libera/#devuan/ Monday, 2024-04-29

joergfirmware?00:20
plasma41archdummy: I've cloned both of those repositories to my local system. I'll take a look at them whenever I find time to set up snapper on my machine.00:34
integfredThunar file manager doesn't copy files using the GUI to external media properly. Copying by GUI results in file corruption. You can test this with cmp comman01:00
integfredUsing a different file manager (spacefm-gtk3) works for the time being01:00
integfredIt's important as I just ripped a computer apart trying to find the source of corruption01:01
integfredPretty severe01:01
ted-iousintegfred: What filesystem are you using?01:07
integfredext401:07
ted-iousOn both?01:07
integfredyes, I used a 6GB file to test01:08
integfredcopying by CLI was fine and copying using spacefm was fine01:08
integfredbut NOT Thunar01:08
ted-iousWhat kind of file?01:08
integfredI tried an ISO originally, then I dded some random stuff to a file to try that as well01:09
integfredIt only happens with external media01:09
ted-iousAnd you're testing for corruption with something like sha*sum?01:09
integfredcmp01:09
ted-iousOh right.01:09
integfredI'm testing one file against the other with cmp01:09
integfredbut shasum would work too01:10
ted-iousOk let's see if we can devise a test that will determine the nature of the corruption.01:10
integfredI also mentioned this on xfce IRC just now01:10
integfreddd if=/dev/urandom of=somefile bs=1M count=600001:11
integfredcopy the file via gui01:11
integfredthen01:11
integfredcmp somefile /path/to/externalmedia/somefile01:11
ted-iousDoes it corrupt with just one small file?01:11
integfredI haven't tried one small file01:12
integfredbut it's possible01:12
integfredI was getting corruptions because I wrote a script to check for silent corruption01:12
ted-iousI would try making a simple script to do larger and larger files until the cmp fails.01:12
ted-iousThat's good.01:12
integfredhttps://github.com/publicsite/consistencyChecker01:13
ted-iousI started using archive file formats for copying files around for the same reason.01:13
ted-iousOr else creating a md5sum file and comparing it on the other end.01:13
ted-ious7z and tarlz are really good for this.01:14
integfredi'm in the process of trying to make it use CRC-24 sum, converted to base64 (which is 4 bytes) and using a 4 byte CRC per 4KB01:14
ted-iousYou could just use tarlz and it will do all the checksums internally for you.01:15
integfredyeah, i don't want the compression01:15
ted-iousThat just needs a -0. :)01:15
ted-iousAlthough you can get minor compression for free with -3 since your cpu will be bored waiting for the usb and wants to do something more than twiddle its thumbs. :)01:16
integfredyeah ... I have slight paranoia haha01:18
ted-iousTrying to avoid any possible corruption?01:19
ted-iousI respect such paranoia. :)01:19
ted-iousYou should read about tarlz then.01:19
ted-iousI think it's the only foss archive format with compression that can both detect corrupt bits and recover from bad blocks.01:21
integfredactually i was part way through writing recovery too01:21
integfredbut this is not in the git yet01:21
ted-iousHave you looked at par2 yet?01:21
ted-iousThere's supposed to be a par3 in development but it will take years before everybody is satisfied and trusts it as much as par2.01:22
integfredI have never tried par2, I guess what I'm after is similar01:23
integfredbut part of the fun for me is to try and reinvent the wheel :)01:24
ted-iousDon't let me stop you then. :)01:24
ted-iousDoes your script do things that rsync -c doesn't?01:27
integfredi don't know what rsync -c does01:37
ted-iousIt forces rsync to compare every file instead of just looking at the date and time.01:54
integfredthe script I wrote in the git at the moment, as it stands, basically does that (compare every file in 2 different directories against each other). I am working on the other things I mentioned like CRC check and recovery to replace the cmp aspect to clarify. I am doing quite a bit of the work in C, but I haven't uploaded that bit yet.01:59
integfredIt's also worth noting that the bug in Thunar is distributed with devuan iso if you didn't already know this :)))02:47
JorilMmh... devuan.org is down...?10:44
Jorilback online now10:54
gnarfaceted-ious: if integfred comes back, tell them to see if it's also happening in debian, and if so report it there11:09
gnarfacei only skimmed the listing for thunar on their bug tracker, but a couple of the things on there looked possibly related11:10
gnarfaceif it's not actually happening in debian too, then i'm sure that info will incite someone to figure out why11:11
ErRandirTrying to upgrade tzdata on daedalus reports a grave bug "b1 - #1069265 - tzdata: Upgrade from 2023c-2 to 2024a-0+deb10u1 corrupts zoneinfo files". I don't want that bug. Looking at it in https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1069265 I don't see if or when it will be fixed. Is there anything in the bug that will tell me?19:19
fsmithredErRandir, I don't see that version in daedalus. It's up to +deb12u119:27
fsmithredand the version in daedalus is not referenced in the changelog.19:36
ErRandirThanks. I see that version in https://pkginfo.devuan.org/cgi-bin/policy-query.html?c=package&q=tzdata&x=submit but http://deb.devuan.org/ does not resolve for me. In /etc/apt I have http://packages.devuan.org/merged. Is that wrong?19:40
gnarfaceyou're the issue is that you're probably missing daedalus-updates and daedalus-security19:54
gnarfacedeb.devuan.org really should resolve though...19:54
ErRandirThanks, it's a bit better now. It tries to upgrade to 2024a-0+deb12u1, with the same bug as a blocker.20:00
gnarfaceErRandir: paste your /etc/apt/sources.list at paste.debian.net and then drop the link here, i'll sanity check it for you20:06
gnarfacemake sure your system clock is relatively accurate too (within about 5 minutes at least of the real time)20:07
ted-iousgnarface: Why would somebody go thru the effort to sync their clock and let it still be 5 minutes off? :)20:17
gnarfacewell, obviously it's easy to just use ntpdate or something like that, but in case they're setting it by hand they have to know it matters more now than it used to in the old days, and some people are tripped up because they don't expect that20:19
gnarfaceit used to be that you could trust everything to work if you were at least just set to some time later than whatever the most recent times in the logs were, and even then it usually wasn't a big deal to be years off, but now days you need to be comfortably within 5 minutes or you can't even trust DNS to work20:21
ErRandirhttps://paste.debian.net/1315565/20:23
ErRandirmy clock looks good.20:23
ted-iousgnarface: This works to within a second or two even if you don't have an ntp server. https://gist.github.com/alexellis/4ce511f08313a16d5a945a5ad27b0f2520:25
fsmithredErRandir, here's a package mirror list in case you want to pick one that's close instead of using deb.devuan.org20:26
fsmithredhttps://pkgmaster.devuan.org/mirror_list.txt20:26
gnarfaceErRandir: deb.devaun.org is a dns round-robin... is it possible you just hit a broken mirror and didn't retry? other than that this sources.list looks fine20:26
fsmithredwe prefer that people use a mirror rather than pkgmaster or packages.d.o.20:26
gnarfaceErRandir: your DNS servers could be suspect...20:27
ErRandirOk, I'll try a mirror.20:29
gnarfaceErRandir: oh, backports isn't supposed to cause problems, but sometimes it has for me, so usually i leave that one commented out except when i need something specific from there.20:30
gnarfaceyou generally do not want to actually do a full upgrade using backports, since those packages aren't usually tested well against each other20:30
ErRandirEven with a mirror it tries to go to 2024a-0+deb12u1, with the same bug as a blocker. Sort-of makes sense since the bug is against 2024a-0+deb10u1 and hasn't been fixed yet.20:47
gnarfaceErRandir: you're getting an error on install you mean? can you show it?20:49
gnarfacepaste.debian.net again20:50
ErRandirhttps://paste.debian.net/1315569/20:54
gnarfaceErRandir: what do you have installed that's actually doing the bug check there? it seems like it has a bug.20:55
gnarfaceit installed just fine for me, but now i'm wondering if my timezone data is corrupt...20:56
gnarfacebut i don't see why it'd be triggering on a bug for oldoldstable unless it was some sort of version parsing error20:57
ErRandirIt comes from package apt-listbugs, which runs /usr/bin/apt-listbugs during the install.20:59
gnarfacecheck what version of apt-listbugs you have21:00
ErRandir0.1.40, no newer version available for me. Maybe I should backup /usr/share/zoneinfo, install anyway, and debug what looks wrong.21:02
gnarfacewell, i'm wondering if you're actually seeing something like this: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=102311121:03
gnarface... so, i guess, maybe yea21:03
gnarfaceif apt-listbugs doesn't parse the part of the version string after the "+" and doesn't check release tags either, i can see why it would get a false-positive here21:07
gnarfacebut if that were the case it does seem to me like false positives would be very common...21:08
ErRandirI wasn't aware of that bug. I've done the install, I'll monitor if any bad stuff happens. Lots of the files in /usr/share/zoneinfo/ are binary.21:14
gnarfacewell, keep me posted if anything happens21:15
gnarfacekeep all of us posted21:15
bgstack15Hi, my latest Devuan install is on an old Acer Chromebook. Not sure if that affects my problem: I cannot figure out how to ignore the power button. I have tried modifying /etc/elogin/logind.conf to set all the possibly-related values to =ignore, I have tried chmod -x /etc/acpi/powerbtn-acpi-support.sh. Does anyone have any additional ideas?21:54
bgstack15I use elogind+lightdm.21:54
gnarfacebgstack15: i'm not sure why you'd want to do that, but fyi on most hardware, "10 seconds to power off" isn't something you can actually prevent without altering bios code21:56
gnarfacefor ignoring regular power button signals besides that though... maybe try lightdm or maybe look for a kernel module you can unload, not sure21:58
bgstack15the power button is right above the backspace key, on a small-ish keyboard. one (ahem) might press the power button when fixing one of four thousand typos a day.21:58
gnarfacethe login manager might have something to do with it since elogind is involved too, but i'm not sure21:59
gnarfaceactually does it have a bios/efi type interface? look in there for power button behavior first, if it does21:59
gnarfaceon a good bios it would definitely have an option22:00
bgstack15I'm not trying to ignore a 5-second press, just the regular little tap. It's an old chromebook so I think whatever SeaBIOS is, would be very locked down22:00
bgstack15Doesn't the OS get to decide what to do with a power key press though?22:00
gnarfacechromebook... yea, wouldn't expect any bios access on non-PC hardware22:00
gnarfacewith a regular power-button press on a chromebook, the OS certainly would if it was chromeos22:01
bgstack15it is x86_64, Celeron 2955U though22:01
gnarfaceif you're running linux on it... anybody's guess whether they even have drivers22:01
gnarfaceceleron chromebook... very interesting22:01
gnarfaceyou sure it doesn't have a regular bios? worth looking...22:02
bgstack15oh, the Devuan 5.0 netinstall worked great on it! (After using the chrx script that partitioned the drive and initially installed Xubuntu on it)22:02
gnarfacehuh, interesting22:03
bgstack15but then I installed Devuan to a usb drive because I was afraid to mess with the partition table of the 16GB onboard storage and ruining ChromeOS's ability to restore itself (if I decide to sell it).22:03
gnarfacedo you have the kernel config?22:04
gnarfacesee what this config var is set to: CONFIG_ACPI_BUTTON=m22:04
bgstack15er, 6.7.12-amd64 stock whatever's in Devuan unstable right now.22:04
gnarface(check in /boot/ or in /proc/config.gz)22:04
bgstack15CONFIG_ACPI=y22:05
gnarfacethat's the only ACPI variable in your kernel config!?22:05
bgstack15would you like a debian paste grep ACPI_ /boot/config-`uname -r`?22:05
gnarfaceno, just check if ACPI_BUTTON support is included, and if it's a module, unload it22:06
bgstack15ooh, so because it was y, it was built-in to this build of the kernel so I can't just unload it, I presume?22:06
bgstack15oh wait, CONFIG_ACPI_BUTTON=m22:07
bgstack15lol, `modprobe -r chromeos_acpi ; [press power button]` still shuts the system down.22:08
gnarfacei forget the generalized kernel command-line syntax for disabling modules... i think it can work if it's not a module but i guess we're not finding out today22:08
gnarfaceisn't there one just called "button" ?22:08
bgstack15Ah, I didn't check for "button." Let me look for that22:08
bgstack15this is the fastest rebooting device I've ever seen in person!22:08
gnarfaceyour window manager might also have some option, depending on the WM22:09
gnarfaceif it even can be intercepted that is22:09
bgstack15fluxbox.22:09
bgstack15But if we need to switch to xfce to get some config options that might work in any sense, I'm game.22:09
bgstack15I will try to modprobe -r button right now.22:10
bgstack15still shuts down without button module loaded.22:10
gnarfacei don't know much about fluxbox, but i think enlightenment would have a way to rebind it22:11
bgstack15Dare I say that it appears the systemd people have something that works.... https://unix.stackexchange.com/questions/288731/disable-the-power-button-on-a-chromebook22:11
bgstack15but I've already tried that of course in the /etc/elogind/logind.conf22:11
gnarfacethat /etc/acpi/powerbtn-acpi-support.sh script, you just chmod -x'd it? try renaming it22:14
gnarfacein fact, rename it and then restart acpid22:14
gnarfacethen try the power button22:14
gnarfacejust curious22:14
bgstack15yeah, I had already renamed that file about 10 reboots ago.22:15
gnarfacethat second suggestion from the stackexchange link might work22:15
ErRandirCould try to use xmodmap to reassign the button.22:15
gnarfacechanging the actual X11 keyboard mapping for it in /usr/share/X11/xkb/symbols/inet22:15
bgstack15after changing that inet file, would I need to reboot for it to take effect?22:17
gnarfaceprobably just need to restart X and lightdm, would be my guess22:18
gnarfacefull reboot might be easier considering your short boot times...22:18
gnarfacebut, looking at that file here right now i can see familiar looking symbols that confirm what i suspected, which is that enlightenment does have controls for this, so fluxbox might too22:19
gnarfacemaybe in some optional control panel you didn't install?22:20
gnarfacethat's often the answer with xfce anyway22:20
bgstack15I replaced every XF86PowerOff with Backspace in that inet file, rebooted, and still no go. I tried disabling the xinput devices that were labeled power button and sleep button. Still no go.22:23
bgstack15I'll try the xfce route I guess22:23
gnarfacethere's every possibility it's wired up to bypass the software, but there's no reason for that to be mutually exclusive either, i just don't know. i'd be curious to ask the elogind devs what they think...22:24
gnarfacewait, xfce? no, no, i was trying to get you to try enlightenment. the xfce thing was just brought up because it's an example of a desktop with lots of optional control panels people don't know about22:25
gnarfacei was specifically curious if these enlightenment controls i'm seeing can be made to affect it22:25
gnarface(on my regular desktop pc at some point in some relatively recent major release, my on-keyboard extra sleep button started working unexpectedly, so there's reason to believe it's gonna support weird buttons where other WMs may not)22:26
bgstack15OK, I'm going to install --no-install-recommends enlightenment right now.22:29
gnarfacewhen you get it up and running, check out settings -> settings panel -> input -> key bindings as well as acpi bindings in the same tab, i see power button controls under both22:31
gnarfacenot really sure how to use them though...22:31
bgstack15lol, startup sound for Enlightment...22:34
gnarfacei usually unload the keyboard map changer widget thingy because i find it causes conflicts sometimes with the system settings, which i had already set the way i wanted22:37
gnarface...but ymmv22:37
gnarfaceit has a really exhaustive set of control panels22:38
bgstack15enlightenment: I changed the "key bindings" for all XF86[power-sounding keys] to just do "File Manager" action. When I press the power button, it does pop up with the little "Power off/Suspend/Lock/Reboot/Hibernate/Log out" menu though.22:38
bgstack15So while my keybinding configs didn't work, it is at least intercepting that button.22:38
gnarfaceinteresting22:38
gnarfacemaybe better luck in the ACPI bindings control panel?22:38
bgstack15so there's hope!22:38
gnarfaceyea22:39
bgstack15I find Enlightenment... uninteresting to me. I'll check out the ACPI control panel next.22:39
gnarfacei just like that it's so flexible22:39
gnarfacedon't sweat the big memory footprint, it's all cache. it's actually very well optimized22:40
gnarfacethe multi-display support is pretty nice22:41
gnarfacethe default theme isn't flashy though22:42
bgstack15the ACPI commands worked; I configured the power/suspend type commands to File Manager, and it pops up a file manager when I press the power button.22:43
gnarfaceexcellent, so we know it can be done22:43
gnarfacethat's big22:43
bgstack15So I guess I need to research fuller acpi config/control.22:43
bgstack15I have to go for now, but thank you for the pointing in the right direction!22:43
gnarfaceyea, now just to figure out how to make fluxbox do it...22:43
gnarfaceno problem22:43

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