| sanitar | i wasn't able to get the live iso to boot when it reached "unmounting filesystem". | 02:58 |
|---|---|---|
| gnarface | sanitar: what did you use for boot media? sometimes it can working but just unexpectedly slowly... | 03:00 |
| gnarface | something with a traffic indicator on it helps | 03:00 |
| sanitar | i booted it from disk | 03:00 |
| gnarface | hmm | 03:01 |
| gnarface | which iso exactly? | 03:01 |
| gnarface | show the link you used | 03:01 |
| sanitar | desktop-iso, daedalus | 03:01 |
| sanitar | gnarface: wait | 03:01 |
| gnarface | this one? https://files.devuan.org/devuan_daedalus/desktop-live/devuan_daedalus_5.0.0_amd64_desktop-live.iso | 03:01 |
| sanitar | here: https://mirrors.ocf.berkeley.edu/devuan-cd/devuan_daedalus/ | 03:02 |
| gnarface | maybe verify the checksum | 03:02 |
| gnarface | ... | 03:03 |
| gnarface | sigh | 03:03 |
| fluffywolf | minor bug report... the reference [numbers] in README.txt don't match up with the links under them. | 05:26 |
| fluffywolf | [5] and later are off by one | 05:26 |
| fluffywolf | also, release notes looks like it was meant to be input for some preprocessor that was never ran? | 05:34 |
| rrq | fluffywolf: ta | 06:22 |
| * fluffywolf is now distracted LOLing at the usrmerge discussion | 06:37 | |
| fluffywolf | bbl, wolfy bedtime | 06:54 |
| peterrooney | fluffywolf: it's absolutely amazing the corner usrmerge got painted into | 06:58 |
| Igor2 | hi; you are probably aware of this, but there seems to be a problem with the usrmerge package | 10:38 |
| Igor2 | I had a small Chimaera install which I dist-upgraded to Daedalus | 10:38 |
| Igor2 | right after that I tried to dist-upgrade it to Excalibur | 10:39 |
| Igor2 | that process failed because I didn't have usrmerge installed before the Daedalus->Excalibur dist-upgrade | 10:39 |
| Igor2 | so during the dist-upgrade to Excalibur, some packages depended on usrmerge already being installed, so the process failed, but it left the system in a state that I couldn't install usrmrerge anymore | 10:41 |
| Igor2 | I've restarted the whole thing and now manually installed usrmerge before the dist-upgrade to Excalibur step and it's going better now (will report when finished) | 10:42 |
| Igor2 | as a user I'd expect dist-upgrade to always work to upgrade a clean, up-to-date system to the very next version | 10:43 |
| CueXXIII | yeah, that is a bug from debian upstream, you have to install usrmerge first if you upgrade from daedalus or debian bookworm. but that should be a task for debian to figure out | 10:43 |
| Igor2 | so they are probably aware of this | 10:44 |
| Igor2 | (I can't find the related debian bugreport, tho) | 10:47 |
| Igor2 | https://lists.debian.org/debian-doc/2023/06/msg00060.html | 10:56 |
| Igor2 | this is the only relevant thing I can find ant it claims dist-upgrade automatically "pulls it in" | 10:56 |
| CueXXIII | did you have held back packages during the dist-upgrade? | 10:58 |
| Igor2 | sorry, I don't know what a held back package is | 10:59 |
| CueXXIII | packages which won't be upgraded; apt should list them as "the following packages are held back" or such | 10:59 |
| Igor2 | the system was clean in the sense I didn't have any manual per package trickery, pinning, etc | 10:59 |
| Igor2 | I dn't have the logs, but I have a snapshot of the Daedalus state so I could later repeate the same dist-upgrade | 11:00 |
| CueXXIII | i'm thinking of a way to reproduce your problem… which init system are you running? | 11:00 |
| Igor2 | sysvinit | 11:00 |
| Igor2 | if you are interested I could save any state you need from that Daedalus starting point, then save full logs of the upgrade and have the broken system at hand | 11:01 |
| Igor2 | I just need to finish the now-working upgrade first | 11:01 |
| CueXXIII | yeah, the output of that dist-upgrade would be interesting | 11:01 |
| Igor2 | thanks, give me some time to finish the upgrade then I can boot up the to-be-broken system again | 11:02 |
| CueXXIII | and probably the output of dpkg -l before the upgrade | 11:03 |
| Igor2 | ok, I can do that | 11:03 |
| Igor2 | any olther states I should dump before and after? | 11:03 |
| Igor2 | s/olther/other/ | 11:03 |
| CueXXIII | non i can think of now… | 11:10 |
| CueXXIII | none | 11:10 |
| Igor2 | ok, doing the upgrade that should break, logging on | 11:18 |
| Igor2 | (will take some time, my disk IO is rather slow for unknown reasons) | 11:19 |
| Igor2 | it's base-files that breaks | 11:20 |
| CueXXIII | yeah, that should tell you to install usrmerge first | 11:21 |
| CueXXIII | but apt-get install usrmerge won't work anymore at that point? | 11:21 |
| Xenguy_ | Igor2, I am late to this conversation, but: https://www.devuan.org/os/announce/excalibur-usrmerge-announce-2024-02-20.html | 11:22 |
| Igor2 | hmm, it worked now | 11:23 |
| Igor2 | I don't know why it failed in my previous attempt | 11:23 |
| Igor2 | (and it was the second time I ran into this) | 11:23 |
| CueXXIII | Xenguy_: people are going to simply use dist-upgrade without reading that… | 11:23 |
| djph | people read? | 11:23 |
| CueXXIII | Igor2: thanks for trying | 11:24 |
| Xenguy_ | CueXXIII, Okay, how can things be improved? | 11:24 |
| Igor2 | (I can confirm: I started using Debian in 1999 and I simply always trusted dist-upgrade to do the right thing and never read announcements) | 11:24 |
| Igor2 | I'd expect dist-upgrade to install usrmerge before base-files, if that's needed | 11:25 |
| CueXXIII | Igor2: you would have to change apt to do that, and that change would make it only into excalibur, aka too late | 11:26 |
| Igor2 | CueXXIII, now the only thing I can think of is that there could be something undeterministic in the process, like different order of package installation on equal level dependencies, and maybe the problem is triggered with one legit ordering and not with the other | 11:27 |
| Igor2 | but I most probalby won't be able to reproduce it (I won't sit down ant try to run this upgrade a few times) | 11:27 |
| CueXXIII | Xenguy_: not sure how debian did this for bookworm, maybe pre-depend on usrmerge in base-files and remove that only after excalibur+1 | 11:28 |
| Igor2 | hmm, I thought some pre-depend in base-files to usrmerge could solve that without special casing in apt | 11:28 |
| Igor2 | but of course I don't have the full picture on thos usrmerge | 11:28 |
| Igor2 | to me, a plain user, this whole usrmerge is just yet another upgrade/change I never needed or wanted | 11:29 |
| Xenguy_ | CueXXIII, The whole usrmerge thing is a clusterfuck honestly, and then we're left trying to find a way to 'do it right' | 11:30 |
| Xenguy_ | And here we are | 11:30 |
| Igor2 | hehe | 11:30 |
| Igor2 | also, if that pre-depend thing wouldn't work, and it indeed can only be hacked through a special case in apt, I'd at least give it time | 11:31 |
| Igor2 | like first do the apt modifications, wait a release, then do the usrmerge thing | 11:31 |
| CueXXIII | Igor2: that would mean forking hundreds of debian packages for one release | 11:33 |
| Igor2 | yeah, I didn't mean in Devuan, but in Debian, obviously hypotethically | 11:33 |
| CueXXIII | just for undoing the usrmerge commitments that debian does now | 11:33 |
| Igor2 | (meanwhile I let the upgade of this test finish in the background, just to make sure this time it really doesn't fail) | 11:34 |
| CueXXIII | yeah, debian did the usrmerge in bookworm | 11:34 |
| Igor2 | btw, even when it doesn't fail like in my 2 past cases, the fact that dist-upgrade stops and requires manual intervention... I think that's suboptimal, such things somewhat degrades my trust in the system | 11:35 |
| Igor2 | I guess there must be some real stron reasons to do this usrmerge thing if debian accepted this manual workaround setup | 11:35 |
| Igor2 | (assuming the same message is printed on dist-upgrade in debian) | 11:36 |
| djph | Igor2: from what I'm reading, they're about as strong as "use systemd" | 11:36 |
| Igor2 | hehe | 11:36 |
| Xenguy_ | Yeah, they jumped the shark and went for systemd, and the rest is history | 11:36 |
| Xenguy_ | That's when we knew Debian was not the same Debian we knew and trusted | 11:37 |
| Xenguy_ | Hence the fork | 11:38 |
| djph | I mean, I'm probably just a curmudgeon, and seeign it as "change for the sake of change" | 11:40 |
| Xenguy_ | More like 'pain for the sake of pain' | 11:41 |
| Xenguy_ | I mean why would anyone decide to do this? | 11:41 |
| djph | Xenguy_: "change for the sake of change" | 11:45 |
| djph | also kids | 11:45 |
| djph | also pottering. | 11:45 |
| djph | (I have no idea) | 11:45 |
| djph | only thing I do know is that I'm not installing Excalibur til I get a new SSD, and therefore having to usrmerge isn't going to affect my upgrade path too much | 11:46 |
| Igor2 | now that I am reading after it: also freedesktop.... | 11:46 |
| djph | (granted, I'll probably also do installs to new VMs as well, and migrate) | 11:46 |
| Xenguy_ | djph, A 'demarcation point' is probably a wise idea in this case... | 11:47 |
| djph | seems so | 11:47 |
| djph | They've all been pulled forward since Jessie/Ascii, so ... probably about time anyway :D | 11:48 |
| Xenguy_ | I've survived thus far by sticking with the 'trailing edge' (currently running Chimaera here) | 11:48 |
| Igor2 | https://www.freedesktop.org/wiki/Software/systemd/TheCaseForTheUsrMerge/ | 11:49 |
| Igor2 | this is the closest I got to the reasons | 11:49 |
| Xenguy_ | djph, That was always Debian's superpower, for me: being able to upgrade in place online | 11:50 |
| Igor2 | oracle solaris.... hehehehe | 11:50 |
| djph | Xenguy_: yeah, most of my servers are chimaera, with one or two less important ones being daedalus (as in, "if they die, oh well") | 11:50 |
| Igor2 | Xenguy_, I agree, and that's what I gradually lost my trust in, a few years ago, after Devuan Beowulf :/ | 11:51 |
| Xenguy_ | djph, Seems prudent nowadays, if you ask me | 11:51 |
| djph | I mean, that was always fairly prudent | 11:51 |
| CueXXIII | been reading https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=994388 a bit and the route taken in debian is no longer open for devuan… | 11:52 |
| Xenguy_ | Igor2, Caution is needed now I suppose. The system is so good they're having trouble breaking it = ) | 11:52 |
| Igor2 | so freedesktop lists a few reasons, all about being compatible with things... that are either largely irrelevant (like orace solaris) or got totally upside down (like gnu autotools: I thought that whole autotool monster was about figuring your system and is not a system in itself you try to bend your OS to) | 11:52 |
| CueXXIII | maybe a few pre-depends: on usrmerge could help that the dist-upgrade works, but i won't guarantee it | 11:52 |
| Xenguy_ | No wonder I'm up at odd hours, the revelations keep coming | 11:52 |
| CueXXIII | people upgrading to excalibur are required to read release notes :( | 11:53 |
| Xenguy_ | CueXXIII, Yes... | 11:54 |
| Xenguy_ | That is always good policy, but I do feel that somehow the 'news' should be made more prominent | 11:55 |
| Xenguy_ | Suggestions for improvement are most welcome | 11:55 |
| Igor2 | Xenguy_, my weak an uninformed and ignorant sugegstion is still a pre-depends usrmerge in base-files | 11:56 |
| CueXXIII | and maybe put a "tl;dr: install usrmerge first with the neccessary precations in {section usrmerge}" at the top of the release notes | 11:56 |
| Igor2 | (base-files and whatever other packages that have pre-install script or other mechanism that detect missing usrmerge and bail out) | 11:57 |
| Xenguy_ | CueXXIII, Good idea (Excalibur has yet to be released to stable of course, so this might well be the way to go) | 11:57 |
| Igor2 | meanwhile my test upgrade succeeded, so this is really just unreproducible; thank you all for the support | 11:58 |
| Xenguy_ | Igor2, Above my pay-grade, but hopefully the wizards here are watching and taking note | 11:58 |
| Igor2 | hmm, another though-experiment | 11:58 |
| Igor2 | let's say I stop trusting Debian/Devuan dist-upgrades | 11:58 |
| Igor2 | what if I do this: I keep a list of packages I manually installed (I believe this can be extracted using apt somehow) | 11:59 |
| Igor2 | instead of dist-upgade I just install a system anew and apt-get install the said packages | 11:59 |
| CueXXIII | then you would lose all your previous configs | 12:00 |
| Igor2 | hmm, true | 12:00 |
| Igor2 | plus if there are changes in package names or packages substituted with alternatives this also breaks | 12:00 |
| Xenguy_ | Once Excalibur has been released to Stable, the Release Notes will be in full view. It's only until then that folks will get tripped up by UsrMerge, amiwrite? | 12:00 |
| Igor2 | Xenguy_, I never cared about official releases and honestly, didn't read Release notes from when I had broadband internet | 12:01 |
| Xenguy_ | It's kind of required, to be fair | 12:01 |
| Igor2 | ignorant users like me, doing this in a rolling upgrade manner just take these things granted :( | 12:01 |
| CueXXIII | Xenguy_: i didn't sign any contract with devuan requiring me to read anything | 12:02 |
| Xenguy_ | Well, welcome to the New Improved Debian /s | 12:02 |
| Igor2 | well, I am not trying to convince you about the opposite, but to me it doesn't feel natural | 12:02 |
| Xenguy_ | CueXXIII, Fair, but that's on you, buyer beware, and all that rot | 12:02 |
| Igor2 | I mean my "user experience" for the past 20+ years is that for me there are no special release events | 12:02 |
| Xenguy_ | Igor2, No argument here, there's nothing natural about it at all | 12:03 |
| Igor2 | I mean Debian has specific releases, but it's just a name in apt sources file and the annoyance that I have to use dist-upgrade instead of upgrade | 12:03 |
| CueXXIII | it would be soo nice if everyone wuld read release notes | 12:03 |
| Igor2 | yeah | 12:03 |
| Igor2 | I am an upstream maintaner/developer of an EDA package | 12:04 |
| Igor2 | it has 2 large and a small GUI application | 12:04 |
| Xenguy_ | CueXXIII, Honestly I always have and continue to do so, but I'm naturally cautious in my ways | 12:04 |
| Igor2 | I spend hours writing a reasonable changelog and release notes for each and every release | 12:04 |
| Igor2 | but I know very very very few people read it | 12:04 |
| CueXXIII | but live just does not want to be nice | 12:04 |
| Igor2 | and that mostly happens only when something already went wrong | 12:05 |
| Igor2 | hmm, wild idea: | 12:06 |
| Igor2 | what if apt-get dist-upgrade gets a feature so that it can print important messages if the state of the system matches some certain pattern | 12:06 |
| Xenguy_ | Sounds like a brainwave | 12:07 |
| Igor2 | so besides the release notes the messages file would also get a specific warning about this one thing of having to install usrmerge | 12:07 |
| Xenguy_ | One wonders why it hasn't happened already | 12:07 |
| CueXXIII | there is apt-changelogs, that does this, as plugin package for apt | 12:07 |
| Igor2 | and this triggers only when dist-upgrading and only when the affected package(s) would break or something | 12:07 |
| Xenguy_ | But again, above my paygrade, alas | 12:07 |
| djph | Xenguy_: which 'it' ? | 12:07 |
| Xenguy_ | djph, Why didn't a warning get generated in use cases where usrmerge is about to break a system? | 12:08 |
| Xenguy_ | Surely that is doable | 12:08 |
| djph | define "break" ? | 12:08 |
| CueXXIII | if you put something like that into vanilla apt that has to happen one release befor it should display such breakages | 12:08 |
| Igor2 | well, in fact the warning was generated, but it was a tiny bit later than I'd expect | 12:09 |
| Xenguy_ | CueXXIII, Yeah, we're in the grey zone right now, if you will | 12:09 |
| Igor2 | it was generated when dist-upgrade was already running, downloaded tons of packages and started to install base-files | 12:09 |
| Igor2 | (whereas the alternative would be to throw a warning and maybe pause for user input in interactive case before doing anything) | 12:09 |
| Xenguy_ | djph, People's systems are getting hosed cos they aren't running usrmerge before they upgrade to Excalibur (I think that's what we're talking about) | 12:10 |
| Igor2 | s/getting hosed/may get hosed/ | 12:11 |
| Igor2 | (I had it twice, but failed to reproduce it) | 12:11 |
| djph | Xenguy_: ah, TBH I'm not fully up to speed on what "could" go wrong specifically in excalibur that isn't a "could go wrong with usrmerge in general" | 12:11 |
| Xenguy_ | djph, All I know is that usrmerge needs to be run *before* the upgrade | 12:12 |
| Xenguy_ | I've avoided the pain by not going there yet | 12:12 |
| CueXXIII | djph: packages in trixie are now moving ther files into a usr merged layout without having pre-dependencies on usrmerge. when these get upgraded before base-files fails, the system might be in a bad state | 12:13 |
| Xenguy_ | For all I know I may be running BSD before I even get there; it's that awful to me | 12:13 |
| djph | .. that is certainly going to be very bad. | 12:13 |
| Igor2 | btw, I am migrating away from Devuan too, partly | 12:13 |
| Igor2 | Devuan did a great job in undoing some of the damage done by Debian lately (e.g. systemd) | 12:14 |
| Xenguy_ | All hail Devuan | 12:14 |
| Igor2 | but there are annoyances Devuan won't be able to address | 12:14 |
| Xenguy_ | We cross bridges when we come to them | 12:15 |
| Igor2 | (so I can relate to that BSD comment) | 12:15 |
| djph | wait | 12:15 |
| djph | we're supposed to *cross* them? | 12:15 |
| Igor2 | hehe | 12:15 |
| * djph puts away the kerosene and matches ... | 12:16 | |
| Xenguy_ | I know, I can't remember life before Debian anymore | 12:16 |
| Xenguy_ | hah | 12:16 |
| Igor2 | I can: first *NIX system I got access to was a SunOS running on some SPARC | 12:17 |
| Xenguy_ | I even had a Debian coffee mug, and just recently it 'broke' | 12:17 |
| Igor2 | (back before solaris) | 12:17 |
| Xenguy_ | Here comes the Sun | 12:17 |
| CueXXIII | didn't solaris start with the usrmerge? | 12:17 |
| Igor2 | I think so, yes, the above freedesktop reasoning says that too | 12:18 |
| Igor2 | nto sure if it was the first UNIX to do that, but it was certainly the most popular one to do it | 12:18 |
| Xenguy_ | Really, I had no idea | 12:19 |
| Igor2 | I am a fan of old UNIX systems, I have a running Irix on SGI hardware and a running old solaris on SPARC hardware, but I really find it funny to cite solaris compatibility as a reason for *anything* in 2024 | 12:19 |
| djph | heh | 12:21 |
| djph | isn't "solaris" that oracle-cloud crapware too now? | 12:21 |
| Igor2 | I guess so, freedesktop calls it oracle solaris | 12:21 |
| Xenguy_ | ugh, sounds awful | 12:22 |
| Igor2 | would be funny to rename the fs root from / to c:\, for compatibility with windows | 12:23 |
| Xenguy_ | Why did Oracle have to usurp such a cool name, and then turn out so bad? | 12:23 |
| Igor2 | ("you just need to install this kernel patch before upgrade", lol) | 12:23 |
| * Xenguy_ turns and runs for the hills... | 12:24 | |
| * rrq join #devuan-offtopic | 12:24 | |
| Xenguy_ | ah, yes | 12:24 |
| opv | Hi all. I am trying to set up a VNC server, but neither Tiger, nor Tight, nor X11 have an initscript. I've checked out the orphan-sysvinit-scripts package but it just contains a handful of scripts? I must admit to be a tiny bit lost here. Do I need to fire up an initscript-generator? Thank you. | 15:05 |
| Hurgotron | I often just put stuff into /etc/rc.local instead of writing init scripts. May be sufficient here, too | 15:10 |
| CueXXIII | xservers also have no initscript, and vnc server is just an x server that outputs to the vnc protocol | 15:11 |
| opv | fair point | 15:11 |
| CueXXIII | and you could start the vnc server via your x login manager to get a login prompt | 15:12 |
| CueXXIII | otherwise you need to give it a session, too, so there is no easy way to write an initscript that covers all usages | 15:12 |
| opv | yeah, also a good point | 15:18 |
| fsmithred | opv, I log in through ssh and start x11vnc on the remote then start client on local machine. | 15:36 |
| fsmithred | I have it scripted and tied to a desktop icon so I'm just a double-click away from a remote desktop session | 15:39 |
| opv | yeah, i just wrote a shell function too | 15:42 |
| opv | thx guys | 15:43 |
| opv | hm, on a completely unrelated matter... any python aficionados in the room? I got a project where I make live music on nothing but a Raspi and free software, and I need some scripting done. I know this is not really appropriate, but I'm taking the shot. | 15:52 |
| AlexLikeRock | hi people | 17:55 |
| AlexLikeRock | how its going today at DEVUAN world ? | 17:55 |
| buZz | functional | 17:57 |
| morenonatural | AlexLikeRock, developing on wayland-daedalus everyday | 17:58 |
| AlexLikeRock | morenonatural, what a head breaker !! hehehehe , htanks for this work guys ! | 18:46 |
| sanitar | gnarface: i checked it | 20:55 |
| sanitar | https://litter.catbox.moe/0wfu42.png | 20:55 |
| sanitar | the checksums okay, but what do i do now? | 20:59 |
| gnarface | sanitar: well, you might have to resort to a different image... i don't really have any other ideas. the live iso should work on most stuff but if you don't have enough RAM it might not, or there might be other bugs. | 21:00 |
| sanitar | gnarface: ok | 21:01 |
| gnarface | what hardware are you actually trying to boot this on? there might be known issues... | 21:01 |
| sanitar | gnarface: an optiplex | 21:01 |
| gnarface | model #? | 21:01 |
| sanitar | 3040 | 21:01 |
| gnarface | how much RAM? | 21:02 |
| AlexLikeRock | to know the RAM; run command: "free -h " | 21:02 |
| AlexLikeRock | at terminal | 21:02 |
| sanitar | AlexLikeRock: thanks | 21:03 |
| gnarface | dell's website says this model came with anywhere from 2GB to 16GB; 16GB would be plenty, 2GB might be barely not enough | 21:03 |
| sanitar | gnarface: it's showing me 11 gigs | 21:03 |
| gnarface | though maybe an older release could fit in 2GB... | 21:03 |
| gnarface | 11 is a weird number but should be enough | 21:03 |
| gnarface | how long did you actually wait at the unpacking phase you said it stalled on? | 21:04 |
| gnarface | maybe just wait longer | 21:04 |
| sanitar | 4 seconds, i think | 21:04 |
| gnarface | heh | 21:04 |
| gnarface | try giving it 10 minutes | 21:04 |
| gnarface | it might be working fine but just booting slower than you expect | 21:04 |
| sanitar | gnarface: no, but then it fallbacks into an error after 4 seconds | 21:05 |
| gnarface | processor might matter here too, i notice this model could have one of 3 different classes of processors, and the second two look pretty weak | 21:05 |
| gnarface | oh, what is the actual error text it outputs? | 21:05 |
| sanitar | gnarface: where can i find and read /boot.log? | 21:06 |
| sanitar | gnarface: i think it's in here somewhere | 21:06 |
| gnarface | i'm not sure the live iso will even log that, but all logs go in /var/log | 21:07 |
| gnarface | i suspect in fact the live iso will not log that, at least for sure not anywhere it'll still exist after a reboot | 21:08 |
| gnarface | you might just be able to pass a couple kernel command-line parameters to the boot prompt to make this work though, there should have been some suggestions for safe mode and compatibility options on the boot screen somewhere, i thought... | 21:08 |
| gnarface | stuff like acpi=off | 21:09 |
| gnarface | noapic, nolapic? | 21:09 |
| gnarface | been a while since i looked at this | 21:09 |
| sanitar | oh wait, i remember that when i cat into boot.log there was a line that said it panicked because it couldn't find the live system | 21:09 |
| gnarface | hmm, that's weird that it would do that. that error usually means you're missing something | 21:10 |
| gnarface | what are you actually trying to do here? just resurrect an old dell? the normal netinstall installer might work better | 21:10 |
| gnarface | find out which cpu you have too, it might be relevant | 21:10 |
| gnarface | cat /proc/cpuinfo |grep model.name | 21:11 |
| sanitar | gnarface: i don't really want the install to connect to the internet | 21:11 |
| gnarface | oh, i see | 21:11 |
| sanitar | gnarface: and also i'm using an intel core i5-6500 | 21:12 |
| gnarface | hmm, that should be 64-bit, so there shouldn't be a problem with that | 21:12 |
| gnarface | you said you were booting the live iso from the harddrive though? maybe the problem is how you copied it over, or how you set up the grub boot menu option for it | 21:13 |
| gnarface | it's not really intended for that type of boot, and it can be made to work but i know there's a trick to it that i can't remember... | 21:13 |
| gnarface | fsmithred might know though | 21:13 |
| gnarface | or someone else around here | 21:14 |
| sanitar | i don't really know how to explain this, but i didn't create a new partition for the iso and used an existing one | 21:14 |
| sanitar | so i think that's whats caused the issue | 21:14 |
| gnarface | yea, kernel panic for not finding the live system would suggest it didn't find it where it expected to | 21:15 |
| gnarface | so maybe something about the disk #'s changing during boot or something like that... i've had that problem with Dell bioses before | 21:15 |
| gnarface | do you have nothing else you can boot it off? the live iso itself has a built-in tool to install to the harddrive if you can just get it booting from optical or usb first | 21:16 |
| sanitar | i didn't have a usb/cd to boot it to the computer | 21:17 |
| sanitar | so i wanted to boot from the computer directly | 21:17 |
| gnarface | https://forums.debian.net/viewtopic.php?t=155900 | 21:19 |
| gnarface | see if this technique will work? | 21:19 |
| sanitar | gnarface: i'll try it | 21:19 |
| junalio | hi! what might be the reason that when trying to install devuan daedalus with the netinstaller image, the apt configuration part always gets stuck at 33%? i tried it 3 times | 23:15 |
| gnarface | junalio: seems weird... maybe a network stall or out of disk space? one of the other virtual consoles might have an error log (try alt+f4) | 23:16 |
| gnarface | and when you say it gets stuck at 33%, have you verified there's no harddrive or network traffic while that's apparently stalled? | 23:17 |
| gnarface | or does it throw an actual error rather than just stall? | 23:17 |
| junalio | the last message in the log from 10 minutes ago is "choose-mirror: DEBUG: command: wget --no-verbose http:/deb.devuan.org/merged/dists/daedalus /Release -O - | Ggrep -E '^(Suite|Codename|Architectures):' | 23:18 |
| gnarface | what happens if you try to bring that URL up in a browser right now? | 23:19 |
| gnarface | i assume you're using the same network connection here that you used to try to install, right? | 23:19 |
| junalio | yes | 23:19 |
| junalio | i can access that | 23:20 |
| gnarface | hmmm | 23:20 |
| gnarface | so, when it gets stuck at 33%, do you see any network traffic lights blinking, either on the computer itself or on the router? | 23:20 |
| gnarface | that's a DNS round-robin, and not all the mirrors in the list are super fast, it's entirely possible it's working but just going so slow it seems stalled | 23:21 |
| AlexLikeRock | wget: unable to resolve host address ‘http’ | 23:21 |
| AlexLikeRock | forget one "/" | 23:21 |
| gnarface | AlexLikeRock: you're alleging it's a bug in the installer? | 23:21 |
| AlexLikeRock | http:/deb.devuan ->>>> http://deb.devuan | 23:22 |
| gnarface | oh wow | 23:22 |
| gnarface | wait, this is the same installer everyone has been using though, or did a buggy one just go up? | 23:22 |
| gnarface | i totally missed that | 23:22 |
| junalio | i just forgot to type the second /, sorry... i had to copy the alt f4 log by hand into this chat | 23:23 |
| gnarface | oh, whew | 23:23 |
| junalio | sorry for the confusion :) | 23:23 |
| gnarface | is this the exact iso you used? https://files.devuan.org/devuan_daedalus/installer-iso/devuan_daedalus_5.0.1_amd64_netinstall.iso | 23:23 |
| junalio | so, i checked the activity on my router, there is hardly any activity right now, the few KBs are probably from irc ;) | 23:23 |
| junalio | yes, that one | 23:24 |
| junalio | 5.01 | 23:24 |
| gnarface | did you auto-partition or manual? | 23:24 |
| junalio | auto | 23:25 |
| gnarface | the auto-partition logic isn't very smart, and one thing that comes to mind is maybe it didn't leave you enough space in /var/ | 23:25 |
| gnarface | or something like that | 23:26 |
| junalio | hmm, the thing is i did the same process on the same hdd a few months ago | 23:26 |
| junalio | with the same (or previous) devuan netinstaller image | 23:26 |
| gnarface | hmm, well you should be able to get a shell open on an unused virtual console to check to be sure | 23:26 |
| gnarface | without any real error message i can only speculate here | 23:27 |
| gnarface | these installers are usually very reliable, but the mirrors less so | 23:28 |
| junalio | df says use of all devices is below 100 except for /cdrom | 23:30 |
| gnarface | well below 100%, or are any of them above 90%? | 23:30 |
| gnarface | keep in mind ext4 reserves 10% by default | 23:30 |
| junalio | ah and in the alt+f4 log it says that the cd-rom mount point is /media/cdrom. should it be 100%? | 23:30 |
| junalio | all are at 0% except /boot which is at 9% | 23:31 |
| gnarface | uh, yea i think optical media is expected to register as 100%, that seems normal to my recollection... | 23:31 |
| gnarface | how about anything like disk i/o errors in the dmesg output? | 23:31 |
| junalio | no, none | 23:32 |
| gnarface | been a couple releases since i've tested one of these, can you just skip that step and complete the install? | 23:34 |
| gnarface | as long as the system can boot, you should be able to configure apt afterwards | 23:34 |
| gnarface | then if there's a network issue i'll be easier to diagnose, at least | 23:34 |
| junalio | will try, one sec | 23:34 |
| gnarface | anything the netinstall would have downloaded to install you should be able to install just fine afterwards too | 23:35 |
| junalio | ah well, so keyboard input on the installer does not even register it seems | 23:35 |
| gnarface | hotplug the keyboard? | 23:36 |
| junalio | no matter if i try cancel or esc or anything else | 23:36 |
| junalio | it's a laptop | 23:36 |
| gnarface | is it on battery or AC? | 23:36 |
| gnarface | wait, keyboard was working though, wasn't it? you would have had to use it to switch virtual terminals, wouldn't you? | 23:37 |
| junalio | yes, the keyboard is working | 23:37 |
| junalio | on battery, just plugged it in to ac now | 23:37 |
| gnarface | i'm confused about whether the keyboard is working or not, you seem to have contradicted yourself... | 23:38 |
| junalio | all the alt+f1/2/3/4 switches work but when i am in the installer window (f1) and press enter, which should select "<Cancel>", nothing happens | 23:39 |
| gnarface | hmm, you're using the installer in graphical mode? | 23:39 |
| junalio | no, text | 23:40 |
| gnarface | hmm, something weird going on there, but i wonder if since it's a laptop maybe you plugged in the wrong keyboard mapping | 23:40 |
| junalio | i mean, the first one on the list of available installers when booting :) that is what i mean by text installer | 23:40 |
| junalio | i selected a german layout | 23:41 |
| gnarface | i haven't seen this revision, so i don't know what is first on the list, usually that's not the text mode though | 23:41 |
| gnarface | but i would imagine they are pretty distinct... | 23:42 |
| junalio | i will restart the installer in another mode, and see if it works then | 23:42 |
| gnarface | look for expert text mode | 23:42 |
| gnarface | expert mode will ask more questions | 23:42 |
| gnarface | and text mode should have less bugs | 23:42 |
| junalio | ok, expert install it now is | 23:43 |
| gnarface | for the keyboard mapping, you might need some alternate selection since it's a laptop | 23:43 |
| gnarface | something about "hidden altGR keys?" | 23:43 |
| gnarface | doesn't explain the stall though... | 23:43 |
| junalio | no, and it worked before, on the same device (it's a thinkpad x220) | 23:45 |
| gnarface | yea, i understand, and i'm pretty sure that one is known to work, but there's only so many things that could have changed | 23:48 |
| gnarface | currently the symptoms you report are possible matches for either a failing power supply or a network problem upstream of you, but even then without any errors it's still quite ambiguous | 23:50 |
| gnarface | are you able to get through the expert install just skipping the apt part? | 23:50 |
| gnarface | if it's a routing issue at your ISP, you may be able to get around it by dialing in a static mirror entry from the list after you're booting | 23:51 |
| gnarface | the netinstall should be able to complete setting up a bootable system without actually using the network, you just won't have much installed at first | 23:52 |
| gnarface | and of course there's also one of the full installer isos.... though i get that that's not the point, it might obviate a network problem at your ISP that you can't fix yourself | 23:52 |
| gnarface | if you have another network connection you can use, it might be worth a try in fact... like maybe you can use your cellphone as a hotspot just to complete a minimal install? just brainstorming here... | 23:53 |
| junalio | i just went through the expert installer up to the apt part and there it's stuck at 33% again... i even chose a different mirror this time | 23:53 |
| gnarface | but i think you can just back out and go on to the next steps and leave apt unconfigured | 23:54 |
| junalio | i only have that one connection. i will try a previous devuan version and see if that works | 23:54 |
| gnarface | that's also an option | 23:54 |
| junalio | no, i cant back out because enter key doesnt work again | 23:54 |
| gnarface | that's weird | 23:54 |
| gnarface | are you able to skip that step entirely though? i think you can | 23:54 |
| gnarface | if you don't start it, it might not go off the rails in the first place | 23:54 |
| gnarface | if the previous release works as an installer though you should be able to just upgrade from it no problem | 23:55 |
| junalio | i will try with a prev release and if that doesnt work i'll skip it :) will report back | 23:55 |
| gnarface | ok | 23:55 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
