I am not "the best of us"01:59
systemdleteLOLOLOL01:59
systemdlete(because I am definitely not.  I can type really fast, 61 WPM, but that's it.)01:59
gnarfaceit's just a saying01:59
systemdlete(I know)02:00
nemoheh. 61wpm is fast? ☺02:00
systemdleteubuntu bug?02:00
fsmithredyeah, that's pretty fast02:00
systemdletewell, faster than most, nemo02:00
gnarfacei meant it to sound less derogatory than "even monkeys fall out of trees" but i don't know if i hit the mark02:00
systemdleteI feel very guilty for wasting all your time02:00
systemdleteagain02:00
systemdleteand02:00
systemdleteagain02:00
fsmithredI watched someone type 90 wpm once, and she was obviously in some kind of trance state02:00
systemdleteand...02:00
systemdletewell, weed is legal here in California, so maybe I'll try that trancing sometime02:01
systemdlete(just to see)02:01
fsmithredit was an awesome site02:01
golinuxsystemdlete: You'll sleep well tonight after solving this.02:01
systemdletegnarface:  Which ubuntu bug was that?02:02
fsmithredI'm sure we could fill a book with stories about all the stupid mistakes we've made02:02
systemdleteamazing you could remember that.02:02
systemdleteI could fill a book with just my own.02:02
gnarfacesystemdlete: it gets mentioned here.  apparently something to do specifically just with keyboard shortcuts and gnome.  https://ubuntuforums.org/showthread.php?t=146924002:02
systemdleteNow, why was it I tripped over this problem... I forgot.  Got so caught up in this issue...02:03
systemdleteOh.02:03
gnarfacesystemdlete: oh, i didn't remember it from 2010.  i asked google 4 minutes ago02:03
fsmithredshhh!02:03
systemdleteI think I was trying to update vbox from 5.2.22 to 5.2.24 on my testbox (not this host)02:03
systemdletefunny fsmithred!02:04
systemdleteyeah, keep it a secret02:04
fsmithredI hate updating vbox. I may stop using it for that reason.02:04
systemdleteelse pretty soon EVERYONE will be asking google...02:04
systemdleteI've considered dropping vbox myself.  But I haven't had time to evaluate the alternates.  I just moved about a month ago and I'm still settling in here02:05
nemosystemdlete: https://www.typingtest.com/result.html?acc=100&nwpm=82&gwpm=82&ncpm=410&gcpm=410&dur=60&time=60&chksum=41013&unit=wpm&kh=998&td=null&err=0&hits=410  hm. dropped ab it02:05
fsmithredit's too easy not to use it02:05
nemowas 85 last time02:05
gnarfacesystemdlete: for hardware support quality, i recommend qemu-kvm.  it's definitely not very user friendly though, and libvirt isn't of high quality so i would recommend against pairing the two02:06
gnarfacebut if you only have one type of guest, you only need to figure out the command-line once02:06
gnarfaceso there's that02:07
nemosystemdlete: but apparently they agree with you on "fast" though ☺02:07
rwpI'm not unhappy with libvirt. That isn't a glowing recommendation. But it hasn't sucked for me yet. I am still using it.02:07
systemdletenemo:  85 -- YOUR speed?  Really?02:07
nemosystemdlete: on the astronaut sample02:07
nemosystemdlete: fun one to try is one of the foreign languages.  Good XCompose digraph practice 😃02:08
systemdletenemo:  This typingtest.com is very smart!  It says I type 82WPM and I haven't ever visited it yet!02:08
systemdleteeasy, nemo.  These increasingly arthritic fingers have their challenges...02:09
systemdletetypingtest.com must have googled my wpm02:09
systemdlete(right, gnarface?)02:09
systemdleteOr maybe it picked up on my last pastebin copy-and-paste.02:11
gnarfaceheh02:11
systemdlete(using some cross-scripting exploit in mozilla?)02:11
fsmithredmaybe. I didn't see the score until I allowed js.02:11
systemdleteok, I will "RETAKE" the typing test, the one I never took.02:11
fsmithredI think the 82 was nemo's result02:12
gnarfaceit would be a simple matter in javascript to measure your WPM but it would be just as simple to cheat...02:12
gnarface(session security is surprisingly not as easy, apparently... the 82 could have been nemo's result)02:14
systemdlete61, errors 10, for adjusted 51 wpm.02:14
systemdleteI was nervous.02:14
gnarfacei'm not even gonna click on it02:14
systemdleteI was thinking of bad things.02:14
systemdletegnarface, maybe I can find out where you live, or even hack your cam...02:14
systemdletenemo: That was the space cowboys test02:15
gnarfacei just assume there's ads on it and they're gonna measure a lot more than my typing speed, fuck that02:15
systemdleteI'll try it again in a moment02:15
gnarfacewe're WAY offtopic here... this should really be in #debianfork...02:15
systemdleteYou're right, gnarface.  Really, they should consider your COMBINED typing and clicking speed.  They could test users to see how facile they are texting and purchasing from Amazon simultaneously.02:16
systemdlete(and that should include how fast they can whip out their credit card)02:16
systemdleteThe test should be relevant to the millenial age, not us old farts02:17
gnarfacenemo: i admit i haven't tried libvirt recently02:17
* systemdlete tries hard to remember what he was REALLY trying to do when this other bug came up...02:19
* systemdlete heads to debianfork so as not to disturb serious people in this channel02:20
systemdletenow I remember:  I was trying to figure out what the problem is which my KVM switch.02:22
systemdleteWhen I switch to the testbox, sometimes the mouse quits on me.  But when I return to this host, all is well again.  KB and Mouse work correctly.02:22
systemdletethe host on the testbox is Ascii, btw.02:23
systemdleteI hit capslock-capslock+Fx (F1 for this host, F2 for testbox) to switch.02:23
systemdleteany thoughts on why mouse might quit only on one PC (the one running ascii)?02:24
systemdleteNOt sure if there is a way to change the key sequence for the kvm switch.02:26
gnarfaceany way to recreate the error?  how sure are you it's not the connections?  the KVM would still have separate wires to each computer...02:26
systemdleteright.02:26
gnarfaceany of the jacks a bit loose?02:27
gnarfaceif they're different hardware, it could be other things too...02:27
gnarfacethings that are very unlikely if they are the same hardware02:28
systemdleteThis switch http://byteccusa.com/index.php/4x1-usb-hdmir-kvm-switch-4k2k.html (for reference)02:31
systemdleteOk, so it has 2 USB cables going in, one for each PC.  Also 2 HDMI cables, one for each machine.  But video is fine across the whole kvm02:32
systemdleteit's just the mouse.02:32
gnarfaceone of those usb cables for each pc has gotta be the mouse02:32
systemdleteIt appears the mouse and keyboard are combined for the trip to each PC because there is only one USB cable per PC02:33
gnarfaceoh02:33
gnarfaceis it a wireless mouse?02:33
systemdleteAnd the "common" cables all go to the same devices.  Yes, the mouse is wireless (but the kb is wired USB)02:33
gnarfacei misunderstood, i thought you meant there were two usb cables and an hdmi cable for each pc02:34
gnarfaceon the pc side02:34
systemdleteI doubt it is a problem with the "front end" (the real (kb) and virtual (Mouse) connections) because they work consistently fine with the first PC02:34
systemdlete it has 2 USB cables going in, one for each PC.  Also 2 HDMI cables, one for each machine.02:34
systemdletesorry if That was confusing.02:35
gnarfacewell what i was thinking is that a bios difference or a driver difference could be... i dunno, putting it or the USB port to sleep due to power management02:35
systemdletemaybe i should have said, it has 1 usb and 1 hdmi cable for each PC02:35
gnarfacethat's the only idea i've got02:35
gnarfacesorry02:35
systemdletegnarface:  That is most likely it.  Now, given the target having difficulty is our own beloved OS (ascii), it seems like we should be able to tackle this02:36
gnarfacewell i asked already, are they different hardware?02:36
systemdleteyes.02:36
gnarfacedifferent models02:36
gnarfaceok so what model is the one that is having problems02:36
systemdletePC #1 ("this" box) and PC #2 ("the testbox"), for future consistency02:36
systemdletePC #1 is Asus board.  All USB is connected directly to the back 6 ports, or 2 in front which are connected from the case module to the board.02:37
systemdletebut the USB cable connnected to the KVM is off the back of PC #!02:38
systemdletePC #102:38
systemdleteNow, on PC2 (the ascii machine), the mainboard is MSI 7461.  Again, the USB connection to the switch is from the back USB port02:38
systemdlete(one of 6 on the back)02:39
systemdletethe testbox (PC2) has 2 front USB ports, but I do not use them for the KVM)02:39
systemdleteSo, in sum:  The usb cables are connected DIRECTLY to USB ports on the back of each PC02:39
systemdlete(I hope this clarifies things)02:39
systemdleteI could try a different USB cable, maybe swap them.02:40
systemdleteOr try different USB ports02:40
gnarfaceyou should check the bios settings too02:40
systemdleteAH!  Maybe the unit demands only USB 3.0 ports02:40
systemdleteI wouldn't think so, since it usually works fine, but who knows.02:41
gnarfacewell does it quit while you're using it?02:41
gnarfaceor after it has been sitting idle for a bit?02:41
gnarfacedoes it quit while the mouse is actually moving or clicking?02:41
systemdleteThe mouse quits.  And it is immediately upon completing the switch to the testbox02:41
systemdleteno, prior to that02:42
systemdlete(see above)02:42
gnarfaceoh, so it never works AT ALL on the test box?02:42
systemdletenot quite, no02:42
gnarfaceonly works in the bios?02:42
systemdleteusually, it works oK.  But after some number of switches to the testbox, the mouse fails02:42
gnarfacehmmm.  and what brings it back?02:43
systemdleteI think I can see the mouse pointer, but PC2 does not react to the mouse02:43
systemdletea few things.02:43
systemdlete1) rebooting the Testbox (PC2)02:43
systemdlete2) switching back to the PC1 one or more times02:43
systemdlete3) (maybe, but I'm not sure about this:) unplugging and replugging the usb dongle on the FRONT end... (why that would work, idk)02:44
systemdleteprobably I should try the 3rd trick unplugging the USB on the back of the testbox.  Let me try that now (the mouse is not working on the testbox now)02:44
fsmithredcheck for any free-play in the plugs02:45
fsmithredsometimes they can move a little bit and lose connection02:45
systemdleteOK, here is what's interesting.  After unplugging and replugging to an adjacent port on the back of PC2/testbox, when I move the mouse, I still don't see the mouse pointer, but I do notice that the images on the screen do react a little bit more02:47
gnarface?02:48
systemdletee.g., I have firefox open, and it wants to restore my session, but I haven't clicked to do so yet.  When I move the mouse around I can see the mouse focus passing over the list of tabs to be restored (they flicker as each tab is passed over)02:48
fsmithredclick on something02:49
gnarfaceyea, does it react to clicks?02:49
systemdleteno02:50
systemdletebut02:50
systemdleteget this: When I wtich the usb port to different "set" on the back, the mouse works again02:50
gnarfacehmm02:50
systemdlete(but I only have this success once so far, so not sure it "fixes" it)02:50
fsmithredtail -f /var/log/messages02:50
systemdletewhat I mean by "set" is that the USB ports come in pairs02:50
fsmithredand then wiggle your wires02:51
fsmithredbrb02:51
gnarfacethe cases may have usb cables inside them too... going from the back ports to the motherboard02:51
systemdletewhen I said I switched to an "adjacent" port, I meant in the same pair of ports02:51
systemdleteeh, no.02:51
systemdletethere is a front panel for the case, and those USB wires do connect to the mainboard.02:52
gnarfaceor the front02:52
systemdletebut they are a specific interface02:52
systemdleteON the board, not from the back02:52
systemdlete(not sure if I make sense)02:52
gnarfaceyes, that is what i'm talking about02:53
gnarfacethat will also be a separate USB hub to the linux kernel02:53
systemdleteI am NOT using the front ports02:53
systemdlete(right)02:53
gnarfacethis could still either be a software or hardware issue as far as i can tell, but said you moved it to the 3rd seat and it started working again, so i'd say leave it there for a while and see if it still acts up eventually.02:53
systemdleteJust fyi, I am NOT using front ports for the switch on either PC.  They are both from a rear port on each.02:54
systemdleteyeah.02:54
systemdletewill do.02:54
systemdleteI'll keep switching back and forth between the two PCs and try to "force" a mouse fail again.02:54
gnarfaceeven on the ports that are directly soldered to the motherboard, if you have more than 2, you are likely looking at actually more than 1 separate USB hub devices too02:55
gnarfaceif that vendor chose to use a different model for that particular usb bank, it might have different drivers or different power management characteristics02:55
gnarfaceso it could still be a software issue02:55
systemdleteI just switched again, and now the mouse is stuck again.02:55
systemdleteSo the different set of ports didn't really matter.02:55
systemdletesame problem02:56
systemdletebut it could be the physical cable.02:56
gnarfacehmm.  then if you... move them back to the 2nd seat again, it starts working, again?02:56
systemdleteI'll try another USB cable (I've got tons of USB A to USB B port cables02:56
systemdlete2nd seat?  you mean 2nd set?02:56
gnarfacethe slot on the switch the test box was formerly in02:57
gnarfacebefore you moved it02:57
systemdleteFirst, I'll try moving it back to another port on the back...02:57
systemdletebb one sec02:57
gnarfaceand how long are these USB cables here?02:57
systemdleteheh.  That worked, switching to a different USB port.02:58
systemdletemouse works again on PC2.02:58
systemdletebut it might not when I switch again... wait.\02:58
systemdletenope.  But after I swwitched it again, the mouse works agian (I hear the KVM chirp after moving to different port)02:59
systemdletethese are pretty standard 5' or maybe 6' cables02:59
systemdleteI have tons of them, mainly from printers over the years...03:00
systemdletepretty funky03:00
systemdleteLet me look at the messages log, as per fsmithred's suggestion...03:00
gnarfacepeople having this trouble in windows on other hardware have been suggested to avoid the usb3 ports03:02
gnarfaceadvised i mean03:02
systemdletehmm03:03
systemdleteI hear that there have been problems with 3.0 vs 3.1 USB ports?03:03
systemdleteI believe mine are all 3.0 or 2.0; I don't think any are 3.103:03
gnarfacei'm not sure but i know that could ultimately mean some sort of driver issue.  the usb3 ports would be using a different driver03:03
systemdleteOK...03:04
systemdleteLet me look at the board's manual...03:04
systemdleteLooks like it's USB 2.0 for all ports, unless I am mis-reading this.  http://www.manualsdir.com/manuals/467058/msi-760gm-p34-fx.html?page=14&original=103:08
systemdletehttp://www.manualsdir.com/manuals/467058/msi-760gm-p34-fx.html?page=11&original=103:09
gnarfacelast bios update 2015-04-23, version "H.H"03:11
gnarfacegot that one?03:11
gnarfaceversion H.E, released 2013-03-25... "Improved gaming mouse compatibility." ....03:12
systemdleteI'm pretty sure it is uptodate, but I'll double check.  Hold on.03:19
systemdlete17.1503:20
gnarfacei dunno, their website doesn't list that version03:22
gnarfacehttps://www.msi.com/Motherboard/support/760GM-P34-FX03:22
systemdleteI know.  The board says MSI 7641 on it, but googling it brings us to msi's site for a whole family of boards.03:26
systemdleteI think this is 760GM-P23, but I am not certain.03:26
gnarfaceit should be silkscreened ON the motherboard somewhere03:27
gnarfacebut you might be able to find it in the output of dmidecode too03:27
gnarfacethe 760GM-P23 lists the same fixes03:28
gnarfacesame versions too03:29
systemdleteYep, it is that model, 760GM-P23, because I have the box it came in right here.03:29
systemdletedmidecode!03:30
gnarfacehttps://www.msi.com/Motherboard/support/760GM-P23-FX03:30
gnarfacei don't know if it's the literal same file on their servers but it's got the same updates on the same days03:31
gnarfaceat least the last 403:31
systemdleteThe "silk screen"'d id is 764103:32
systemdlete(not helpful, yes)03:32
gnarfaceactually it is03:33
gnarfacethat checks out03:33
gnarface7641 is in the bios zip file03:33
gnarfacein the file name03:33
gnarfacefor both of them.  looks like it is the same file03:33
systemdleteLooks like the "same" board model number is used for a number of products, including mine.03:37
systemdletebut why does the BIOS UI say it is v17.15?  That's not even a thing on their site... (I'm looking at forums now)03:37
systemdlete(MSI forums)03:38
systemdletegnarface: I'm sending MSI support a question.  Let them sort it out.  I don't want to risk bricking the board.  This is only a testbox, but I like the idea of continuing use of the board.  I willfollow their directions, if an update is available, and if it bricks, itwill be on THEM.03:42
gnarfacesystemdlete: sorry, got distracted there.  yea, that sounds like a good plan.  i don't have any idea why they do it that way.03:44
gnarfaceseems annoying.  i wouldn't worry about the 7641 thing03:44
gnarfacethat's common03:44
gnarfaceusually they have several different models based on the same exact motherboard, just with different combinations of I/O ports soldered on03:45
systemdleteand graphics cards, in this case03:45
systemdletethe "760GM" is the graphics I believe.03:46
systemdleteBut, yeah.03:46
systemdleteHOWEVER03:46
systemdleteWhere does this BIOS v17.15 come from?  It is not even available on their website.03:46
systemdleteUsually, the mfr keeps old BIOS versions around03:46
systemdleteThe numbering scheme is completely different from the (apparent) scheme my BIOS currently has.03:47
systemdleteI still think a message to their support is in order.03:47
gnarfaceyes, that seems like a valid question03:47
systemdleteOK, filled out their form, which is almost unusable.  They must have had Microsoft engineers design their site.03:59
systemdleteEverything is Windows-centric, even though even though the briefest of scans through their forums show that there are gamers, Linux users.04:00
systemdleteWhen oh when will these companies wake up to reality?04:00
systemdleteIt's annoying being required to fill in a form, and when I get to the drop-down for type of issue, it shoots me over to another page where I have to fill in the same info all over again.  This time, I can enter phone number free style, rather than the first page's strict input form for phone.04:01
systemdleteThey claim they get back in 24-48 hours.  Maybe they mean days, or years.   In the meantime, I will keep trying to manipulate the hardware a bit, trying different cables, ports, etc.04:04
gnarfacei wonder if you can wake the mouse back up from software alone, by running some hotplug event script...04:06
gnarfaceyou should also probably be tailing some log file to see if it spits out any relevant info at the point when the mouse dies04:07
golinuxWhat a marathon!  You're still at it!!04:08
Criggiew04:15
Criggiefree04:15
systemdletegolinux:  Intermittent problems like these often involve lengthy investigation and debugging.   I don't recall a time when this was not the case.04:26
golinuxVery true.04:30
systemdletethank you for your understanding05:09
_abc_Hi. I asked yesterday if the link (1) is somewhat relevant to refracta2usb generated boot environments on media. What is the real reason for using isolinux on a fat32 and not grub on an ext2 or other partition for refracta2usb? https://www.pendrivelinux.com/boot-multiple-iso-from-usb-via-grub2-using-linux/08:25
_abc_Is it FAT32 just for convenience for windows users?08:27
gnarface_abc_: no, there's some technical reason.08:29
_abc_Well using isolinux for boot manager is one reason.08:30
_abc_I don't remember if it uses isolinux or grub.08:30
gnarfaceit's got something to do with fat32 being the only filesystem supported everywhere without drivers08:30
_abc_Ah. Everywhere meaning?08:30
gnarfaceat a bios level08:30
_abc_Ok, that makes sense. But the actual boot loader would be grub, which does support raw disks and ext2 iirc.08:31
_abc_And vfat and other things too.08:31
gnarfacei don't know if grub has a problem with hybrid iso images or what the deal is08:31
gnarfacemabye it's just personal preference of the developers08:31
_abc_Well I hope someone does know.08:31
_abc_(ping fsmithred :)08:32
gnarfacesomeone does.  i remember the discussion i just don't remember the details08:32
_abc_I'll wait.08:32
gnarfaceyou can still boot from USB with grub but then you need a bios that can look for a MBR on a USB key08:33
gnarfacewhereas a hybrid iso image is made to boot like an optical disk08:33
gnarfaceor something like that08:33
_abc_I see what you mean, I think.08:34
_abc_And I think I meant syslinux not isolinux but it's the same08:34
gnarfaceit could come down to just legacy software too.  grub is the newcomer to this scene08:35
_abc_grub2 is the newcomer08:35
gnarfacesemantics08:35
gnarfacei think they did put some convenience info in the fat32 partition for windows users though08:36
gnarfaceso it shows up as something that has a readme08:36
gnarfaceor some sort of identifying information at least08:37
gnarfacethere might even be a launch option08:37
_abc_this has some tangent explanations https://www.pendrivelinux.com/yumi-multiboot-usb-creator/08:37
fsmithred_abc_, I don't know the reason for using syslinux instead of grub. Maybe because it's easier to deal with.16:54
fsmithredI have made live-usb with grub2 bootloader. It works fine, but have to do it manually.16:54
_abc_fsmithred: ok, thanks for answering17:55
gnarfacesystemdlete: go into the bios on that second machine please and see if there's an option for something about USB Legacy support20:58
gnarfacejust another possible fix that came to mind20:59
gnarfacedifferent hardware might have different defaults for whether that's on or off20:59
gnarfaceand it affects bios-level USB device support20:59

Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!