rrq | hook54321: did you try again? | 00:13 |
---|---|---|
hook54321 | rrq: golinux unblocked it, it's working so far. | 00:14 |
golinux | Not afaik . . . I wandered off to real life for a while . . . so he may not have seen my pm | 00:14 |
rrq | should be fine now; the baddie trap was a bit too eager | 00:16 |
hook54321 | is there anything i should do to avoid hitting it again? | 00:19 |
golinux | Not that I know of | 00:21 |
rrq | we had a small powwow with the trap bot imploring it to be kinder to well designed crawling | 00:26 |
rrq | ... as you might notice | 00:27 |
hook54321 | lol | 02:00 |
golinux | hook54321: No more problems? | 02:11 |
hook54321 | checking | 02:12 |
hook54321 | doesn't look it. running around 2-4 requests per second fine. | 02:13 |
hook54321 | errr, maybe 6-8, idk. | 02:13 |
search_social | Hello, I have just installed devuan and every few minutes or so my console is getting spammed with an "i2c_hid_get_input: incomplete report" message. I know the message is harmless and I'm wondering how to avoid seeing it on the console | 07:41 |
pablocastellanos | Hi, any ideas which packages provides the smallest window manager? | 08:03 |
tarzeau_ | amiwm and wmaker | 08:05 |
lts- | twm comes with X | 08:05 |
tarzeau_ | i read smallest+usable | 08:06 |
tarzeau_ | you can further make them smaller using upx on the binary | 08:06 |
pablocastellanos | By smallest I mean less space (I'm running beowulf on a text live CD headless) | 08:07 |
pablocastellanos | I only need to show one window | 08:07 |
pablocastellanos | twm sounds perfect | 08:08 |
tarzeau_ | ratpoision is what i use for kiosk mode | 08:08 |
lts- | You likely could run a single window with just xinit | 08:09 |
lts- | s/likely/possibly/ | 08:09 |
pablocastellanos | tarzeau_: Thats a good idea, but a non-experienced user needs to configure something using this window and ratpoison is not very user friendly | 08:09 |
tarzeau_ | pablocastellanos: that is fullscreen browser, borderless :) kiosk mode | 08:10 |
tarzeau_ | i doubt twm will be usable by non-experienced users | 08:10 |
pablocastellanos | tarzeau_: I have friends that use ratpoison, and awesome. | 08:10 |
onefang | Not sure why you need to manage windows if there's only the one window. | 08:11 |
pablocastellanos | tarzeau_: It has three buttons at the upper left corner | 08:11 |
pablocastellanos | tarzeau_: As far as I remember | 08:11 |
tarzeau_ | https://commons.wikimedia.org/wiki/File:FreeBSD_TWM_Window_Manager_Desktop.png | 08:11 |
pablocastellanos | onefang: Because this configuration window can open other windows as part of the same program, like opening files to import configuration | 08:12 |
onefang | The solution I used for that sort of thing was to use framebuffer, but that was a custom application. | 08:12 |
pandakekok9 | Is there someone looking at the inbox of repository@devuan.org? I've sent an encrypted email there yesterday | 08:14 |
onefang | I didn't even know that was such an email. What is your email about? | 08:15 |
pandakekok9 | onefang: It's about the SSH fingerprint of the gitea repo | 08:16 |
pablocastellanos | onefang: Looks like a lot of work. My setup will be needed only for a few hours. LOL | 08:16 |
pandakekok9 | Just making sure that I'm connecting to the right one | 08:16 |
onefang | Ah the source code repository, not the package repository. | 08:16 |
pandakekok9 | Oh, so they are different? I thought it also covered gitea... | 08:17 |
onefang | Our gitea is used for those few packages we dont inherit from Debian. | 08:17 |
golinux | I have never once heard of that address in the years I've been here. | 08:18 |
pandakekok9 | I just need to know what is the correct SSH fingerprint for the gitea repo | 08:18 |
onefang | Though mostly those we fork I think. I dunno much about that, I only run the package mirrors. | 08:18 |
pandakekok9 | Might be a good thing to document what SSH fingerprint you use on the website | 08:19 |
golinux | https://www.devuan.org/os/keyring | 08:20 |
golinux | Don't know if that's what you're looking for. | 08:20 |
golinux | Probably not. | 08:21 |
pandakekok9 | golinux: That's for pgp, I need ssh | 08:21 |
pandakekok9 | (And btw, you might want to update the devuan-keyring package, since some keys there are expired) | 08:21 |
golinux | Yeah. Do you have an account on gitea? | 08:21 |
pandakekok9 | golinux: Yes, it's jobbautista9 | 08:21 |
golinux | Ah, right. | 08:21 |
golinux | Unfortunately the gitea master is on holiday | 08:22 |
pandakekok9 | Oh, that sucks. Any idea when they are returning? | 08:23 |
pandakekok9 | It's not urgent btw, I can still use https for pushing | 08:23 |
golinux | And I haven't a clue about such things. You might open an issue there. He might see that. Or ask on the #devuan-dev channel. | 08:23 |
pandakekok9 | Ok, thanks | 08:24 |
golinux | LeePen would know and others there also | 08:24 |
golinux | Most everyone is sleeping now though. | 08:24 |
* systemdlete has buyers remorse re: btrfs... | 11:07 | |
systemdlete | I lost some data in some files on a btrfs file system on one system. Not sure if it is due to btrfs, but I'm wondering. | 11:07 |
systemdlete | I may have deployed btrfs too much at once. I thought because it had been working well on some systems it was safe to proceed with further rollout. I may have been wrong. | 11:08 |
systemdlete | Anyway, only SOME of the data in SOME files was lost. I have backups, luckily, heheheh | 11:08 |
systemdlete | I am restoring them now. | 11:08 |
systemdlete | whoa... | 11:10 |
systemdlete | now they are back? | 11:10 |
systemdlete | I did a restore, but to a /tmp directory. I haven't copied the files back. | 11:10 |
systemdlete | But somehow the files look right again. | 11:10 |
systemdlete | (previously, they had only one or two truncated lines in each of the files in question) | 11:11 |
systemdlete | wtf? | 11:11 |
systemdlete | I hope this sort of thing will not persist... | 11:12 |
systemdlete | no, I see what happened. | 11:13 |
systemdlete | The data is still missing. | 11:13 |
systemdlete | Some of the data is not missing. | 11:14 |
systemdlete | nvm. I will look into this some more. | 11:14 |
* systemdlete is redfaced with embarassment | 11:22 | |
systemdlete | Could someone please help me get my foot out of my mouth? | 11:22 |
systemdlete | I was relying on a construct to determine which files I wanted to gather data on. But that's not reliable, and I thought I had fixed that long ago. | 11:22 |
systemdlete | crud. | 11:22 |
systemdlete | but it is too early for most people here. Sorry for my "outburst" -- but at least I figured it out. It had nothing to do with btrfs... unless you count the fact that there is no mount.btrfs, which was the bad coding on my part... | 11:23 |
systemdlete | I have a better way to do this now (using lsblk) | 11:24 |
ShorTie | what is the big thing about btrfs ?? | 12:32 |
ShorTie | ext4 just works | 12:33 |
rm | snapshots | 12:35 |
ShorTie | guess i'm doning sumfin wrong | 12:38 |
ShorTie | i like never need to use a backup/snapshot | 12:39 |
MinceR | it destroys itself :> | 12:47 |
tarzeau_ | btrfs has live compression, and performs way better than ext4 (like xfs as well) | 12:52 |
* tarzeau_ hasn't touched ext4 since 15+ years | 12:52 | |
Unit193 | I dunno, https://launchpad.net/apt-btrfs-snapshot seems like an amazing idea. >_> | 12:54 |
GyrosGeier | xfs has never performed better than ext* for me | 12:57 |
GyrosGeier | my personal benchmark is "how long does it take to delete a linux kernel tree?" | 12:57 |
DPA | It would be nice if ext4 could do CoW, I have some libvirt-lxc containers where it would be nice if I could dedublicate some of their files. | 12:58 |
DPA | I don't like btrfs, though. Where did the space go, is there an old snapshot, does it need balancing, why does it suddently need so much CPU, etc. | 12:58 |
GyrosGeier | the last time I tried xfs, it took 45 minutes, while ext3 took 30 seconds | 12:58 |
tarzeau_ | GyrosGeier: try creating 10000 files, then removing them (or add a few more 0s) | 12:58 |
tarzeau_ | not sure ext3 had limits on # of inodes, unlike xfs/btrfs | 12:59 |
tarzeau_ | if that was fixed with ext4 | 12:59 |
GyrosGeier | it still has limits, but you can tell on creation what the limit should be | 12:59 |
GyrosGeier | same as in ext2 | 12:59 |
tarzeau_ | i'm aware of creation time limits, but i never know ahead whaty my users do, so i'm fine with xfs/btrfs | 12:59 |
GyrosGeier | there is a reason why d-i gives you an option "how is this file system going to be used?" | 12:59 |
tarzeau_ | never seen that question (using preseeded installs), does it have the option "no idea, but crazy unusual stuff" | 13:00 |
GyrosGeier | there is not much harm in using lots of inodes, except that it wastes space | 13:01 |
tarzeau_ | i've kept running out of them 15+ years ago | 13:02 |
GyrosGeier | so if you expect your users to do weird stuff, you can increase the number for /home | 13:02 |
tarzeau_ | and that filesystem hasn't been checked for 180 days stuff also bugged me | 13:02 |
GyrosGeier | or have inode quotas to stop them from breaking each other's stuff | 13:02 |
tarzeau_ | they use local disk space, not /home (way too slow in our setup) | 13:02 |
tarzeau_ | or just use xfs or btrfs :) | 13:03 |
GyrosGeier | in traditional sysadmining, you have a separate /var/spool with lots of inodes if you run a news server | 13:03 |
GyrosGeier | normal user homes tend to have a good mix of large and small files, and I've seldom run out of inodes on homes | 13:04 |
GyrosGeier | and for multiuser I have quotas | 13:04 |
onefang | Think you have gone #devuan-offtopic, or should. | 13:04 |
Kitty | I am trying to use udev rules to name my ethernet devices properly | 15:15 |
Kitty | the file is being picked up | 15:15 |
Kitty | and then sometime later in the boot process, the device is being renamed | 15:15 |
Kitty | overwriting it | 15:15 |
gnarface | Kitty: check if it's cached in this file /etc/udev/rules.d/70-persistent-net.rules | 15:19 |
gnarface | or something similar | 15:20 |
gnarface | i think it might get applied last, overriding your custom rules until you delete the old line manually | 15:21 |
sadsnork | Turns out my devuan mirror filled up my drive pretty fast. This morning I had to run the rsync with --delete and it freed up a couple dozen GB of space. Should I continue to use --delete or was there a specific problem that created the recent increase in space? | 15:30 |
onefang | Keep using --delete. | 15:31 |
sadsnork | Deal! :-) | 15:33 |
sadsnork | PS: Thanks onefang. | 15:34 |
onefang | You are welcome. Thanks for running your mirror. | 15:35 |
* sadsnork tips his hat | 15:37 | |
sadsnork | To be honest, a lot of what drives me is principle... and my use of Devuan is certainly no exception. Lately [thanks to links here] I have been reading a bit of that unixsheikh and related stuff - and I have been finding myself quite pissed off at the direction and decisions of the Linux community at large. | 15:39 |
sadsnork | I'm glad Devuan is an exception to that and want to encourage it. | 15:40 |
Ankokukishi | me too bro | 15:42 |
Ankokukishi | i just started using linux last year (wish it was sooner tbh) and what the mainstream distros do just pisses me off. My friend recommended refracta to me, and now its all i use | 15:43 |
Ankokukishi | i wish i had the programming chops to make a version that was compatible with laptops | 15:43 |
Ankokukishi | be the one distro that actually cared about laptops | 15:43 |
* sadsnork cheers | 15:44 | |
sadsnork | Nice Ankokukishi! I also wish my dev skills were better. I guess my skills being more suited to infrastructure led me to running a mirror. | 15:45 |
Ankokukishi | i tried main devuan before, i just like the interface of refracta more, but id recommend devuan to anyone. Programmers who actually have principles and care about functionality over bloat | 15:45 |
Ankokukishi | i have zero dev skills | 15:46 |
Ankokukishi | lol | 15:46 |
Ankokukishi | i wanna remedy that eventually but a new job and having to learn so much for that right now is gonna keep me away from being able to learn to dev | 15:46 |
Ankokukishi | lol | 15:46 |
Kitty | Hi, I am trying to use custom network device names. | 16:02 |
Kitty | I have ten network devices in this box, and the order from the kernel, either in the eth* format or the enp101s0f1DEADCAT format is a pain in the arse to use | 16:03 |
Kitty | I am trying to use 70-persistent-net.rules | 16:03 |
Kitty | but it doesn't work | 16:04 |
gnarface | permissions maybe? | 16:04 |
gnarface | on the rules file? | 16:04 |
Kitty | root.root 644 | 16:07 |
Kitty | if I set net.ifnames=0, then it picks up on the fact I tried to rename eth0-4, and just uses eth10-13 instead, with nothing having eth0-4, if I use ifnames=1, then it just gets eno? and enp101sf10p1a981 names | 16:09 |
gnarface | hmm, i'm not sure what's up | 16:11 |
GyrosGeier | net.ifnames should be largely irrelevant with udev rules | 16:12 |
GyrosGeier | that is just how the kernel names them initially, before the rules are applied | 16:12 |
GyrosGeier | the generated rules map from MAC address to persistent interface name | 16:13 |
gnarface | could it be caching the persistent names elsewhere? in a second file somewhere that's still not being cleared? | 16:13 |
GyrosGeier | if some of these are virtual functions, they may have random MAC addresses, which would allocate new names on every boot | 16:14 |
GyrosGeier | I think it shouldn't, but I haven't looked at it for a long time | 16:14 |
gnarface | oh, yea if the mac addresses are new every boot they'd also look like brand new devices in this case | 16:14 |
gnarface | it wouldn't be on by default unless something has gone very wrong | 16:14 |
GyrosGeier | I have configured the VF interfaces to use the vfio driver, which kind of stops them from getting a name | 16:15 |
GyrosGeier | well, how many VF devices show up can be set in the PF via kernel parameter, and on some machines through the BIOS | 16:16 |
Kitty | they are real hardware devices | 16:16 |
GyrosGeier | $ lspci | grep Ethernet | wc -l | 16:17 |
GyrosGeier | 16 | 16:17 |
GyrosGeier | all of these talk to the same port | 16:17 |
Kitty | 10 | 16:18 |
GyrosGeier | 14 have random MAC addresses that change on every boot | 16:18 |
Kitty | 4 are 1Gbps, 2 are 10Gbps copper, and 4 are 10Gbps fibre | 16:18 |
Kitty | all have hardwired mac addresses | 16:18 |
GyrosGeier | ah | 16:18 |
GyrosGeier | because I just oversimplified things | 16:18 |
GyrosGeier | in fact, my 16 PCIe devices talk to the same 4 ports through an internal switch :) | 16:19 |
Kitty | is there some other location udev rules might hide? | 16:21 |
gnarface | in /lib/udev/rules.d maybe? | 16:22 |
gnarface | i didn't think any should be going there from the persistent rules but i dunno for sure really | 16:22 |
gnarface | also something could go wrong | 16:22 |
gnarface | also, in the past it's been somewhat uncooperative for me and i've had to restart it | 16:22 |
gnarface | sometimes it seems to exhibit tweaky behavior based on file timestamps like pam does | 16:23 |
Kitty | and suddenly it works | 16:31 |
Kitty | what | 16:31 |
Kitty | the | 16:31 |
Kitty | utter | 16:31 |
Kitty | fuck | 16:31 |
gnarface | yea i dunno | 16:33 |
gnarface | suspicious to say the least | 16:33 |
gnarface | ran into similar issues when i was making custom rules for my steam controller | 16:33 |
gnarface | after i got it working the first time it was fine | 16:34 |
Kitty | I removed KERNEL== bit from the udev rules file | 16:36 |
Kitty | so it just has subsystem, action, drivers, attr(address) and name | 16:37 |
gnarface | what had you set it to? | 16:38 |
Kitty | various of 0000:66:00.0 | 16:38 |
Kitty | and "eno*" | 16:39 |
gnarface | hmm, i still dunno | 16:39 |
gnarface | as long as you got it working now though, that's the important part | 16:39 |
Kitty | let's reboot and prove it's not a fluke | 16:40 |
rkta | There is an jessi image for the raspi, but non for beowolf. Is support for the raspi dropped? If yes, are there recommended alternatives? | 17:41 |
buZz | did you find http://arm-files.devuan.org/ yet? | 17:43 |
buZz | rkta: ^^ | 17:44 |
rkta | no | 17:44 |
rkta | That's what I was looking for, thx buZz | 17:45 |
hook54321 | golinux: connect timeouts again | 20:03 |
gour | hello, i'm unning ceres/runit on my desktop machine and wonder which packages you recommend for "ntp" & "cron" jobs? | 21:23 |
gour | *running | 21:24 |
golinux | hook54321: Yeah, you're back on the list. | 21:27 |
golinux | Maybe you need to slow it down a bit? I really don't want to have to keep doing this dance. | 21:28 |
golinux | Or pause till rrq gets back and finds a way around it. | 21:31 |
hook54321 | golinux: yeah, slowing it way down | 22:25 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!