concatime | Does work: http://31.220.0.151/devuan/merged/pool/DEVUAN/main/u/util-linux/util-linux_2.32.1-0.1+devuan2.1_amd64.deb | 00:05 |
---|---|---|
concatime | Does work: http://31.220.0.151/devuan/pool/main/u/util-linux/util-linux_2.32.1-0.1+devuan2.1_amd64.deb | 00:06 |
concatime | Does work: http://31.220.0.151/devuan/devuan/merged/pool/DEVUAN/main/u/util-linux/util-linux_2.32.1-0.1+devuan2.1_amd64.deb | 00:06 |
concatime | Does not work: http://31.220.0.151/devuan/devuan/pool/main/u/util-linux/util-linux_2.32.1-0.1+devuan2.1_amd64.deb | 00:06 |
onefang | I just woke up, and haven't had brekky yet. This is the sort of thing that the mirror checker script I'm writing is for. It's at the "seems to work" stage, still needs to get to the "actually does a proper job" stage. | 00:07 |
onefang | Gimme a few minutes, I'l see what it reports for that file. | 00:08 |
concatime | Men, I’m lost in all these redirections! Take your time. | 00:09 |
onefang | My mirror checker script specifically delves deep into the rediretions, and multiple DNS records, probing each one all the way through. | 00:10 |
onefang | What results are you seeing? 404? | 00:18 |
concatime | yes | 00:23 |
onefang | https://mirror.koddos.net/devuan/devuan/merged/pool/DEVUAN/main/u/util-linux/util-linux_2.32.1-0.1+devuan2.1_amd64.deb works for me. The official mirror list says that mirror is at mirror.koddos.net/devuan/devuan Not mirror.koddos.net/devuan | 00:25 |
onefang | http://31.220.0.151/devuan/devuan/merged/pool/DEVUAN/main/u/util-linux/util-linux_2.32.1-0.1+devuan2.1_amd64.deb also works. | 00:26 |
onefang | https://31.220.0.151/devuan/devuan/merged/pool/DEVUAN/main/u/util-linux/util-linux_2.32.1-0.1+devuan2.1_amd64.deb fails, coz the certificate is for the domain name, not the IP. | 00:27 |
onefang | And I just uncovered a minor flaw in my checker script. lol | 00:29 |
onefang | But the above results are valid, the checker script was just not checking one of those. | 00:30 |
concatime | Ah ok, thank you | 00:31 |
concatime | You should add HTTPS to koddos in http://pkgmaster.devuan.org/mirror_list.txt | 00:33 |
onefang | Yeah a few of those Protocols lines are inaccurate, Evilham said he'll double check them all soon. The one for my mirror was to, but he fixed that at least. | 00:35 |
onefang | I'm the new Devaun mirror admin, but very new, I haven't been given the keys to the kingdom yet. | 00:35 |
onefang | Or "mirror herder" as I like to call it. B-) | 00:36 |
concatime | Hahah | 00:37 |
concatime | One last question, I should replace `deb http://mirror.koddos.net/devuan/merged ascii` by what | 00:37 |
concatime | deb http://mirror.koddos.net/devuan/devuan ? | 00:38 |
onefang | deb http://mirror.koddos.net/devuan/devuan/merged ascii Or for one of the other NL mirrors (mine) deb http://sledjhamr.org/devuan/merged ascii | 00:39 |
concatime | http://mirror.koddos.net/devuan/devuan/merged 404 | 00:40 |
concatime | Error 404 | 00:41 |
onefang | Hmmm, the mirror list says devuan/devuan, and that works when grabbing the file using wget, but not when putting http://mirror.koddos.net/devuan/devuan/merged into a web browser, where a single /devuan works. | 00:43 |
onefang | Try mine. At least i can fix that. B-) | 00:44 |
concatime | That what I said | 00:44 |
concatime | The issue from the beguining | 00:44 |
concatime | Dont’t take it personal, but I prefer a company-backed mirror. | 00:44 |
concatime | http://31.220.0.151/merged seems to work | 00:45 |
concatime | (koddos) | 00:45 |
onefang | Might be something odd in their rewrite rules, as you suggested. | 00:45 |
onefang | Something my mirror checker will need to test for. | 00:45 |
onefang | Does devuan/devuan/merged actually work if it's in your sources.lst? | 00:46 |
concatime | Only the ip one | 00:47 |
concatime | http://maemo.cloud-7.de/irclogs/freenode/_devuan/_devuan.2018-11-16.log.html#t2018-11-16T18:57:16 | 00:47 |
onefang | Hmm, I'll need to investigate more after I wake up properly. | 00:48 |
onefang | Or maybe Evilham can shed more light on this. | 00:49 |
concatime | I think `deb http://mirror.koddos.net/devuan/devuan` work. | 00:49 |
concatime | nop | 00:50 |
concatime | does not | 00:50 |
Evilham | Fun :-) | 10:12 |
onefang | I haven't managed to look at that yet. Been busy all day. | 10:13 |
Evilham | Fwiw: I said I'd check consistency, not accuracy :-D. Some things I've patched as I spot them | 10:13 |
Evilham | Will try to take a look over the day at this mirror | 10:13 |
onefang | Cool. | 10:13 |
onefang | And let me know the full details of the problem. I can't spot it. | 10:14 |
Evilham | I've focused more on keeping deb.devuan.org tested since that affects most people when misbehaving | 10:14 |
Evilham | Assuming I figure it out! Yes :-D | 10:14 |
onefang | Is there something subtle apt is doing that we are not testing for maybe? | 10:14 |
Evilham | concatime: it looks to me like mirror.koddos.net/devuan/devuan should work | 10:53 |
Evilham | (notice, without merged) | 10:54 |
onefang | concatime left many hours ago. | 11:00 |
Evilham | pf actually that mirror is being super strange | 11:05 |
onefang | We noticed. lol | 11:05 |
Evilham | it looked like it was working | 11:06 |
onefang | Yep, my script didn't find any problem. | 11:06 |
onefang | Does yours? | 11:06 |
onefang | That's the problem with these sorts of checkers, the goal is to find the problems. Hard to make sure you have tested for all of the strange problems. | 11:09 |
Evilham | it's not a strange problem, the rewrites do work with /devuan/devuan as a base URL, but the distribution files don't | 11:14 |
Evilham | it's something that we should test for in a standard fashion | 11:15 |
Evilham | aka, this list of files: https://dev1.cloud/mirrors | 11:15 |
Evilham | so, current script, the one that sends the email to the mirrors ML doesn't check for release files and the like and this particular mirror doesn't have a base URL that does both things right | 11:16 |
Evilham | it either has the dist files but no redirects or it has redirects but no dist files | 11:16 |
onefang | Those are the files I'm testing for, got that from your scripts. | 11:16 |
Evilham | then you should see it fail either on Release and Packages.gz or on /pool/DE{VU,BI}AN | 11:17 |
Evilham | cd | 11:17 |
Evilham | oops | 11:17 |
onefang | Yes, getting 404s on the Release files. | 11:18 |
Evilham | there you go then, it is properly setup for deb.devuan.org, so that's good | 11:19 |
Evilham | I added a note that the mirror is only usable over deb.devuan.org, we should ping the mirror admins | 11:20 |
Evilham | (the note should be visible in about 40 mins) | 11:20 |
onefang | Or less. B-) | 11:20 |
Evilham | anyway, everything looks decent on mirror land, I'm off \o | 11:20 |
onefang | See you. | 11:21 |
kts | Is there KDE in Devuan? | 11:32 |
onefang | A friend of mine is trying to get Devuan stuff installed on a QEMU VM running under Windows 10. Twice now an apt or apt-get install command screwed with her screen, spewing blank lines all over it. Has anyone seen that sort of thing before? i haven't. | 12:14 |
climbingturtle | onefang: Does it happens inside an DE or at cli? | 12:42 |
onefang | cli, it's a minimal install. | 12:43 |
climbingturtle | oh, I would be it was dirvers for graphics and DE. I have not expereinced it my self, though I have never run qemu on windows. | 12:46 |
onefang | Neither have I. | 12:46 |
climbingturtle | Does it only occurs with apt update and upgrade commands | 12:47 |
climbingturtle | ? | 12:47 |
onefang | I've used it a lot from a variety of Linux OSes. Sometimes with Windows as the guest. | 12:47 |
onefang | No, tee last one was apt-get install a big bunch of stuff. The first one was an update / upgrade. | 12:48 |
jiefk | Hi! I installed Devuan (from devuan_ascii_amd64_netinstall.iso, then upgraded to beowulf) on a laptop today. All is OK so far, except two things : | 17:06 |
jiefk | I have to wait a quite-not-so-short time staring at "Waiting for /dev to be fully populated..." | 17:07 |
jiefk | What could cause such a behaviour? | 17:07 |
jiefk | (I had to install various non-free firmwares (I guess) due to some messages appearing in dmesg | 17:08 |
drawkula | maybe its the action after this that has a delay befor outputting anything? | 17:09 |
jiefk | I don't know, but there's this thing [....] in front of the action that get changed to [.OK.] when done ? | 17:10 |
jiefk | it seems eth0 is taking some time to be configured also. | 17:15 |
jiefk | When I unplug the eth cable, it takes far more time ("waiting for /dev....." and "Configuring network interface"), slowing the boot time to about 2 minutes | 17:17 |
jiefk | is there some way to at least disable the waiting for eth0 to be ready ? (considering it to be a laptop on wifi) | 17:18 |
gnarface | jiefk: maybe a problem with your DHCP server... try a static ip for eth0 | 17:20 |
fsmithred | there's a patch for the eth0 delay - I'll find it | 17:20 |
fsmithred | https://dev1galaxy.org/search.php?search_id=2119130902 | 17:21 |
jiefk | fsmithred: Thanks for your input. Can you please link to the actual topic ? (I can't read the search results... "You are not logged in.") | 17:25 |
fsmithred | sorry | 17:27 |
fsmithred | https://dev1galaxy.org/viewtopic.php?pid=15493#p15493 | 17:27 |
fsmithred | jiefk, I recommend applying the patch manually, in case there's a difference in the script due to version differences | 17:29 |
fsmithred | I've used this same bit of code in ascii and beowulf | 17:29 |
jiefk | yes, thanks ! | 17:30 |
jiefk | I'll try that | 17:31 |
jiefk | yep fsmithred | 17:57 |
jiefk | this works but as Geoff42 from the forum says, $iface and $link are not set | 17:58 |
jiefk | (this works as "boots faster") | 17:58 |
fsmithred | jiefk, $iface and $link get set in the case statement above. If there's more than one in that list, I guess it will be set to the last one. | 18:22 |
fsmithred | (in the for loop above) This: for iface in $(ifquery --list --allow=hotplug) | 18:23 |
jiefk | fsmithred: id the laptop case, I see (at boot time, and also on service restart) `sys/class/net//operstate` (you see the missing $link between both // | 19:46 |
jiefk | -id +in | 19:47 |
jiefk | missing $link in this command : `if [ "$(cat /sys/class/net/$link/operstate)" = up ]` | 19:47 |
fsmithred | jiefk, do you have at least one interface set to "allow-hotplug"? | 19:58 |
jiefk | yep, eth0 ; I even manually added wlan0 too! | 20:00 |
jiefk | # ifquery --list --allow=hotplug | 20:01 |
fsmithred | hm, I switched my eth0 and wlan0 to allow-hotplug for testing, and now I can't access that computer | 20:01 |
fsmithred | shit | 20:01 |
jiefk | shows both interfaces. | 20:01 |
jiefk | argh sorry :/ | 20:02 |
fsmithred | no biggie - I just had to get up out of the chair | 20:03 |
fsmithred | ok, it would also help if I had network cables with little locking clips on the ends | 20:11 |
fsmithred | so they would stay plugged in | 20:11 |
fsmithred | anyway, I'm thinking the "ip link set $iface up" line should be in the for loop or have its own for loop. | 20:12 |
PinkBellyNagger | alright fellas? | 20:53 |
Nematocyst | I need to solve why my system requires /usr/lib/dbus-1.0/dbus-daemon-launch-helper to be group mlocate. every time I update dbus, the packaged one is group messagebus, which prevents slim from launching the DE [slim: open_session: Unable to open session: Failed to execute program org.freedesktop.ConsoleKit: Permission denied] changing back to group mlocate solves it until the next dbus update. what to change to permanently solve this | 21:08 |
Nematocyst | issue? | 21:08 |
fsmithred | jiefk, you here? I tried a modified patch. | 21:40 |
jiefk | yup, I'm here | 21:40 |
fsmithred | I put these two lines: | 21:41 |
fsmithred | ip link set "$iface" up || true | 21:41 |
fsmithred | if [ "$(cat /sysclass/net/$link/operstate)" = up ] | 21:41 |
fsmithred | in place of this line: (gimme a sec) | 21:41 |
fsmithred | if [ -e "/sys/class/net/$link" ] | 21:41 |
fsmithred | all in ifup_hotplug() | 21:41 |
fsmithred | it seems to be working | 21:42 |
fsmithred | sort of | 21:42 |
fsmithred | both eth0 and wlan0 are set to allow-hotplug and dhcp in interfaces | 21:42 |
fsmithred | there's no delay at boot with wire plugged in or not plugged in | 21:43 |
fsmithred | with wire plugged in, eth0 gets an address | 21:43 |
fsmithred | with wire out at boot, if I plug it in, it does not get an address | 21:43 |
jiefk | ok. I have some question : Can I -try to- rewrite this function as a whole ? This script (this function seems a bit hackish... | 21:45 |
fsmithred | sure, you can change it any way you want. Keep a copy of the original. | 21:45 |
fsmithred | caveat: I don't really know what I'm doing. | 21:45 |
jiefk | I'll try something and see if it works. I'll come to you later if I get something less hacked | 21:46 |
fsmithred | great. If you can come up with something better, I can find people to test it. | 21:46 |
fsmithred | maybe this should be part of the devuan-sanity package | 21:47 |
fsmithred | oh. ...if that ever gets packaged. | 21:47 |
jiefk | This script is a spaghetti :D | 22:05 |
jiefk | fsmithred: Can you try this function ? https://paste.debian.net/1089761/ | 22:10 |
jiefk | Hum no not yet, I still have to wait for eth0 :/ | 22:20 |
fsmithred | oh | 22:20 |
jiefk | I see that mooded the script from a pristine networking file :-\ | 22:24 |
jiefk | Modded | 22:24 |
jiefk | I apply now the patch you gave me before | 22:25 |
jiefk | The one from dev1galaxy forum | 22:25 |
fsmithred | I just made one that's a combination of mine and yours | 22:26 |
fsmithred | no delay | 22:26 |
fsmithred | and eth0 gets an address when I plug it in | 22:27 |
fsmithred | https://termbin.com/69we | 22:28 |
fsmithred | jiefk, I tried my first mod again, but I changed ifup $ifaces "$@" to ifup "$@" $ifaces and it works. | 22:37 |
fsmithred | eth0 gets an address when I plug in the cable | 22:38 |
jiefk | I have something that works too :-) | 22:38 |
jiefk | I'll post it somewhere in a few secs | 22:39 |
jiefk | pastebin.debian.net/1089763 | 22:42 |
fsmithred | Is there really an extra close-parenthesis in that function? | 22:46 |
fsmithred | there's on after the ifaces=$(blah) and then another after the "done" | 22:47 |
fsmithred | on/one | 22:47 |
fsmithred | that link isn't working for me | 22:48 |
fsmithred | ok, it's nested parens - I see it now. | 22:49 |
jiefk | https://paste.debian.net/1089763 | 22:49 |
jiefk | sorry, I wrote it from my head | 22:49 |
fsmithred | got it. will test | 22:49 |
fsmithred | no delay on that last one, but also no ip address after plugging cable | 22:57 |
jiefk | yup, just rebooted after plugged-unplugged the cable that's what I observed as well | 22:58 |
fsmithred | jiefk, do you have any other thoughts on this? I'm going to post the two variants that work on that d1g thread. | 23:12 |
jiefk | no but gimme a few minutes :) | 23:12 |
fsmithred | ok | 23:12 |
jiefk | Got an ip !! | 23:13 |
jiefk | from the patch at dev1galaxy | 23:13 |
fsmithred | oh? | 23:13 |
fsmithred | without changing it? | 23:13 |
jiefk | there's this action "echo "$iface" | 23:14 |
fsmithred | yeah, what's that about? | 23:14 |
jiefk | no idea, but I replaced it with an ifup command and I failed to obtain an ip address | 23:14 |
jiefk | I put bach this echo command and I got the IP | 23:15 |
jiefk | I'll reboot a few times more, and I get back at you | 23:15 |
fsmithred | ok | 23:15 |
concatime | yo, why isnt rDNS for ipv6 easy to setup? | 23:17 |
concatime | "we use tools such as sipcalc, subnetcalc to find the usable IPv6 ranges in the subnet" | 23:17 |
concatime | https://bobcares.com/blog/vultr-reverse-dns/ | 23:17 |
jiefk | I suppose the whole function goal is to return the interfaces to be processed later | 23:18 |
jiefk | I was trying to get the interface up inside the function and maybe this was not the time to bring the network up yet...? | 23:18 |
jiefk | I really don't know, . | 23:19 |
jiefk | New reboot : interfaces are ready --unplugged of course, so no IP. When I plug the cable, eth0 got an IP :) | 23:20 |
jiefk | reboot now, with cable plugged | 23:20 |
fsmithred | I just tried the one from the thread, and I get an ip on eth0 when I plug in the cable, but wlan0 is DOWN | 23:21 |
jiefk | Oh and best of all, no delay in the boot sequence :) | 23:21 |
fsmithred | I think that patch will only affect one interface | 23:22 |
jiefk | fsmithred: try from this paste (https://paste.debian.net/1089763) to replace : | 23:23 |
fsmithred | one of my mods has ifup in the for loop and it works | 23:24 |
jiefk | ifup "$@" "$link" || true | 23:24 |
jiefk | with | 23:24 |
jiefk | echo "$iface" | 23:24 |
jiefk | can you post your code, please? | 23:25 |
fsmithred | https://termbin.com/69we | 23:26 |
jiefk | yes, you have both the echo cmd and the ifup cmd | 23:27 |
jiefk | I bet that if you comment the ifup command, it will still work :) | 23:27 |
jiefk | Also | 23:33 |
jiefk | you have a syntax errot : /sysclass/net/$link/operstate is missing a / | 23:34 |
jiefk | after sys | 23:34 |
jiefk | error* | 23:34 |
fsmithred | that's my rubber keyboard - have to press hard | 23:37 |
jiefk | fsmithred: ok now I think I'm satisfied with what I have. Wifi does not get an IP because I did not hardcode SSID and passphrase in network/interfaces | 23:41 |
fsmithred | yeah, it works with the ifup commented out | 23:41 |
fsmithred | yeah, neither did I | 23:42 |
jiefk | but wlan sure if raised up :) | 23:42 |
jiefk | is* | 23:42 |
fsmithred | mine says state DOWN | 23:42 |
jiefk | do you have that allow-hotplug ? | 23:43 |
fsmithred | yes | 23:43 |
jiefk | hum... | 23:43 |
fsmithred | gonna turn off encryption and try it | 23:43 |
jiefk | what gives this command for you ? `ifquery --list --allow=hotplug` | 23:45 |
fsmithred | eth0 and wlan0 on separate lines | 23:47 |
jiefk | yes, that's what is expected. | 23:48 |
jiefk | trien running ip link set wlan0 up | 23:49 |
jiefk | ? | 23:49 |
jiefk | -n +d | 23:49 |
jiefk | dunno what to expect... | 23:50 |
jiefk | check mannually if cat "/sys/class/net/whan0/operstate" says "up" | 23:51 |
jiefk | wlan* | 23:51 |
fsmithred | ip link does nothing. operstate says down | 23:52 |
jiefk | me neither, ip link does nothing ., echo $? just after says 0 | 23:53 |
jiefk | but I have my interface up :/ | 23:54 |
jiefk | oh no it's down | 23:54 |
jiefk | but is it not expected because I did not give SSID and passwd? | 23:55 |
jiefk | I'll try that. | 23:55 |
fsmithred | that's what I'm doing now | 23:56 |
jiefk | but don't remember the syntax. | 23:56 |
jiefk | I misunderstood because wlan was listed in ifconfig | 23:58 |
fsmithred | oh | 23:59 |
fsmithred | I have a sample somewhere | 23:59 |
fsmithred | Scroll down to the bottom: https://refracta.org/docs/Release_Notes_9.0.txt | 23:59 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!