The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

«‹›8 »

cruft: Oct 25, 2007

This is a question about ffmpeg. I don't know where else in the forums to ask it.

I've got this HEVC mkv file and I'm trying to convert it to H.264 using the h264_v4l2m2m codec. The following works fine:

code:

ffmpeg -i in.mkv -map 0:v -vf "scale=-2:'min(200,ih)',format=yuv420p" -b:v 3000k -c:v h264_v4l2m2m interim.mp4
ffmpeg -i in.mkv -c:a copy interim.ac3
ffmpeg -i in.mkv -c:s copy interim.srt
ffmpeg -i interim.mp4 -i interim.ac3 -i interim.srt -map 0:v -c:v copy -map 1:a -c:a copy -map 2:s -c:s copy out.mkv

This, however, fails:

code:

ffmpeg -i in.mkv -map 0:v -vf "scale=-2:'min(200,ih)',format=yuv420p" -b:v 3000k -c:v h264_v4l2m2m -map 0:a -c:a copy -map 0:s -c:s copy -num_capture_buffers 16 -num_output_buffers 32 out.mkv
Could not write header for output file #0 (incorrect codec parameters ?): Invalid data found when processing input
Error initializing output stream 0:0 --

This starts working, then spews a bunch of errors saying "All capture buffers returned to userspace. Increase num_capture_buffers to prevent device deadlock or dropped packets/frames", and eventually hangs. I can give it 128 capture buffers and it does the same thing :

code:

ffmpeg -y -i in.mkv -map 0:v -vf "scale=-2:'min(200,ih)',format=yuv420p" -b:v 3000k -c:v h264_v4l2m2m -map 0:a -c:a copy -map 0:s -c:s mov_text -num_capture_buffers 16 -num_output_buffers 32 out.mp4

This works fine:

code:

ffmpeg -y -i in.mkv -map 0:v -vf "scale=-2:'min(200,ih)',format=yuv420p" -b:v 3000k -c:v h264_v4l2m2m -map 0:a -c:a copy -num_capture_buffers 16 -num_output_buffers 32 out.mp4

This fails with the same "Invalid data found when processing input":

code:

ffmpeg -y -i in.mkv -map 0:v -vf "scale=-2:'min(200,ih)',format=yuv420p" -b:v 3000k -c:v h264_v4l2m2m -map 0:a -c:a copy -num_capture_buffers 16 -num_output_buffers 32 out.mkv

I know it's a long shot asking such a specific software question on SA but maybe there's an ffmpeg nerd here...

# ¿ Mar 1, 2023 00:25

Adbot: ADBOT LOVES YOU

# ¿ May 4, 2024 03:46

cruft: Oct 25, 2007

Rocko Bonaparte posted:

I'm asking this because I was enabling BPF and got a question about lowering the cost to the kernel's footprint by making it a module.

e: to answer your actual question, there is a complicated graph of when stuff can be a module. I don't know for certain about the one you're messing with, but it's entirely likely that some other feature you're configuring forces the BPF module back into the main kernel. Like, I know in the last 5 years, BPF has been overloaded to provide a whole mess of stuff. Maybe some network driver is using it for something, and since the kernel can't call into userspace being like "hey I would like to request you tell me where this module is in order for the network to come up", config marks BPF as Y instead of M.

The thing I edited out was basically "what are you hoping to achieve by recompiling the kernel?" From my perspective, it's equally likely that you have a legitimate situation where a kernel recompile can actually help, or that you're on step 23 of a funroll-loops-esque duck hunt based on a misunderstanding about what "footprint" and "cost" mean.

cruft fucked around with this message at 16:15 on Mar 4, 2023

# ¿ Mar 4, 2023 16:04

cruft: Oct 25, 2007

fletcher posted:

How come nut-scanner says snmp & xml are not available, even though I've installed those packages?

code:

root@d7a3fe62ee9c:~# nut-scanner --available
SNMP library not found. SNMP search disabled.
Neon library not found. XML search disabled.
IPMI library not found. IPMI search disabled.
OLDNUT
USB
EATON_SERIAL


root@d7a3fe62ee9c:~# apt list --installed | grep nut

WARNING: apt does not have a stable CLI interface. Use with caution in scripts.

debianutils/stable,now 4.11.2 amd64 [installed,automatic]
libgnutls30/stable-security,now 3.7.1-5+deb11u3 amd64 [installed,automatic]
libneon27-gnutls/stable,now 0.31.2-1 amd64 [installed,automatic]
libnutscan1/stable,now 2.7.4-13 amd64 [installed,automatic]
nut-client/stable,now 2.7.4-13 amd64 [installed]
nut-server/stable,now 2.7.4-13 amd64 [installed]
nut-snmp/stable,now 2.7.4-13 amd64 [installed]
nut-xml/stable,now 2.7.4-13 amd64 [installed]
nut/stable,now 2.7.4-13 all [installed]

Oh what the hell is this bullshit?

I spun up a debian container, installed whatever this is, and tried to run it. I got the same output. I installed libneon27, same output. I straced it: it appears to do a directory listing of the right directory, but it's not happy with the results.

I found a patch to NixOS meant to fix this very problem (link). It appears to set LD_LIBRARY_PATH, so I did that. Same output.

Next up is going to have to be looking through the source code to figure out what batshit crazy technique "nut" uses to locate dynamic libraries. I mean, this was standardize, I dunno, 20 years ago? But I guess these people think they have a better way, so one of us is going to have to reverse-engineer what their better way is so we can make it work like everything else.

# ¿ Mar 9, 2023 01:11

cruft: Oct 25, 2007

Okay OP, I'm knee deep in the source code here, in a function called get_libname_in_dir in common/common.c. It looks like it opens the directory, which I see in the strace, and then goes through each dirent trying to match it against the library name, which on anything but Windows should be libneon.so. That file exists in /usr/lib/x86_64-linux-gnu so I'm not sure what the problem is.

The source code says I should be able to pass -D to nut-scanner to get all these gorgeous debug messages, but it seems the version in Debian is too old and doesn't have that flag.

Does your version of nut-scanner let you pass -D in? That should make it reveal where it's trying to find libneon.so, which should help...

e:

quote:

I'm running this on debian bullseye.

Womp womp. Okay, I'll keep digging.

# ¿ Mar 9, 2023 01:32

cruft: Oct 25, 2007

I upgraded to nut-client from bookworm (debian testing) and it found the library.

So I'm going to throw in the towel and suggest you do the same thing. There's probably some stupid bug in the version in bullseye and I don't want to roll the source code back to whatever that is just to find out if I can jury-rig some workaround for some rando with a tiny av and a 2003 reg date

# ¿ Mar 9, 2023 01:39

cruft: Oct 25, 2007

I use emacs.

# ¿ Mar 9, 2023 14:52

cruft: Oct 25, 2007

I have a bunch of employees who use vi, which is fine, too.

Actually they swear up and down they only know vim, and then I find out they don't even know what "dd" does. Their "vim" usage technique is to press "i" and use the cursor keys and backspace A TON.

I don't understand how kids today reconcile themselves with such an incredibly painful and laborious method of editing text files.

So this leaves me, an emacs user, explaining to people how to do super basic vi stuff. I guess that makes me a vi expert by today's standards.

# ¿ Mar 9, 2023 15:23

cruft: Oct 25, 2007

Pablo Bluth posted:

I use nano...

Honestly, so do my aforementioned employees who claim to use vi.

Now that it lets you turn off word wrapping, nano is the right editor to teach new users.

# ¿ Mar 9, 2023 15:37

cruft: Oct 25, 2007

AlexDeGruven posted:

One thing that does piss me off is that nano is the default editor in Ubuntu and it's really difficult to change. The cognitive dissonance created when I type visudo and get dumped into nano...

It's not that difficult to change:

code:

export VISUAL=vi

Think of it as visudo, not visudo.

e: f,b

# ¿ Mar 9, 2023 15:44

cruft: Oct 25, 2007

AlexDeGruven posted:

That's the problem I ran into. I don't use Ubuntu much anyway, so the fact that it's now $VISUAL and not $EDITOR threw me.

See my previous statements about inertia, heh.

It's been $VISUAL since at least SunOS 4.2 (1982)...

If $VISUAL isn't set, it falls back to $EDITOR.

Why are there two? Well, in 1982, sometimes you logged in using one of these:

e: f, b AGAIN

`vi` started out as an extension to `ed` (or maybe `ex`, I don't recall). You'd pop into "visual mode" after starting, by typing `vi` at the `:` prompt.

cruft fucked around with this message at 15:51 on Mar 9, 2023

# ¿ Mar 9, 2023 15:49

cruft: Oct 25, 2007

xzzy posted:

The real travesty is hjkl navigation. I know arrow keys didn't exist at that point but I still despise the layout. At least make it ijkl so up/down makes sense.

But that positions your hand so that going northwest or southwest is a big long stretch.

cruft fucked around with this message at 15:55 on Mar 9, 2023

# ¿ Mar 9, 2023 15:52

cruft: Oct 25, 2007

Kibner posted:

In what way? It's just like WASD, but located on a different part of the board.

e: maybe this is just sarcasm that went over my head...

In Nethack, T and Y are NW and NE; B and N are SW and SE.

But the real reason it's hjkl is that ^H is left, ^J is down, ^K is "vertical tab", and ^L is "line feed". ^I is tab, and it would have been weird to make the "go right to the next tab stop" key mean "go up".

# ¿ Mar 9, 2023 16:25

cruft: Oct 25, 2007

fletcher posted:

Got a little further on my nut adventure....

nut-scanner found my device:
code:
root@02d9683e2ed8:~# nut-scanner -S --start_ip=192.168.1.83 --end_ip=192.168.1.83
Scanning SNMP bus.
[nutdev1]
        driver = "snmp-ups"
        port = "192.168.1.83"
        desc = "Eaton 5PX 1500 RT2U Netpack G2"
        mibs = "pw"
        community = "public"
I stuck that in my /etc/nut/ups.conf and then ran:
code:
upsdrvctl start
service nut-server start
Then I tried some of the commands from this page to no avail:
code:
root@02d9683e2ed8:~# upscmd -l nutdev1
Error: Connection failure: Cannot assign requested address
root@02d9683e2ed8:~# upsc nutdev1
Error: Connection failure: Cannot assign requested address
I thought this might have something to do with the fact that I'm trying all this in a docker container and maybe something about the way networking is handled was screwing it up, so I tried adding network_mode: "host" to my docker compose. Now it works!
code:
root@dev-vm:~# upsc nutdev1
Init SSL without certificate database
ambient.count: 0
battery.charge: 100
battery.charger.status: floating
...
I don't understand why I needed to do that though?

So I was actually sort of impressed with the general quality and organization of the nut codebase. But the code running on your UPS? LOL.

SNMP runs over UDP, and maybe it's doing broadcasts? If so, the Linux kernel may not be able to associate the outgoing packet with inbound replies.

I mean, who knows *. If the only thing you're doing in this container is running this client periodically, there's no additional exposure by giving it the host network.

(* I could help you figure this out if you're just dying to know, but it's going to be a long and winding road)

# ¿ Mar 10, 2023 15:20

cruft: Oct 25, 2007

JLaw posted:

Got a display settings question, figured I'd run it by y'all first before diving into reddit etc.

This is in Pop!_OS 22.04 w/ NVidia card/drivers.

For various reasons I want to be able to programmatically switch between a couple of display-settings arrangements, let's call them "profiles". I have a collection of xrandr commands to do that.

In one case I am setting a monitor to 144 Hz. When I do that, the visible scaling goes to 200%. I can open the settings GUI and set "Scale" back to 100%, and all is well, but obviously I want to avoid the GUI.

I don't know what is causing the "Scale" change, and I don't know how to change it back. The output of "xrandr -q --verbose" shows no difference before and after correcting the scaling through the GUI. Same for the output of "dconf dump /".

Anyone know what the "Scale" widget in the display settings GUI is affecting, in Pop!_OS or more generally Gnome?

Hm one other thing actually, "xrdb -query | grep Xft.dpi" shows 192 instead of 96 in the 200% scale case. But I don't know much about what's going on in X resource stuff. It looks like I could force it back to 96 with "xrdb -merge" but that doesn't seem to affect the current active state of things.

Hi, maintainer of 9wm here. Before I dive into trying to answer this, can you tell me what you get as output when you run:

code:

echo $WAYLAND_DISPLAY

# ¿ Mar 10, 2023 20:10

cruft: Oct 25, 2007

JLaw posted:

Yeah, not Wayland-ing here (yet... may try switching to that). That var is empty.

Okay, interesting. All the distros I use, Gnome = Wayland.

I was sort of hoping it was Wayland so there would be an avenue you hadn't checked yet. The fact that the DPI is halving is encouraging: it means X understands everything's gotten blown up.

I reckon the answer is "yes" but have you played with xrandr --scale?

Also, when you use the GUI control to scale back down to 100%, did you make sure you're still using the same frequency? Like, maybe for some weird reason this monitor halves the resolution at 144Hz?

Shooting in the dark here. I was really hoping it was a Weston thing.

# ¿ Mar 10, 2023 23:01

cruft: Oct 25, 2007

JLaw posted:

Yah, it doesn't fix whatever is going on. "--scale 1" leaves it doubled. I remember trying other scale values to see if I could "cancel out" whatever is going on, and it gets the overall icon and window sizes etc back down to the right dimensions but some things are weird like the rightclick menu popup being truncated.

Yeah using the GUI I can for sure get it to normal resolution/scaling at 144 Hz. Running some games in vsync verifies that they hit 144 FPS.

It's tantalizing to know that it's possible to get the correct state using the settings GUI... seems like there would have to be some way to do it with the CLI.

Well, if it wasn't quitting time on a Friday, I would dive into the code for the settings to see what's going on. But it is that.

Maybe I'll get a chance to grub around this weekend. Probably there's some weird quirk they found and coded up and (hopefully) left a comment about, that will hint you about what you need to do in this case. The DPI thing, yeah. That's making it seem within grasp.

# ¿ Mar 10, 2023 23:20

cruft: Oct 25, 2007

JLaw posted:

Thanks for the optimism. :-) I took a quick dive into the Gnome source before needing to head back to work... seems like maybe gnome-control-center is the relevant bit. In panels/display/cc-display-config-dbus.c it looks like it just sets logical_monitor->scale to the scale value. Then later in the config_apply function, the logical_monitor fields are used to construct a call to g_dbus_proxy_call_sync to do "ApplyMonitorsConfig". Can't chase that any further right now. Maybe I'll need to end up using gdbus invocations in my scripts, or something.

ugh, yeah, of course. Something else is listening on dbus for an ApplyMonitorsConfig call.

I'm sure there's something that lets you inspect dbus to see who that is. Have a nice time wading into this swamp!

# ¿ Mar 11, 2023 00:09

cruft: Oct 25, 2007

Subjunctive posted:

what the actual hell people

I'm new to SH/SC. Am I allowed to emptyquote?

# ¿ Mar 13, 2023 20:59

cruft: Oct 25, 2007

Twerk from Home posted:

Both at work and at home, I have machines with a ton of RAM but using network storage that is better at continuous reads rather than low latency IOPS. When I've needed random read access to a file, I've found that it's much faster to start a "cat /path/thefile > /dev/null" as I start accessing it to get the whole file into page cache. This feels like a terrible hack, is there a better way to tell the OS that I want a whole file preloaded into page cache to better consolidate reads?

That's actually very clever, and obvious in hindsight. I'm embarassed I hadn't already thought of it.

# ¿ Mar 14, 2023 01:45

cruft: Oct 25, 2007

VostokProgram posted:

If it's a really big file that might be slow. Maybe a C program that can mmap the file and then read one byte from each page would be better?

But it's got to pull the whole file over the network anyway, which is going to order(s) of magnitude slower than cat > /dev/null

# ¿ Mar 14, 2023 01:57

cruft: Oct 25, 2007

Nitrousoxide posted:

Fedora CoreOS is still being maintained and actively developed.

You're probably thinking of CoreOS Container Linux which was sunsetted ~3 years ago.

Flatcar is also a fork of the origional CoreOS Container Linux which is still being developed.

I'm experimenting with K3OS (which is the replacement for RancherOS) too, though I've not gotten it to use the cloud-init file correctly so far.

We use (and love) Flatcar for our clouds at work, AMA.

# ¿ Mar 14, 2023 16:45

cruft: Oct 25, 2007

I would blow Dane Cook posted:

How do you make the fonts not look like poo poo in chrome on arch?

This is a vague question. What looks lovely about them?

I'm going to hazard a guess that you're using Wayland on a HiDPI monitor and you haven't told Chrome to use Wayland output. This is making it draw fonts at like 72DPI or whatever, and then Wayland scales them up to the 280DPI of your monitor or whatever, and then it's all fuzzy.

Solution: tell Chrome to use Wayland. It's in flags somewhere. Be aware that you'll lose the ability to do screen sharing if you do this: you can only share Chrome tabs. Or at least that was the case back when I was still using Chrome.

# ¿ Mar 20, 2023 15:17

cruft: Oct 25, 2007

Tacos Al Pastor posted:

Does anyone know of a way to stop the script command from logging terminal output? I have a hung process and it seems outside that window I cant simply type "exit" to get script to end. Would like to know of an alternate way to get script to end and write to the file.

did you try kill -1?

# ¿ Mar 23, 2023 22:46

cruft: Oct 25, 2007

I'm so tired of Manjaro audio issues with the browser. Today I'm seeing 10 audio input options, but System default works. There's only one output option: System default, and it doesn't work.

Last month, Chrome got it wrong nearly 100% of the time and Firefox got it right nearly 100% of the time. Now it's switched.

Lately I've given up on the system audio settings dialog and I'm just routing everything to everything in Helvium.

# ¿ Mar 30, 2023 14:09

cruft: Oct 25, 2007

Yeah, things are in transition (still) and I think my two browsers just haven't caught up quite yet. I'm sure it's coming. It's just annoying to deal with right now.

PipeWire does seem like the Right Thing to me. PulseAudio was honestly a step in the right direction, and probably needed to happen, but, yeah, it had a lot of room for improvement.

# ¿ Mar 30, 2023 16:05

cruft: Oct 25, 2007

Klyith posted:

You might want to check what audio packages are actually installed / make sure you're on wireplumber as the session manager. Or even just reinstall the manjaro-pipewire metapackage.

I'm not sure what Manjaro is doing by default right now, because I manually shifted over to pure pipewire+wireplumber myself last year when the distro was still farting around with pulseaudio. Anyways I think they kinda bungled it a bit. And looking at their forums I'm starting to feel like the average manjaro user has lost the plot. Half the people on their boards seem to be complaining about things being updated, "pulseaudio works keep it", et cetera. Which is like, buddy this is supposed to be a rolling release distro, if you want stability you should be somewhere else.

OTOH personally I'd pin an issue like this on chrome over anything else. TBQH chrome is one of the most consistent problem citizens for me, particularly with audio & video stuff.

It's just like the browsers aren't getting the message when I choose a new audio output. And the just dizzying number of options doesn't help, either. Some of that is my fault, the USB docking station, and the HDMI monitor, and the builtin audio, and the bluetooth, and who knows what else, are just a whole lot of possibilities for "the right place to send audio". I assume there's some mechanism for communicating this, but the browser people aren't convinced it's stable yet, so they're trying to make guesses.

I don't know why it doesn't just provide an audio sink named "Default" that everything can use unless the user says differently, but whatever. It also seems to be exacerbated by Video chat apps trying to be clever.

Feels like the kind of mess I'm familiar with when there's an explosion of software ecological niches, honestly. In a year or two it'll settle down and we'll just have one thing that everybody hates but mostly works.

e: I am for sure using the pipewire sink but I suppose it wouldn't hurt to check this again. Thing is, I'm probably switching jobs soon, so I don't feel compelled to sink any time in figuring out what's wrong with this laptop I'm about to give back. Thanks, everyone, for listening to me complain

cruft fucked around with this message at 17:05 on Mar 30, 2023

# ¿ Mar 30, 2023 17:02

cruft: Oct 25, 2007

Keito posted:

Man am I glad I don't need to speak LDAP with my headphones...

Oh sure, first you have to bind (usually anonymously) to the correct domain (in this case your head), then you just do a query on "(|(cn=left)(cn=leftEar)(cn=oreille_gauche))" to retrieve the target (I think that covers all manufacturers but check your local specs for the correct canonical name) and then finally you can write a 16-bit sample to the returned distinguished name.

# ¿ Mar 30, 2023 21:48

cruft: Oct 25, 2007

Subjunctive posted:

code:
for i in *1.mp3; do
  mv $I `echo $i | sed &#8216;s/ 1\.mp3/.mp3/&#8216;`
done
Stick an `echo` in front of the `mv` the first time you run it to make sure I didn’t gently caress it up typing on my phone (also smart quotes are probably going to hurt)

E: ^^^ oh sure, in the last 20 years they’ve written a whole command for it

I gotta throw in on this one

code:


for i in *1.mp3; do echo mv "$i" "${i%1.mp3}.mp3"; done

If it looks right, run it again without the echo.

${i%str} and ${i#str} are Bourne shell things more Unix nerds ought to know about.

# ¿ Apr 5, 2023 04:49

cruft: Oct 25, 2007

Subjunctive posted:

Nice. Bourne or bash?

Bourne! So you can use this in ash, ksh, bash, and maybe zsh?

I take it as a point of pride that I hardly ever use bashisms. It's a nerdy flex, but it's the only one I have.

# ¿ Apr 5, 2023 04:59

cruft: Oct 25, 2007

VictualSquid posted:

Info pages, whatever happened to them.

They're still very much around. Did you know you can also run them through TeX and make a printed manual? Info pages are rad.

# ¿ Apr 5, 2023 13:36

cruft: Oct 25, 2007

ziasquinn posted:

When I got into Linux my friend warned me about posting for help anywhere because Linux users are literally the worst about that poo poo.

Linux users can be pretty bad, but Plan 9 users are the actual worst.

Anyway, if it's worth anything, I enjoy helping people with things they don't know, but I can't do anything about those who live to heap abuse on noobs. I believe this sentiment is the same thing that led Raph to create Freenode, actually.

# ¿ Apr 6, 2023 15:14

cruft: Oct 25, 2007

ziasquinn posted:

Based on just those two log snippets it sure looks like some kind of hardware problem. I'm inclined to suspect RAM, but maybe it's the firmware on the USB controller chip (which you probably can't modify) or... I mean, my work team tried to chase down a glitch like this recently and after about a month of replacing stuff we had to just give up and replace the whole motherboard. I'd say my team is above average competent at this sort of thing, but at some point you're going to need to decide that the time and energy required to figure out the problem is worth more than the money to just start over.

I guess I'd want to at least try a new OS first, though. Maybe some weird bit got flipped in your kernel code or some important daemon.

cruft fucked around with this message at 17:26 on Apr 6, 2023

# ¿ Apr 6, 2023 17:22

cruft: Oct 25, 2007

BattleMaster posted:

Never trust someone else's commands unless it's some grep awk text processing bullshit because gently caress coming up with that independently and it's not like it'll hurt anything if it doesn't work

code:

cat /etc/os-release | awk '{print "lol" > "/etc/shadow"}'

# ¿ Apr 6, 2023 19:16

cruft: Oct 25, 2007

RFC2324 posted:

Bsd absolutely has plenty of technical knowledge.

Its just trapped in the 70s, like his preferred OS

Whoa there. There's cool stuff happening in BSD (the OS) land. And while I'm kind of new to the thread, I haven't seen BSD (the poster) do anything worth all this crap getting shoveled onto them.

In fact BSD (the poster) appears to be one of the few people ITT with enough experience to be able to whip out an awk script to sort /etc/passwd by the 3rd GECOS field, and also tell you why it's called the GECOS field, which is something I at least appreciate.

# ¿ Apr 6, 2023 19:22

cruft: Oct 25, 2007

cruft posted:

code:

cat /etc/os-release | awk '{print "lol" > "/etc/shadow"}'

Been thinking about this all day and I think I have a better one that at least half of my 1st year hires would blithely run as root:

code:

## WARNING: THIS WILL DESTROY EVERYTHING IN /tmp
## WARNING: DO  NOT RUN THIS
wak -F 'ri ' -v a=m -v d=- '{b=FS; sub(/f/, a, FS); c=FS d b "/tmp"; print |c}' /proc/1/environ

I changed it from "/" to "/tmp", so if one of you idiots pastes that into a shell, at least you'll only shoot off part of your foot. It would look a lot less suspicious with "/" instead of "/tmp".

Anyway, I think this could nail a whole lot of people who don't know awk and assume it's "safe". Try it in an alpine container, it's scary fun.

e: also changed "awk" to "wak" just to make double-sure nobody unintentionally screws themselves over.

cruft fucked around with this message at 00:49 on Apr 7, 2023

# ¿ Apr 7, 2023 00:39

cruft: Oct 25, 2007

VostokProgram posted:

`cd -` goes back to the previous directory you were in btw

I can hardly wait until everybody discovers $_

# ¿ Apr 7, 2023 04:36

cruft: Oct 25, 2007

isaboo posted:

I've got 3 drives in my system, one older 500GB Samsung EVO ssd and 2 1TB nvme drives.

I use Arch as my daily, but now I need Windows (ugh) too. I'm planning on wiping the EVO and using that for Windows, and just choose which drive to boot from in the BIOS.

Normally I'd just disconnect the 2 nvme and install windows on the EVO but the nvme drives are under my big rear end GPU and heat sink so I'd rather not fiddle with removing them to access the drives.

Will windows play nice and allow me to select the 500GB EVO and install there without it touching the other drives?

If I were doing this, I would disconnect the Linux drive before installing Windows. It can't screw up a disconnected drive.

I just don't trust Windows or my ability to install it enough to make the risk of losing my Linux stuff less than the cost of pulling the cable under a GPU.

# ¿ Apr 7, 2023 23:20

cruft: Oct 25, 2007

zhar posted:

angrily deal with the consequences were i wrong.

I'm going to predict that this does come to pass.

# ¿ Apr 7, 2023 23:59

cruft: Oct 25, 2007

It's probably bad form to reply to the author's 60 lines of comments on my 3-line patch with "look, bozo," right?

# ¿ Apr 8, 2023 22:12

Adbot: ADBOT LOVES YOU

# ¿ May 4, 2024 03:46

cruft: Oct 25, 2007

Woolie Wool posted:

Very few native Linux programs of similar ages can do that from an existing binary, or even be compiled to run on a modern Linux system.

Let me tell you how much of a pain in the rear it's been to keep forward-porting Ken Arnold's Rogue to compile.

Like, just compile.

In fact, it's been a while, my repo is probably broken again. I should check on that.

# ¿ Apr 8, 2023 23:43

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

«‹›8 »