| lru | TOFU + cert could be useful... that way you have, in theory, some protection against MITM, and should you ever get an error in the future, and the cert allowed a MITM, you get detection from TOFU | 00:04 |
|---|---|---|
| rrq | hmm does it (Trust On First Use) do anything more than hiding the authentication process from the end user? | 02:10 |
| rrq | sorry.. this is for #devuan-offtopic | 02:12 |
| Guest977 | Hi there, how do i set my nameserver to 9.9.9.9 permanently? | 02:19 |
| Guest977 | because whenever i restart my machine the default nameserver is back. | 02:20 |
| adhoc | have a look in /etc/resolv.conf | 02:21 |
| debdog | Guest977: add "nameserver 9.9.9.9" to /etc/resolv.conf | 02:21 |
| rrq | the DHCP protocol includes that the DHCP server suggests DNS service, and most DHCP client software accepts that and updates the local configuration | 02:21 |
| rrq | so configure your DHCP client to not do that | 02:22 |
| Guest977 | how do i configure DHCP client to be permanent nameserver 9.9.9.9? | 02:23 |
| fsmithred | If you're running network-manager it will mess with /etc/resolv.conf | 02:23 |
| fsmithred | I made mine immutable so it will not be changed unless I do it. | 02:23 |
| Guest977 | @fsmithred i just installed devuan | 02:23 |
| Guest977 | i didn't install a network manager | 02:24 |
| Guest977 | how do i make it immutable? | 02:24 |
| fsmithred | how did you install devuan? No desktop or something other than the default desktop? | 02:24 |
| fsmithred | or maybe offline install? | 02:24 |
| Guest977 | yes its the default desktop | 02:24 |
| Guest977 | net install | 02:25 |
| fsmithred | then network-manager should be there | 02:25 |
| fsmithred | and it usually leaves a note in resolv.conf saying that it was there. | 02:25 |
| Guest977 | how do i configure network manager to not reset the changes i made in resolv.conf? | 02:25 |
| fsmithred | chattr +i /etc/resolv.conf | 02:25 |
| fsmithred | chatter -i /etc/resolv.conf to undo it | 02:26 |
| Guest977 | thanks. will try | 02:26 |
| fsmithred | I don't know. Maybe in the gui for n-m you can set it and it will stay. | 02:26 |
| Guest977 | i see. | 02:26 |
| fsmithred | chattr not chatter | 02:31 |
| devuan_ | chattr: Operation not supported while reading flags on /etc/resolv.conf | 02:41 |
| devuan_ | can't make the changes to resolv.conf if the chattr command is not working | 02:43 |
| debdog | you have to be root, of course | 02:43 |
| gnarface | devuan_: it bears mentioning that filesystem issues aside, the more appropriate place to change your DNS server in your case would be in your DHCP server, (if allowed) which is most likely your home router/firewall/modem that your ISP gave you | 02:43 |
| devuan_ | @gnarface i see | 02:44 |
| gnarface | that has the added benefit that this change will propagate through system upgrades and to your other devices | 02:44 |
| freaxeh | use 1.1.1.1 as a secondary dns server | 02:45 |
| freaxeh | i've had 9.9.9.9 go down for me | 02:45 |
| gnarface | devuan_: but feel free to keep learning how to wedge network-manager real good, that's still valuable knowledge :) | 02:45 |
| devuan_ | thanks guys | 02:46 |
| onefang | That conversation about HTTPS vs HTTP for /etc/sources.list happened while I slept (was 1 AM here). I skimmed it, but didn't see the REAL reason. For deb.devuan.org we use HTTP only, coz that's the DNS-RR for package mirrors. Which means each time you ask, you get a different IP which points to a different mirror. Which means we can't share the HTTPS certificate. If you want to use HTTPS, pick a mirro | 04:23 |
| onefang | r that supports it. Alas the person that asked about it isn't here anymore. | 04:23 |
| * onefang goes back to weekend projects. | 04:26 | |
| amarsh04 | now running Linux kernel 6.11 | 04:26 |
| adhoc | are there 6.x kernels back ported to Chimaera ? | 04:34 |
| fluffywolf | hrmm, and distributing a cert that lists all the mirrors would be too insecure, right? | 04:40 |
| adhoc | how would you stop folks from copying and distributing them ? | 04:41 |
| fluffywolf | it'd need to be distributed privately, of course. | 04:41 |
| adhoc | and there would need to be actual DNS records for each of teh mirrors | 04:42 |
| adhoc | a star cert for each of subdomains; | 04:43 |
| adhoc | maybe something like; | 04:43 |
| adhoc | mirror.eu.devuan... | 04:43 |
| adhoc | mirror.ap.devuan... | 04:43 |
| adhoc | perhaps by region rather than by country | 04:43 |
| adhoc | and round robin mirrors by region ? | 04:44 |
| fluffywolf | the risk is that having the same private cert on a whole bunch of systems increases the odds of it no longer being private | 04:44 |
| adhoc | right | 04:44 |
| adhoc | same goes for any other distribution | 04:44 |
| adhoc | wonder what the NetBSD folks do ? | 04:45 |
| rwp | There is no advantage to using https for repositories. There are many disadvantages. Clocks must be set for one. Not having clocks set is a repeating problem people come here with. | 04:51 |
| onefang | #devuan-infra might be a better place to discuss this. And maybe on a day when the package mirror herder (me) isn't trying to just relax and enjoy himself. B-) | 04:53 |
| fluffywolf | I don't know if the clock argument is valid, since it already won't work until they set the clock. lol | 04:56 |
| onefang | So, my weekend activity was to install and try out a bunch of games on my Daedulas install. I have two joysticks, and I managed to find my SpaceMouse Pro. Using spacenavd and libspnav0 as the "driver" things that should work with the SpaceMouse do. Which is Blender and FreeCAD. | 08:59 |
| onefang | The joysticks, or at least one of them, works fine in games where you would expect them to. | 08:59 |
| onefang | But I can't use the SpaceMouse as a joystick. You would think this was an obvious use case, and it should even do that out of the box. spacenavd can even use a joystick as some sort of pretend SpaceMouse, though how isn't documented so I haven't tried it. | 09:01 |
| onefang | Several utilities can map joysticks to keys and mouse movements and such. | 09:01 |
| onefang | But nothing for SpaceMouse -> joystick. | 09:02 |
| onefang | Anyone got any clues? | 09:02 |
| onefang | BTW, the SpaceMouse Pro is the controls used in close ups of spaceship controls in the TV series The Expanse. I saw that after I bought it, was very cool spotting my new nerd toy on a SciFi show. Don't think I have stumbled across any space flying games that are 3D though. Would be a great use for it. | 09:16 |
| CueXXIII | what kind of events does spacenavd generate? does xev show any? or do you have to use a special api to get inputs from spacenavd? | 09:53 |
| CueXXIII | if you can somehow get events you could forward them to X with xev, althogh that still needs work | 09:54 |
| onefang | libspnav0 implements the proprietary protocol as a library. spacenavd provides that protocol as a server. | 09:55 |
| onefang | You can tell spacenav to send keys when buttons are pressed. The device has 14 buttons, some labelled Alt, Ctrrl, Esc, and Shift. Which I mapped to their respective keys. | 09:56 |
| onefang | The MENU button I'm undecided on. I can map it to the Menu key (my keyboard lacks that), which does indeed work. Or I can leave it unmapped, which then means it works by default in the SpaceMouse aware applications. Can't have both. lol | 09:58 |
| onefang | Spacenavd is badly documented. I may have to dive into the source code. There's at least one undocumented config option that I stumbled across, but that was only a mention that it existed, no actual details like what it's called or how to use it. | 10:17 |
| buZz | onefang: cool a spacemouse, i got one too :) | 12:39 |
| buZz | openscad has nice support for it , btw ;) | 12:39 |
| buZz | but never tried to game with it though | 12:40 |
| buZz | a game i'd imagine would love such inputdevice is Descent | 12:40 |
| unclouded | onefang: I don't know if it works the same for you, but I have to stop spacenavd before the spacemouse works at /dev/input/js0. | 12:59 |
| buZz | oh fancy | 13:06 |
| buZz | i got a 'spaceball 5000' , one of the first models with USB ;) | 13:06 |
| onefang | Ah, I was gonna try all the games without spacenavd running, just had not got around to that yet. | 13:22 |
| buZz | i think mine only makes a hiddev :) | 13:23 |
| unclouded | Mine makes a js0, but "jstest" returns all zeros until spacenavd is stopped, then the axes show values with jstest. | 13:30 |
| buZz | nice | 13:30 |
| onefang | Unplugged the joysticks and the SpaceMouse, stopped spacenavd. Plugged the SpaceMouse back in. /dev/input/js0 only appears when the joystick is plugged in. | 13:34 |
| onefang | jstest-gtk has never listed the SpaceMouse. | 13:35 |
| buZz | i think 'spacemouse' is just a collection of hw that changed its modus operandi a lot | 13:36 |
| onefang | There's a bunch with similar names from the same company. "SpaceMouse Pro" is the one I have. | 13:36 |
| buZz | logitech is the company ;) nowadays | 13:37 |
| onefang | 3Dconnexion was the company when I bought it, their web site is still there. Could be Logitech makes some things with similar names? | 13:38 |
| buZz | well, they bought the company | 13:38 |
| onefang | Ah. | 13:38 |
| buZz | iirc ages ago | 13:38 |
| onefang | I bought mine ages ago. lol | 13:39 |
| buZz | oh! | 13:39 |
| buZz | TIL | 13:39 |
| buZz | 3dconnexxion is a subsidiary of logitech since founding | 13:39 |
| buZz | 3Dconnexion was formed in September 2001 by Logitech, combining LogiCAD3D, based in Europe, and Labtec's 3D peripheral business, based in the United States. The two companies combined have over 20 years of experience in 3D input devices. LogiCAD3D's product, the Magellan controller, was used in fields such as automotive design and aerospace. A NASA project used a Magellan product to control a robot in space. | 13:39 |
| buZz | The SpaceBall also had a history in space, having been used to remotely drive the Sojourner robot on Mars.[2] | 13:39 |
| onefang | I had a SpaceBall long ago. | 13:40 |
| onefang | Guess I'll put up with having a joystick on my desk, and unplugging it before I go to sleep, coz the good one has 7 bright lights on it. Wanted to spend the day playing games, not battling my SpaceMouse. lol | 13:55 |
| buZz | onefang: why not a wireless gamepad? | 13:55 |
| buZz | i use a Nintendo Wii-U pro controller with a '8bitdo usb adapter' , gives me a /dev/input/js0 | 13:55 |
| debdog | cover bright lights with stickers? | 13:56 |
| onefang | Coz I want to use my SpaceMouse. | 13:56 |
| buZz | the controller switches off when not used | 13:56 |
| onefang | I have bluetack on lots of the bright lights, doesn't work on somethings. | 13:56 |
| buZz | i use black nailpolish or a solderiron ;) depending on how annoying the light is | 13:57 |
| debdog | wait.... seven.... is ith a Saitek one? | 13:57 |
| onefang | Yep, Saitek. lol | 13:57 |
| debdog | hehe, I feel your... eye pain | 13:57 |
| * onefang unplugs it, yawns, and heads to bed. | 14:13 | |
| Walex | I also had a SpaceBall long ago. | 16:04 |
| devuanconsumer | I upgraded 7 mining desktop computers with exactly the same hardware (cpu, motherboard, ram, psu, etc) and now 3 of them are consistently having random restarts after the chimaera to daedalus upgrades. I have all PC's updated and upgraded to the latest daedalus software. I am thinking this may actually be hardware related, possibly coincidental hardware failure not linked to the daedalus upgrades: psu or ram. I doubt its overheating issues. | 18:58 |
| devuanconsumer | cpu is ryzen 7950x | 18:59 |
| devuanconsumer | what are your thoughts? | 18:59 |
| djph | blame the RAM | 18:59 |
| devuanconsumer | damn. | 18:59 |
| djph | I mean, if it was linux, chances are that it'd be every one. | 18:59 |
| djph | Not a difficult test -- memtest86+ for a few hours and see how it checks out. Might even just be a simple thing like poor contact | 19:00 |
| devuanconsumer | I have tried resetting the bios and this is still occuring | 19:00 |
| devuanconsumer | Ok I will do the memtest86 and reseat the ram | 19:00 |
| devuanconsumer | if this doesn't work | 19:00 |
| devuanconsumer | I will reapply thermal paste to cpu | 19:00 |
| devuanconsumer | and work through replacing the ram and then the psu | 19:01 |
| rwp | 3 out of 7 is a difficult data point to assess. It is unlikely to be any single thing. I would be inclined to boot the previous stable kernel and see if the problem persists. | 19:01 |
| devuanconsumer | i did replace a failing powersupply but it wasn't having random restarts like this. The computer would simply turn off and the psu made weird grinding noises | 19:01 |
| devuanconsumer | yes, I agree. It has me perplexed as well, since 3 out of 7 is a high number lol | 19:02 |
| devuanconsumer | what i hate about this is that i will waste a lot of time solving the issue.. | 19:02 |
| rwp | Since booting the previous kernel is easy on the 3 failing systems and might smooth things out. If it does then the kernel is indicated. Though the root cause may still be marginal ram timing. | 19:02 |
| rwp | The newer Linux kernel may be running the hardware more likely to be marginal or more likely to fail a race condition. | 19:03 |
| devuanconsumer | would adjusting the ram frequency help? | 19:06 |
| devuanconsumer | maybe i can install kernal 6.10 and uninstall 6.25 | 19:07 |
| devuanconsumer | btw thank you guys. I always appreciate the premium support I get here and I love debian very much. | 19:11 |
| rwp | The problem with the Linux kernel crash is that unless something is logged then it's unknown what the root cause is triggering the crash. | 19:15 |
| rwp | Adjusting ram timings won't help if that's not the problem. But if it is the problem then it might. You might try underclocking and see if that gets a more stable system. | 19:16 |
| rwp | I am guessing that the syslog is not logging anything useful. Because it never does for me either when I am having crash problems. You might set up a remote system logger remote system console in order to hopefully capture some last gasp logging that would point to the problem. | 19:17 |
| rwp | I have also just generally disassembled everything and blown the accumulated dust out of everything and checked for fans not turning and whatnot hoping it is a heat problem that might be improved too. | 19:18 |
| gnarface | random restarts might be a power supply problem too... | 20:51 |
| onefang | rwp: I do that once a year at least, open it up and clean the dust off everything. Due to do that again in a fortnight. This time I'm gonna reseat the RAM as well, might kill my RAM gremlin. | 21:06 |
| fsmithred | when I had a cat, I did dust removal about once a month. | 21:10 |
| onefang | cat dust > /dev/null | 21:26 |
| unclouded | onefang: Sorry, I didn't realise that only some spacemice make a js0. Mine is a "spacemouse wireless", so maybe they switched to HID later on? | 22:18 |
| onefang | It's entirely possible that my upgrade to Daedulas broke things. Had not actually used it since before then. | 22:20 |
| onefang | Hmmm, wonder if there's a firmware update for it? | 22:21 |
| unclouded | I'm on deadalus though. `Bus 001 Device 002: ID 256f:c652 3Dconnexion Universal Receiver` | 22:22 |
| onefang | Hmmmm. hid-generic 0003:046D:C62B.00C1: input,hidraw6: USB HID v1.11 Multi-Axis Controller [3Dconnexion SpaceM] on usb-0000:47:00.3-4/input0 | 22:23 |
| onefang | Says it's HID when I plug it in. | 22:23 |
| unclouded | Sorry, I'm an idiot. That particular model might be HID, but not a HID joystick | 22:24 |
| onefang | Yeah, that'll be why it can send Alt, Esc, Ctrl, and Shift. But it does say "HID v1.11 Multi-Axis Controller" which is the bit that's eluding me so far. spacenavd gets the buttons mapped fine. | 22:25 |
| unclouded | I get a lot of output but including `hid-generic 0003:256F:C652.0013: input,hiddev0,hidraw0: USB HID v1.11 Multi-Axis Controller [3Dconnexion 3Dconnexion Universal Receiver] on usb-0000:0f:00.0-4/input0` | 22:25 |
| unclouded | And no mention of a joystick, yet input/js0 does turn up. | 22:26 |
| onefang | Over here js0 only turns up when I plug in an actual joystick, not the SpaceMouse. | 22:27 |
| onefang | Nothing else changes when I plug it in. | 22:27 |
| unclouded | Very confusing. We seem to be almost identical, even down to the receiver. Is it behaving differently based on the spacemouse itself? | 22:30 |
| onefang | That's why I was wondering if I can get a firmware update for it. It is old. | 22:31 |
| unclouded | It's a shame that Linux is still neglected for things beyond the HID standard. At least solaar is awesome. | 22:33 |
| onefang | What's solaar? | 22:33 |
| unclouded | If you have Logitech mice or keyboards, you must install it! It a UI to let you see the battery levels and to bind devices like in Windows. | 22:34 |
| onefang | I have a Logitech MX518 mouse. No battery. Binding the buttons on it to things? They are bound to the correct things out of the box. | 22:36 |
| unclouded | No, I mean binding a keyboard and mouse to the same receiver, or binding a mouse that you lost the RX for to a different RX. | 22:37 |
| onefang | Ah, wireless stuff. I avoid that. | 22:37 |
| gnarface | onefang: your missing js0 may be more of a difference in udev rules than hardware features, i'd try merging the working one with yours | 22:37 |
| gnarface | i seem to recall someone else's devices going missing on a upgrade to daedalus recently... | 22:39 |
| golinux | unclouded: Note that Daedalus is not "Dead" | 22:39 |
| gnarface | oh, wait, no they had updated ceres and it stopped working, but was previously working in daedalus, i think | 22:39 |
| unclouded | golinux: lol, thanks! | 22:40 |
| gnarface | but i'm pretty sure it just ended up being the result of some udev rule culling | 22:40 |
| onefang | /usr/lib/udev/rules.d does have less rules files in it. | 22:42 |
| onefang | "51-these-are-not-joysticks-rm.rules" lol | 22:43 |
| gnarface | wouldn't be the first case we've had of "nobody's using that! i'll just delete it to clean things up a bit..." | 22:44 |
| gnarface | lots of distro maintainers and kernel devs seem to think that since event* is a superset of js* we shouldn't need it anymore, completely ignoring the legacy software issue apparently out of spite | 22:45 |
| gnarface | but aiui anything that can support event* can support js* with a minor udev rule addition, simply because event* is a superset of js* | 22:46 |
| gnarface | maybe i'm wrong about that but that's been the experience here with my devices | 22:46 |
| onefang | It is turning up as event16. | 22:46 |
| gnarface | SUBSYSTEM=="input", KERNEL=="event[0-9]*", ATTRS{name}=="Microsoft X-Box 360 pad [0-9]*", ENV{ID_INPUT_JOYSTICK}="1", GROUP="steam", MODE="0660" | 22:47 |
| gnarface | SUBSYSTEM=="input", KERNEL=="js[0-9]*", ATTRS{name}=="Microsoft X-Box 360 pad [0-9]*", ENV{ID_INPUT_JOYSTICK}="1", GROUP="steam", MODE="0660" | 22:47 |
| gnarface | is it too confusing if just drop these 2 lines from my steam controller config? | 22:47 |
| gnarface | somewhere it's gotta have a matching like like the first one, you just have to re-add one of the same form like the second one | 22:48 |
| gnarface | the only difference you need is KERNEL=="js[0-9]*" instead of KERNEL=="event[0-9]*", and it shouldn't be a problem for 2 rules to match (my steam controller needs a minimum of 5 to make everything work) | 22:49 |
| gnarface | if you just replace both and put them in a file in /etc/udev/rules.d/ higher specificity might allow both to take precedence, i think... i didn't have to worry about that for my steam controller which was completely unrecognized | 22:50 |
| onefang | I'm still searching through the udev rules looking for anything relevant. | 22:51 |
| gnarface | it might be a super simple blanket type catch-all rule of very low specificity | 22:51 |
| gnarface | try grepping for the vendor id | 22:52 |
| onefang | Did that half an hour ago. lol | 22:52 |
| gnarface | oh | 22:52 |
| gnarface | eh, just make your own rule file and try to load it | 22:52 |
| gnarface | if you match on 3 or more fields i'm sure it'll override whatever generic junk is currently matching | 22:53 |
| gnarface | "oh all HID devices should work exactly the same" nice fantasy there, but reality is a harsh master | 22:53 |
| gnarface | make the file named something like 90-spacemouse-local.rules and it should act last | 22:54 |
| gnarface | i've had to do stuff like that for my roccat mouse, my steam controller and a few different Chinese USB adapters for proprietary Nintendo controllers | 22:55 |
| gnarface | nothing like playing Playstation games with a Wii Pro controller :) | 22:56 |
| onefang | I do still have my historically significant Playstation 3, so I have it's controller in a box over there somewhere. | 22:57 |
| onefang | Sooo, how do I let udev know it's rules changed? | 23:01 |
| gnarface | supposedly the PS3 controllers will pair up with bluetooth... haven't tried it | 23:01 |
| onefang | They do. | 23:01 |
| gnarface | in theory you should just have to issue a "/etc/init.d/udev reload" but i've had problems with it seemingly caching stuff and having to flush state manually a couple times with udevadm [something] | 23:01 |
| gnarface | you can check if it read the rules with udevadm [something else] (sorry, i forget what exactly) | 23:02 |
| gnarface | obviously the best check though is seeing /dev/input/js0 appear | 23:02 |
| onefang | Only there is no /etc/init.d/udev only /etc/init.d/eudev | 23:02 |
| gnarface | oh, yea, issue the reload command to that one then | 23:03 |
| onefang | Ah the rule I thought I found was for a slightly different one. | 23:04 |
| gnarface | so, i think the way to flush the rules without the init script is this: udevadm control --reload-rules && udevadm trigger | 23:05 |
| gnarface | try that if "/etc/init.d/eudev reload" doesn't work | 23:05 |
| gnarface | there's also some way to get it to dump the rule stack matching a given device, but i'm having trouble digging it up again, i should have kept better notes on this... | 23:06 |
| onefang | That would be useful. | 23:06 |
| gnarface | the other one i should have written down is the command that dumps the entire state tree of a given device, which is useful for finding field values to match against when composing new rules from scratch | 23:07 |
| gnarface | oh, this is it, i think | 23:08 |
| gnarface | try this: udevadm info -a /dev/input/event16 | 23:08 |
| gnarface | (keep in mind it might change from #16 on hotplug) | 23:08 |
| gnarface | that one should give you some fields to match against though | 23:10 |
| * onefang reads the copious output. | 23:10 | |
| gnarface | i think it also gives you all the fields for all the devices up the usb chain, which includes like your motherboard controller and stuff | 23:10 |
| gnarface | just skip to the block that has recognizeable "spacemouse" descriptive text | 23:10 |
| onefang | Yep, most of the output is for the PCI and USB busses. | 23:11 |
| gnarface | but one of the blocks should have ATTRS{product} field recognizeable as the spacemouse | 23:12 |
| onefang | ATTRS{name}=="3Dconnexion SpaceM" | 23:12 |
| gnarface | take both ATTRS{idProduct} and ATTRS{idVendor} from that block and match on them both, as the combination should be canonically unique | 23:13 |
| gnarface | two 4-digit hex fields | 23:13 |
| onefang | It doesn't have either of those. | 23:13 |
| gnarface | (you're free to match on the text fields too but i've seen them change with firmware updates) | 23:13 |
| gnarface | hmm, it should, by law... check the next block? | 23:13 |
| gnarface | my steam controller for example has ATTRS{name}=="Valve Software Steam Controller Keyboard" on one of the parent device blocks, but that's not the block you'd be looking for | 23:14 |
| onefang | Ah, a few parent devices up. lol | 23:14 |
| onefang | ATTRS{product}=="SpaceM" | 23:15 |
| gnarface | yea, sorry it's a mess. i'm sure if i knew this tool better i'd have been able to give you some more targeted dump | 23:15 |
| gnarface | and keep in mind it's possible for the device itself to physically show up as multiple different devices | 23:15 |
| onefang | So it goes event16 -> input217 (ATTRS{name}=="3Dconnexion SpaceM") -> hid-generic -> usbhid -> ATTRS{product}=="SpaceM" -> ATTRS{product}=="xHCI Host Controller" etc. | 23:17 |
| unclouded | Thanks for that udevadm invocation gnarface. Did you say there was one to show which rules match for a device? I could run that here if it helps. This is a fresh install of daedalus | 23:20 |
| unclouded | Someone suggested "udevadm test". Here's my output: https://pastebin.com/4s31Hatr | 23:26 |
| gnarface | i thought there was but i'm not sure | 23:27 |
| gnarface | pastebin has ads. :( use paste.debian.net or just /msg it to me if you want me to look | 23:28 |
| unclouded | Thanks, sorry about that. I remembered there was one, but searching for "paste debian" didn't find it. | 23:28 |
| gnarface | there's also termbin.com which seems popular, i'm starting to warm up to it | 23:29 |
| gnarface | pastebin just seems... sinister | 23:30 |
| onefang | It's left handed? | 23:30 |
| gnarface | no, like they're being knowingly leveraged by organized crime and it was part of the business plan up front | 23:30 |
| unclouded | OK, no ads: https://paste.debian.net/1329560/ | 23:31 |
| gnarface | oh wow, that's a lot... | 23:31 |
| gnarface | hmm, i wonder if there's a way to identify just the matching rules from this block? seems to be showing everything... | 23:32 |
| unclouded | Might it be hardware afterall? I see ".INPUT_CLASS=joystick" in the output. Does that refer to the USB input class, i.e. the actual descriptor? | 23:33 |
| gnarface | maybe starting at line 136? | 23:33 |
| onefang | When I do that' after reading the rules file I get - unable to open device '/sys/dev/input/event16' | 23:35 |
| onefang | udevadm test /dev/input/event16 | 23:35 |
| unclouded | Is it permissions? | 23:35 |
| gnarface | onefang: remember it might have moved to a different event# | 23:35 |
| onefang | There is no such file, it's not in /sys it's in /dev | 23:35 |
| gnarface | just do "ls -lht /dev/input" and it's probably the one at the top of the list | 23:36 |
| gnarface | whichever was created most recently | 23:36 |
| onefang | I'll say again /dev/input/event16 exists, but when I point that udevadm command at /dev/input/event16, it complains about /sys/dev/input/event16 which doesn't exist. | 23:38 |
| onefang | There is no /sys/dev/input even. | 23:38 |
| gnarface | onefang: 1) add udev rules, 2) reload udev rules 3) hotplug device | 23:40 |
| gnarface | 4) check /dev/input/ again | 23:40 |
| unclouded | I just used strace on my "udevadm test" command. It looks for "/sys/devices/..." but no attempt to use /sys/dev at all. | 23:41 |
| gnarface | hmm, at some point for my steam controller i had to add uinput to /etc/modules | 23:42 |
| gnarface | it seems it was loading automatically and then suddenly wasn't, or else it was suddenly needed where before it wasn't, not sure | 23:42 |
| gnarface | not sure if that's relevant here | 23:42 |
| gnarface | hmm | 23:45 |
| gnarface | onefang: show me the rule you're trying to add | 23:46 |
| onefang | I was trying to comment out two rules, but turned out that was for a different device. | 23:46 |
| gnarface | oh | 23:46 |
| onefang | Thanks for trying to help, but I've spent way too much time on this rabbit hole. I'll just put up with plugging in a joystick when I need it for a game. I should be getting some work done today instead, no longer weekend mode. | 23:48 |
| gnarface | onefang: oh, well suit yourself but i just remembered something that might help | 23:48 |
| onefang | OK, one last rabbit. lol | 23:49 |
| gnarface | i just recalled that i caught my steam controller spawning virtual devices that didn't have idProduct or idVendor specified on them, just the string "Microsoft X-Box 360 pad " suffixed with a # | 23:49 |
| gnarface | those are the two i pasted earlier, that's why i didn't have idProduct and idVendor in them | 23:49 |
| gnarface | so the initial instructions i gave you for which block to find in the udevadm output were wrong | 23:51 |
| gnarface | or might have been wrong anyway | 23:51 |
| gnarface | additionally, they don't show up unless Steam is running | 23:51 |
| gnarface | so you might have to load your spacemouse driver and userspace daemon or whatever first | 23:52 |
| onefang | I don't even have Steam installed, I much prefer open source games I can get from the Devuan repos. | 23:52 |
| gnarface | no no, i'm not saying you need Steam, i'm just saying that your device might also need something running in userspace for the relevant devices to even appear | 23:52 |
| onefang | So it might be a timing thing and a reboot might help? | 23:53 |
| gnarface | yea, whatever you were using before might work just might need to make sure it's actually running before you turn on and plug in the device | 23:54 |
| gnarface | with the right udev rules already loaded | 23:54 |
| gnarface | anyway, i understand the need to get back to work, but i'll try to help you write some udev rules for it at some other time if you want | 23:55 |
| onefang | When I boot, I don't want to have to plug in devices afterwards. | 23:55 |
| onefang | Thanks again. | 23:55 |
| gnarface | in a fair world you shouldn't have to reboot or hotplug devices for any of this to work, but like i said early on there seemed to be some finnicky "caching" behavior going on i had to get past | 23:56 |
| gnarface | when i was writing steam controller rules | 23:56 |
| gnarface | so there were a handful of things i did just to try to force it to respond at all, to figure out what if anything was happening | 23:57 |
| onefang | As I mentioned before, I'm scheduling a "pull my computer apart and clean everything" first thing next month. It'll get rebooted then anyway. | 23:57 |
| onefang | Also entirely possible my RAM gremlin is just getting in the way, which is why the cleaning will include reseating my RAM. | 23:59 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!