linearain | well i tried converting my buster 10.1 to devuan 3.0.0 on a test system and it was a complete mess. At one point it asked me if i want to delete the kernel? hundreds of packages removed and upgraded etc. I'm not a fan of upgrades and updates and all that stuff. I like to use what I got and only upgrade it if it breaks. At least now I know the only option is to live with systemd or just install a fresh | 13:38 |
---|---|---|
linearain | systemd-less OS on a different partition and then slowly decomission the old os partition lol | 13:38 |
ShorTie | just install a fresh is really the only option imho | 13:39 |
linearain | yeah, debian sucks that it doesnt support normal init systems | 13:40 |
ShorTie | thats the adavantage of playing arm in a way i guess | 13:41 |
linearain | arm? | 13:42 |
ShorTie | 10 bucks gets you a new sdcard | 13:42 |
ShorTie | fresh starts are cheap and easy .. :/~ | 13:43 |
linearain | but sometimes its unconvenient when u have lots of files and configuration on a single filesystem | 13:44 |
gnarface | linearain: if you know what you're doing that upgrade process still works. it's easy to screw up, but it's actually possible to complete the upgrade as long as you know to install a kernel before the next time you reboot | 15:28 |
gnarface | some tangled nest of dependencies meant that many of the replaced packages have to be removed before they can be replaced, and in some cases one of those packages becomes the running kernel | 15:30 |
gnarface | a fresh install is certainly easier | 15:30 |
gateway2000 | I have installed beowulf for the first time today and I am enjoying it so far | 15:43 |
Ankokukishi | Im on refracta on both my desktop and laptop | 15:45 |
Ankokukishi | my friend told me about refracta, tried it on one laptop, and when the mint devs cucked to a bunch of crybabies over chromium | 15:46 |
Ankokukishi | that was my sign to go full in | 15:46 |
gateway2000 | whats a difference between devuan and refracta? | 15:47 |
Ankokukishi | refracta is just a fork of devuan | 15:47 |
Ankokukishi | basically the same thing, fsmithred is a devuan dev and decided to do his own thing with refracta | 15:48 |
gateway2000 | o ok nice | 15:48 |
Ankokukishi | i tried the main devuan distro once, i liked it but refracta is a tad lighter | 15:48 |
Ankokukishi | i think its like 50-100mb on ram lighter | 15:49 |
Ankokukishi | plus it comes with refracta2usb, meaning you can write your current refracta settings to a USB and take your exact settings with you anywhere | 15:50 |
Ankokukishi | devuan can install it, i heard of people on other distros using it too | 15:50 |
Ankokukishi | but its built in with refracta | 15:50 |
Ankokukishi | i mean there arent THAT many differences between them, i just prefer refracta | 15:52 |
gateway2000 | sure whatever works | 15:53 |
Ankokukishi | yeah | 15:53 |
Ankokukishi | im happy as long as i dont have that garbage systemd in my os | 15:53 |
Ankokukishi | switching to devuan/refracta made me spend more time in the terminal, which is why i switched to linux in the first place for anyway | 15:54 |
Ankokukishi | im old school, i was using dos 5.0 WAAAAAY back in the day lol | 15:55 |
Ankokukishi | first pc i ever used was in 1987, i think it was a mac (before apple bought them) with the green and black screens lol | 15:55 |
crashoverride | The people who wrote RFC5322 should be locked into a room filling with concrete, that can only be open if they write a regular expression that can parse the format they described for email addresses. They should have 4h until the room is full of concrete. | 15:55 |
Ankokukishi | lol | 15:56 |
crashoverride | I mean... "space and "(),:;<>@[\] characters are allowed with restrictions (they are only allowed inside a quoted string, and in addition, a backslash or double-quote must be preceded by a backslash)" | 15:58 |
crashoverride | so you CAN have spaces in email addresses | 15:58 |
crashoverride | but it has to be "my super email address"@domain.com | 15:59 |
crashoverride | WHY was ever that a requirement?! | 15:59 |
Ankokukishi | probably because of microsoft? | 15:59 |
Ankokukishi | considering it was the 90s, and we ALL know what happened in the 90s | 15:59 |
crashoverride | sounds believable | 15:59 |
Ankokukishi | i mean if someone REALLY wanted to, couldnt they change the format and make something new? | 16:00 |
crashoverride | no. | 16:00 |
crashoverride | that's actually much harder. | 16:00 |
Ankokukishi | gah | 16:00 |
crashoverride | if you ever worked in a company who has a small product that has a HUGE technical debt, you'd know that :D | 16:01 |
crashoverride | "can't we re-implement?" | 16:01 |
Ankokukishi | lol | 16:01 |
Ankokukishi | i havent, so i dont know that lol | 16:01 |
crashoverride | "well, our customers have various versions of the product, with various different bugs and features combinations, so our backend is mixed javascript, visual basic, perl, prolog, ocaml and COBOL." | 16:02 |
Ankokukishi | yuck | 16:03 |
Ankokukishi | jabbashit | 16:03 |
Ankokukishi | im building a website for my friend in html/css only | 16:03 |
Ankokukishi | and css is only for the look of the page | 16:04 |
Ankokukishi | i refuse to use jabbashit at all | 16:04 |
crashoverride | "but by all means, if you want to re-implement that, AND DOCUMENT IT, you're welcome to do it. We just won't put it in production in the next 3 years because it has to be fully reviewed, and tested, and "we don't fix something that works"(tm); and also, you have to do it on your free time because we won't pay you to do that. And if still want to do it, you could look at the attempts from Oleg, | 16:05 |
crashoverride | Francis, Tamal, Ekon and Brad. They all tried and failed." | 16:05 |
Ankokukishi | lol | 16:06 |
crashoverride | Also, when you fix such a thing, you always have a customer who needed the feature | 16:07 |
crashoverride | for example, user.name(fromaccounting)@domain.com will be used by a dozen of people | 16:08 |
crashoverride | and they will all take many hours of customer support when you'll push your "fix" | 16:08 |
Ankokukishi | thats dumb | 16:09 |
crashoverride | and then there's THIS guy | 16:09 |
crashoverride | https://xkcd.com/1172/ | 16:09 |
Ankokukishi | computers used to be about innovation | 16:09 |
Ankokukishi | LMAO | 16:09 |
crashoverride | Ankokukishi: you're new to this aren't you? | 16:10 |
crashoverride | "computers used to be about innovation" | 16:10 |
crashoverride | that was true in the 70s | 16:10 |
crashoverride | maybe a bit in the early 80s | 16:10 |
Ankokukishi | good point | 16:10 |
Ankokukishi | lol | 16:10 |
crashoverride | but when most of the traders realized they machines were helping them big time; innovation was just for the general public. | 16:11 |
crashoverride | and this, right here, is why COBOL is still big. | 16:11 |
crashoverride | Look | 16:12 |
crashoverride | https://www.nytimes.com/2019/10/24/us/nuclear-weapons-floppy-disks.html | 16:12 |
crashoverride | this is from 2019 | 16:12 |
Ankokukishi | ew fake news website | 16:12 |
Ankokukishi | lol | 16:12 |
Ankokukishi | HISS | 16:12 |
crashoverride | And then there's this: https://hackaday.com/2015/07/23/this-little-amiga-still-runs-school-districts-hvac/ | 16:15 |
Ankokukishi | that just means amiga 2000s are chad computers :) | 16:16 |
crashoverride | myeah | 16:17 |
crashoverride | anyway | 16:17 |
crashoverride | What we need is to: | 16:17 |
crashoverride | 1. formalize better specs, that are based on current ones but with more restrictive requirements | 16:17 |
crashoverride | 2. implement systems that support BOTH specs, while making an effort to slowly transition | 16:18 |
Ankokukishi | true | 16:18 |
crashoverride | 3. sunset the old specs, while progressively adding artificial execution delays to those techs | 16:19 |
crashoverride | (and this last method is the most effective way. If you simply stop supporting it, people will complain. If you don't add delays, they will say "we will change" and never will) | 16:19 |
crashoverride | and that's it. | 16:19 |
crashoverride | you never fully stop supporting the old techs, or... you can, when the artificial delay is over 50 years. | 16:20 |
crashoverride | and even then you should be careful, someone might just sue you. | 16:20 |
crashoverride | but yeah, that's about it. | 16:21 |
crashoverride | so login to your mail provider with your "email that really is so funny"@domain.com? Login accepted. With a notice tho, that the login is sunset. | 16:21 |
crashoverride | and as time passes, the login page will be slower and slower. | 16:21 |
buZz | my 2012 motorola phone runs modern devuan really nicely though :) | 16:22 |
crashoverride | eventually, you will surely go for email_that_really_is_so_funny@domain.com and login with no delay. | 16:22 |
buZz | thanks to #maemo-leste team | 16:22 |
crashoverride | buZz: nice :) | 16:22 |
buZz | https://en.wikipedia.org/wiki/Droid_4 | 16:22 |
buZz | https://leste.maemo.org/Motorola_Droid_4 | 16:22 |
buZz | :) | 16:22 |
crashoverride | I should really get one of those. | 16:22 |
buZz | they're cheap if you're near US | 16:22 |
crashoverride | supports 4G man | 16:23 |
crashoverride | buZz: I'm smack in the middle of the EU | 16:23 |
crashoverride | so... | 16:23 |
buZz | i dont think it support EU 4G | 16:23 |
buZz | but no clue, i dont really use cellular stuff | 16:23 |
buZz | its just a PDA to me :) | 16:23 |
crashoverride | well, I need something that supports 4G | 16:23 |
crashoverride | and I would sell my soul for a physical keyboard. | 16:23 |
buZz | 'something' is easy | 16:24 |
Ankokukishi | im in japan motorola doesnt exist here lol | 16:24 |
buZz | 'something with devuan' | 16:24 |
buZz | is a lot harder | 16:24 |
crashoverride | buZz: actually I'm gonna put NetBSD on it. | 16:24 |
crashoverride | so... you know. | 16:24 |
buZz | rofl | 16:24 |
buZz | you're not so troubled by lack of support then, nice | 16:25 |
crashoverride | the lack of what? | 16:25 |
buZz | you know, working drivers | 16:25 |
crashoverride | you mean, my ability to copy paste code from freebsd? :D | 16:25 |
buZz | or even booting :P | 16:25 |
crashoverride | what's booting anyway? | 16:25 |
buZz | if you never boot it, it cant crash either! | 16:25 |
crashoverride | true systemd philosohpy. | 16:26 |
crashoverride | You can get it for 40 bucks on amazon US tho | 16:26 |
crashoverride | wow. | 16:26 |
buZz | mine was funded by EU | 16:27 |
buZz | or well, by NLNet who got their funding from EU afaik | 16:27 |
crashoverride | funded? | 16:27 |
crashoverride | what do you mean? | 16:27 |
buZz | many EU govs want more opensource dev | 16:27 |
buZz | they fund good projects with money | 16:27 |
buZz | maemo leste project used funding to buy a pile of devices and distribute them amongst devs | 16:28 |
crashoverride | ah yeah | 16:28 |
crashoverride | is the droid ultra similar? | 16:29 |
buZz | no, completely different, see the specs | 16:29 |
buZz | droid 4 is very much alike the N900 | 16:29 |
buZz | but more faster cores, more ram, etc | 16:29 |
crashoverride | ah | 16:30 |
buZz | maemo leste obviously targets n900 ;) but also some other devices, like droid4 | 16:30 |
crashoverride | yeah | 16:30 |
buZz | btw plain devuan (and probably also some of those BSDs you mentioned) should work on n900 totally fine | 16:30 |
crashoverride | I know about the N900, but for this VERY reason, it's impossible to find. | 16:31 |
buZz | really? i got 3 and i barely tried | 16:31 |
crashoverride | the N900? | 16:31 |
buZz | yeah | 16:31 |
crashoverride | Where?! | 16:31 |
buZz | also have the N810 N800 and 770 internettablet | 16:31 |
buZz | find some n900 usergroups, ask around :) | 16:32 |
crashoverride | yeah well | 16:32 |
crashoverride | in the EU, it's gonna be 'spensive | 16:33 |
crashoverride | you're in the US, right? | 16:33 |
buZz | i'm in netherlands | 16:34 |
crashoverride | < buZz> they're cheap if you're near US | 16:49 |
crashoverride | I assumed the US because of this | 16:49 |
buZz | hmhm, the point was, they only got sold on US market afaik | 16:50 |
buZz | they only started to work in EU after some firmware upgrades later in its life | 16:50 |
crashoverride | aah | 16:50 |
crashoverride | so they're common place in the US | 16:50 |
crashoverride | and rather uncommon here | 16:51 |
crashoverride | correct? | 16:51 |
buZz | from what i've seen, yeah | 16:52 |
buZz | the one i have has 'verizon' on the front panel :) | 16:53 |
crashoverride | that's the only one I can seem to find. | 16:58 |
buZz | i'm not sure a different version even exists | 17:00 |
crashoverride | wikipedia can be really stupidly written, sometimes. | 17:01 |
crashoverride | "space and special characters "(),:;<>@[\] are allowed with restrictions (they are only allowed inside a quoted string, as described in the paragraph below, and in addition, a backslash or double-quote must be preceded by a backslash)" | 17:01 |
linearain | i installed sysvinit-core with synaptic on debian buster and it just works somehow | 17:01 |
buZz | linearain: it removed systemd? | 17:01 |
crashoverride | until "quoted string", it works. | 17:01 |
crashoverride | and then everything falls apart. | 17:02 |
linearain | i think not, but it booted with sysvinit instead of systemd | 17:02 |
crashoverride | first, the text instructs that any backslash has to be an infinite loop of backslashes | 17:02 |
fsmithred | linearain, did you get any further than that with the migration? | 17:02 |
buZz | crashoverride: seems pretty irrelevant to devuan | 17:02 |
crashoverride | buZz: fine. | 17:03 |
crashoverride | like I could predict someone would come writing about systemd. | 17:03 |
linearain | well i got so lost with that migration thing... so i just decided to try installing sysvinit | 17:03 |
fsmithred | well, that's the first step in nixer's instructions | 17:03 |
linearain | i just wanna learn a bit more about how things work | 17:03 |
crashoverride | fwiw it's usually just simpler to install devuan. | 17:03 |
crashoverride | if your goal is learning, however, simplicity is usually conflicting. | 17:04 |
linearain | oh i followed the instructions in devuan.org | 17:04 |
linearain | main page | 17:04 |
fsmithred | yeah, we discussed this the other day | 17:04 |
buZz | ah the ones ppl said wouldnt work :P | 17:04 |
linearain | yeah sorry too much info forgot | 17:04 |
crashoverride | linearain: and you ended up with a deBIan system? :P | 17:04 |
fsmithred | that set does not have you install sysvinit-core - it's supposed to be pulled in automatically when you install eudev | 17:04 |
linearain | somehow eudev failed to install if i recall correctly | 17:05 |
fsmithred | that's normal | 17:05 |
fsmithred | then you do it again after reboot | 17:05 |
linearain | but again, my buster is not updated. I installed it in late 2019 and never updated any packages since then | 17:05 |
fsmithred | at least that's how it was when it was working | 17:05 |
fsmithred | I don't think that matters | 17:06 |
linearain | good to know | 17:06 |
linearain | i will surely try migrating again, now im a bit worked up | 17:06 |
fsmithred | the guy who wrote the directions on the website also did stretch or jessie to beowulf migration | 17:06 |
linearain | so i keep copy pasting partitions with gparted because i keep messing them up | 17:07 |
fsmithred | you can learn things by seeing how many ways you can break something | 17:07 |
linearain | anyway, after i booted with sysvinit, everything seemed to work, except internet hostnames or whatever, they didnt work (like devuan.org), and therefore the URLs in sources.list would not resolve, i couldnt ping devuan.org, but i could ping it by IP | 17:09 |
fsmithred | /etc/resolv.conf probably got screwed up. Or did that get replaced with a systemd thing? | 17:10 |
linearain | to be honest i don't remember | 17:10 |
linearain | is it a hard fix? | 17:10 |
fsmithred | no, just put one line in the file | 17:11 |
linearain | alright | 17:11 |
fsmithred | nameserver <some-ip-address> | 17:11 |
fsmithred | the address can be your router's address on the local network | 17:12 |
linearain | yeah i just noticed it | 17:12 |
fsmithred | 192.168.1.1 or maybe .0.1 | 17:12 |
linearain | anyway, after booting with sysvinit, after the login welcome message, it gets stuck for a few minutes until the prompt appears | 17:12 |
linearain | crashoverride: well its a server with no gui, so i guess kind of | 17:15 |
fsmithred | linearain, here's the summary of nixer's method. Scroll up for more detail. https://dev1galaxy.org/viewtopic.php?pid=18390#p18390 | 17:15 |
linearain | i got that thread bookmark, alright ill look at it now | 17:16 |
Batoeh | Good morning everyone. Something I am running into with the minimal iso is that it will not boot on my lenovo w530. Both on a cd or a usb but it seems to work fine on my desktop. I feel like I ran into this issue once before but I am not sure. Only seems to be devuan so far in my testing. Stuff like Ubuntu, FreeBSD, and OpenBSD are able to boot into their install media no problem. Tried different USBs, | 17:21 |
Batoeh | CDs, and ISO files. Do any of you have any suggestions? | 17:21 |
crashoverride | linearain: I didn't get that | 17:22 |
fsmithred | Batoeh, is it uefi? | 17:23 |
Batoeh | Oh almost forgot the really interesting part is that the laptop can boot into the devuan-desktop install media no issue too. Just seems to be server. | 17:23 |
mason | Ooh, that's notable, yeah. | 17:23 |
Batoeh | Yee but I am fairly sure I adjusted the appropriate settings for it. I mean the desktop I tried was a threadripper system so fairly recent hardware vs the laptop. Thanks for the assistance so far. | 17:24 |
crashoverride | Batoeh: then install via PXE? :D | 17:24 |
fsmithred | server iso or minimal-live iso? | 17:24 |
fsmithred | the minimal-live will not boot in uefi mode | 17:25 |
crashoverride | well, here you have it. | 17:25 |
crashoverride | personally, with UEFI, I usually use PXE. | 17:25 |
fsmithred | neither will the i386 desktop-live | 17:25 |
crashoverride | fsmithred: if they run a threadripper, I sure hope they don't use the i386 arch. | 17:25 |
Batoeh | I might just give that a shot crashoverride but I never tried a pxe boot/install. Yeah but I am pretty sure I am set it to legacy boot. No worries. Worst case if all it needs is a pxe install I am down to learn. | 17:26 |
fsmithred | lol, yeah good point | 17:26 |
fsmithred | I'm still not clear on which iso won't boot | 17:26 |
fsmithred | and I really would like to know | 17:27 |
crashoverride | I'm assuming minimal-live. | 17:27 |
crashoverride | x86_64. | 17:27 |
fsmithred | that would be good | 17:27 |
crashoverride | Batoeh: ^ ?? | 17:27 |
fsmithred | if it's anything else, then something needs to be fixed | 17:27 |
crashoverride | yep | 17:28 |
crashoverride | also, regarding UEFI and boot | 17:28 |
crashoverride | that's a bug I wanted to report. | 17:28 |
fsmithred | also, if it's the minimal live, I could post a link to one I made that does boot uefi | 17:28 |
crashoverride | with some UEFI implementations, the EFI program isn't found. | 17:28 |
crashoverride | because it is not in the standard, specififd path. | 17:28 |
crashoverride | specified* | 17:28 |
fsmithred | yeah, it's a real crap-shoot with motherboards. | 17:29 |
crashoverride | so, more modern UEFI firmwares that "discover" the firmwares are booting fine | 17:29 |
crashoverride | fsmithred: it's easy to fix tho. | 17:29 |
crashoverride | just copy the grubx64.efi firmware | 17:29 |
crashoverride | with the correct name | 17:29 |
crashoverride | that's how I solved my problem. | 17:29 |
Batoeh | Sorry I should have been clear it is x86_64. Yeah I will try a few more things and if all else fails I will try that pxe install. I really appreciate the assistance. | 17:30 |
crashoverride | Batoeh: also, minimal-live? | 17:30 |
Batoeh | Yep! | 17:30 |
crashoverride | thanks! | 17:30 |
tuxd3v | me too :) | 17:30 |
fsmithred | cool | 17:30 |
crashoverride | tuxd3v: you too, what? :D | 17:30 |
Batoeh | They also appreciate the assistance. | 17:30 |
tuxd3v | hehehe | 17:30 |
crashoverride | ah | 17:30 |
fsmithred | tux is usually giving assistance | 17:31 |
crashoverride | I thought they also had problems to boot. | 17:31 |
tuxd3v | other subject not related :D | 17:31 |
Batoeh | It is a rich tapestry. I will let you all know how it goes. Thank you again. Have a wonderful day. | 17:32 |
crashoverride | thank you :) | 17:32 |
fsmithred | Batoeh, that iso won't boot uefi | 17:32 |
tuxd3v | I messed up in the wrong place :/ | 17:32 |
fsmithred | if you have another linux installed and you're d | 17:32 |
fsmithred | oh well | 17:32 |
crashoverride | :DD | 17:32 |
fsmithred | boot from grub command line, install without bootloader, use other linux to boot | 17:33 |
crashoverride | yeah, altho, using grub's rescue shell is no party. | 17:33 |
fsmithred | ew, not the rescue shell, just the regular grub prompt | 17:34 |
fsmithred | I still haven't figured out how to use the rescue prompt | 17:34 |
tuxd3v | good Easter for those who it apply :) | 17:34 |
onefang | Installing rEFInd is a good idea. Replacing grub with it even better. Though it's only for EFI. | 17:34 |
mason | The rescue prompt is unpleasant. | 17:34 |
mason | Another option: Linux includes a stub loader. Use the kernel directly and win. | 17:35 |
fsmithred | I haven't tried stub loader yet. I have used rEFInd and it's good. | 17:35 |
fsmithred | and easy | 17:35 |
* fluffywolf does not celebrate zombie jesus day | 17:36 | |
fsmithred | and in the repo now | 17:36 |
fsmithred | not even with rabbit ears? | 17:36 |
crashoverride | fsmithred: the rescue prompt is easier for debugging a grub install | 17:36 |
onefang | Scientists have found The FSM. | 17:36 |
crashoverride | fsmithred: because it is fully sequential, and imperative. | 17:36 |
fsmithred | I could never figure out what commands to use there | 17:37 |
onefang | And I better not say any more about that, coz #devuan-offtopic. | 17:37 |
crashoverride | use tab | 17:37 |
crashoverride | fsmithred: also, help | 17:37 |
mason | fluffywolf: You should join #devuan-offtopic as I wanted to reply there. | 17:37 |
fsmithred | you're talking about the grub prompt that actually has the word 'rescue' in it? | 17:37 |
crashoverride | no idea what word it has in it. | 17:38 |
crashoverride | v_v | 17:38 |
fsmithred | or just grub> | 17:38 |
crashoverride | I think that last one. | 17:38 |
fsmithred | regular grub prompt is wonderful | 17:38 |
crashoverride | aren't they the same thing? | 17:38 |
fsmithred | been using that for years | 17:38 |
fsmithred | nope | 17:38 |
fsmithred | there's one that says grub_rescue> | 17:38 |
fsmithred | it is hell | 17:38 |
crashoverride | the one you get when you press 'c' at the grub menu | 17:38 |
fsmithred | that's regular grub command line | 17:38 |
crashoverride | ok | 17:39 |
fsmithred | good stuff | 17:39 |
crashoverride | well, it is useful, as it is fully imperative/sequential, as I said | 17:39 |
crashoverride | so, remember when I wanted to determine what grub modules I should insmod? | 17:39 |
crashoverride | that was how I found out. | 17:39 |
crashoverride | also, it's cool because it can be scripted if you have a macro keyboard | 17:39 |
crashoverride | so you can edit the macro | 17:40 |
fsmithred | :) | 17:40 |
crashoverride | and see what happens | 17:40 |
crashoverride | python strings are a fucking chore. | 17:44 |
crashoverride | I'd argue it was less bad in python2. | 17:44 |
crashoverride | so basically, if your input is '\test', it will be understood as '\\test' or '\test' depending on the context and how you parse it. | 17:52 |
crashoverride | if your input is '\\test' is will be '\\test', unless it is '\\\\test' | 17:53 |
crashoverride | if your input is '\\\\test', it can also be parsed as '\test' in some cases. | 17:53 |
crashoverride | nice. | 17:53 |
crashoverride | so, with my current script, if I input '\test' or '\\test', I get errors, if I input '\\\test' or '\\\\test', I get the same behavior. | 17:55 |
crashoverride | feels like python has been implemented in shell. | 17:56 |
crashoverride | actually, no, my reporting is incorrect. I crash with '\\@' but '\\t' parses as a tab. | 17:58 |
EHeM | Issue actually looks like pkgmaster has had vhosts enabled and is giving 404 for <2-letter-country-code>.deb.devuan.org. | 18:01 |
onefang | <2-letter-country-code>.deb.devuan.org is deprecated, has been for along time. | 18:03 |
onefang | Though I plan on bringing it back, but just don't use it now. | 18:04 |
crashoverride | onefang: are you in charge of running the *.deb.devuan.org servers? | 18:05 |
EHeM | Keeping it live means your plan will be exposed to greater testing sooner... | 18:05 |
onefang | I'm the Devuan package mirror herder, yes. | 18:05 |
crashoverride | ok | 18:06 |
furrymcgee | there are tools for building devuan live images | 18:16 |
furrymcgee | what do think of an image to create a local debian mirror for pxe netinstall? | 18:17 |
fsmithred | I like the idea | 18:18 |
fsmithred | a local amprolla? I know aitor has done that, but he didn't make a live-iso of it. | 18:18 |
onefang | Or a script to setup a mirror the usual way, but leaves out the "tell onefang about it" step, coz it's not a public mirror? | 18:20 |
fsmithred | oh, that sounds easier | 18:21 |
mason | I like the local-coalescer idea because it'd give you a local Debian mirror. | 18:21 |
mason | Otherwise you get whatever geoip gives you after hitting the mirror | 18:21 |
furrymcgee | there will be a lot more attention to an offical netinstall-client and netinstall-server images | 18:30 |
Batoeh | Good afternoon. I was able to thankfully get Devuan booting. I am fairly sure the iwlwifi firmware that it is telling me I am missing is just the wifi driver instead of ethernet but it won't let me skip past it and manually putting the ucode files on the installer usb or on another usb as instructed online does not seem to work either. I am using the server iso now but I faced the same issue on the netinstall | 20:43 |
mason | fsmithred: ^ I thought it was possible to skip the iwlwifi warnings, but maybe I'm thinking of something else | 20:45 |
fsmithred | I have always skipped past that window | 20:46 |
Batoeh | Hmm you know what I am sorry hang tight. I skipped before and it just kept popping up but it just seems that I have to keep doing that. I will let you know. Sorry. | 20:46 |
mason | No worries. Tell us how it goes. | 20:47 |
Batoeh | hmm I had to fanagle the menu manually and like several tries later but I am booting into my devuan system :) was close to ripping out the wifi card til my install was done. | 20:52 |
alv | hi golinux | 21:20 |
alv | hi fsmithred | 21:21 |
alv | the dyne mail it's fixed | 21:22 |
fsmithred | thanks, alv | 21:24 |
alv | hi | 21:26 |
fsmithred | hi | 21:27 |
alv | how is going ?? | 21:27 |
fsmithred | ok here. | 21:27 |
fsmithred | quiet day. No Easter gatherings this year. | 21:27 |
alv | me too still chilling | 21:27 |
alv | no | 21:27 |
golinux | Thank you alv! | 21:31 |
golinux | Since I couldn't do email I decided to eat so been afk | 21:32 |
golinux | Was now able to login. :D | 21:34 |
fsmithred | I'm gonna go ahead and build the deepsea them package. There doesn't seem to be much activity on the bug reports and I'm tired of waiting. | 21:36 |
alv | great i'm going to eat right now | 21:47 |
alv | Happy Easter to All !!!! | 21:48 |
gnarface | same to you alv | 21:59 |
Batoeh | Everything is working good! I really appreciated the help I received earlier. Thank you. | 23:04 |
systemdlete | So I decided to just install beowulf from scratch. I don't understand why the dist-upgrade did not work, and I didn't want to waste any more time on that approach. But after spending an hour installing beowulf from scratch, it will not boot. I think this is a EFI issue, but I'm not sure. | 23:09 |
systemdlete | Install did install grub, which used to work for 2 or 3 OSs I had on the laptop, which is EFI not BIOS. I am probably confused as to exactly what the devuan beowulf installer did and maybe what I told it to do. | 23:11 |
systemdlete | I tried using a grub repair tool, but it was not able to repair it. It complains about a missing boot bios, which was there (otherwise, it would never have booted prior to the beowulf upgrade attempt). | 23:12 |
systemdlete | So now I have a laptop which cannot boot ANYTHING, let alone beowulf. I am willing to do this all over again, from scratch, but I want to know what I need to look out for when installing on EFI hardware | 23:13 |
systemdlete | It is a Cherry Trail based system. BTW, there is nothing on the laptop's disk that I need to save (all important matter is backed up). So I can blow the drive away if needed. | 23:14 |
systemdlete | I just don't want to repeat this process if it will fail again. | 23:14 |
systemdlete | Might refracta or starlinux be any better at properly detecting and configuring EFI? | 23:16 |
fsmithred | do you get a grub prompt? | 23:16 |
systemdlete | Yes. | 23:16 |
fsmithred | and yes, refracta will work better, but for a simple reason | 23:17 |
fsmithred | grub-efi-amd64-signed was removed | 23:17 |
fsmithred | you can do that in a chroot or after booting from grub commands | 23:17 |
fsmithred | set root=(hd0,gpt... | 23:17 |
systemdlete | right | 23:17 |
fsmithred | linux /vmlinuz ro root=/d | 23:17 |
fsmithred | you know it | 23:17 |
fsmithred | then remove the signed package. The unsigned will remain. | 23:18 |
fsmithred | I assume you don't need secure boot. | 23:18 |
systemdlete | grub-efi-... was removed from... refracta? or removed from devuan? | 23:18 |
fsmithred | from refracta | 23:18 |
fsmithred | well... | 23:18 |
fsmithred | maybe from the desktop-live devuan isos, too. I don't remember. | 23:18 |
fsmithred | but easier to remove it yourself now that the system is installed | 23:19 |
systemdlete | fsmithred: Did I miss some fine print about installing to EFI ? I know at one point it asks if additional space should be allocated, something like that... | 23:19 |
fsmithred | you used one of the installer isos or a live iso? | 23:20 |
systemdlete | I tried booting from grub by setting linux and initrd (but not root=) but it gives me the rescue prompt | 23:20 |
fsmithred | yeah, the prefix is wrong because the location of the bootloader is hard-coded | 23:21 |
fsmithred | in EFI/debian | 23:21 |
fsmithred | one fix is to put a grub.cfg there | 23:21 |
fsmithred | you need root= two times to boot | 23:22 |
systemdlete | I elected to use one file system, rather than splitting /boot off by itself | 23:22 |
fsmithred | once is: set root=(hd-something | 23:22 |
fsmithred | the other is on the linux line, root=/dev/sd-whatever for the root filesystem | 23:22 |
systemdlete | netinstall | 23:22 |
fsmithred | yeah, you get the signed package that way | 23:23 |
systemdlete | So, let me understand this: I want the UNsigned package, not the signed one? That sounds... idk, backwards? | 23:23 |
fsmithred | if you need a microsoft-approved signature, go for it | 23:24 |
systemdlete | More Debian-OOPS? | 23:24 |
fsmithred | no | 23:24 |
systemdlete | Oh... I see. | 23:24 |
fsmithred | debian has a key for secure boot, and that means the bootloader is found in EFI/debian | 23:24 |
fsmithred | if you need secure boot, you should make your own key | 23:25 |
systemdlete | No, I don't want secure boot. | 23:25 |
fsmithred | right | 23:25 |
fsmithred | just remove the signed package. | 23:25 |
fsmithred | you don't have to add anything | 23:25 |
systemdlete | ok, I have the laptop here. It may take me a few minutes to figure out how to get it to turn on again. Sometimes, it doesn't shut down completely, other times it takes holding down the power for 10 secs... sheesh. | 23:26 |
systemdlete | Have to hold it down 3 seconds, otherwise, it just displays the battery level and turns back off. | 23:27 |
systemdlete | seems like the hw boot is not consistent. | 23:27 |
systemdlete | When I do "ls" at the grub prompt, I see hd0 with about 7 partitions (which is right; I did that myself), followed by (hd1) (hd2). AFAIK, there is only one SSD in this machine. | 23:29 |
fsmithred | good | 23:29 |
systemdlete | so what are hd1 and hd2 then? | 23:29 |
systemdlete | no thumb drives are connected atm | 23:29 |
fsmithred | I don't know | 23:29 |
systemdlete | lol | 23:30 |
fsmithred | they shouldn't be there | 23:30 |
systemdlete | I think I agree. | 23:30 |
systemdlete | Strongly agree. | 23:30 |
fsmithred | set root=(hd1) and then ls | 23:30 |
fsmithred | just to see what it says | 23:30 |
systemdlete | ls (hd1) or ls (hd2) reports no known fs detected, then sector size, etc. Both are about 4MiB | 23:31 |
systemdlete | 4096.5KiB actually | 23:32 |
onefang | Laptops internal recovery "disks"? | 23:32 |
systemdlete | onefang: Bang! Probably. | 23:33 |
systemdlete | There was Windows originally. I blew it away. Or thought I did. | 23:33 |
systemdlete | But I've been using the laptop for over a year without incident. | 23:33 |
systemdlete | Running devuan that is | 23:33 |
fsmithred | set root again to make it right | 23:34 |
systemdlete | devuan ascii | 23:34 |
systemdlete | set root for hd0? | 23:34 |
fsmithred | make it =(hd0,gptX) | 23:34 |
fsmithred | where X is the right number | 23:34 |
fsmithred | for the root fs | 23:35 |
fsmithred | linux /vmlinuz ro root=/dev/sd<that same number> | 23:36 |
fsmithred | initrd /initrd.img | 23:36 |
fsmithred | boot | 23:36 |
systemdlete | "Begin: Running /scripts/local-block ... done." about 20-25x, then drops me into initramfs | 23:39 |
systemdlete | busybox shell | 23:39 |
systemdlete | (I've gotten to this point in the past, just fyi) | 23:40 |
systemdlete | also, thanks fsmithred, for the assist. I'm sorry I am unable to figure this out myself. | 23:41 |
fsmithred | boot the installer media, go Advanced, Rescue, and open a console on the installed system | 23:41 |
systemdlete | I'm still getting the hang of this EFI firmware stuff. After finally understanding grub, along comes EFI. | 23:41 |
fsmithred | then remove the grub siggned package | 23:42 |
systemdlete | ok. | 23:42 |
fsmithred | check out rodsboooks.com | 23:42 |
fsmithred | https://www.rodsbooks.com/efi-bootloaders/index.html | 23:42 |
systemdlete | I've actually read those pages -- very helpful! | 23:42 |
systemdlete | But I still haven't mastered the beast... obviously. :D | 23:43 |
systemdlete | I did try this rescue mode, but did not know about removing that package. | 23:44 |
systemdlete | I opened a console; the main screen kept prompting me for the ath10k driver | 23:48 |
systemdlete | but it can't find apt or apt-get | 23:48 |
systemdlete | or dpkg | 23:48 |
systemdlete | now it is prompting me for the wireless to use "during the installations process" uh, maybe just a harmless string, but another thing that should be fixed that probably never will be | 23:50 |
systemdlete | I didn't get this before because I was using ethernet dongle. | 23:50 |
systemdlete | but at least it DID find my wireless lan | 23:51 |
systemdlete | trying to connect now | 23:51 |
systemdlete | they're exchanging keys now. | 23:51 |
systemdlete | got network... | 23:54 |
systemdlete | waiting for chance to open a console somewhere | 23:54 |
systemdlete | after I select the root device, it tells me it appears that /boot is a separate device. As I mentioned, I used one partition for the whole shebang. | 23:56 |
systemdlete | It might be detecting the separate /boot partition from the ascii that was installed on this machine. | 23:56 |
systemdlete | I just checked /target/boot and it is correct. I will not mount that other paritition | 23:58 |
systemdlete | fsmithred: Signed package is gone. Reboot? | 23:59 |
systemdlete | or do I need to do something else? | 23:59 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!