| + erlehmann (~erlehmann@x59cc89d2.dyn.telefonica.de) | 00:29 | |
| m2 | mntmn thank you for the kernel, sending from reform on wifi now | 00:40 |
|---|---|---|
| mntmn | m2 oh cool that it works! | 00:41 |
| m2 | mntmn thank you for the kernel, sending from reform on wifi now | 00:41 |
| mntmn | m2 cool cool! got your message twice now :D | 00:42 |
| m2 | yea, i'm that excited :) j/k, i think there was a bit of lag and i didn't see the local echo so i repeated | 00:42 |
| m2 | only minor issues remain, but i think they're basic debian config things i'm unfamiliar with, unrelated to new custom hardware | 00:43 |
| m2 | 1) rather than wlan0 as the interface name, it's coming back as a long ugly 16 character serial number looking thing. | 00:44 |
| mntmn | yeah i think udev renames the device | 00:44 |
| mntmn | i’m not sure why they do that tbh | 00:45 |
| mntmn | maybe it’s possible to disable that “feature” | 00:45 |
| zakx | it's called predictable interface naming, a systemd feature | 00:45 |
| zakx | it has an actual, useful purpose: making sure that your network device is the one you expect it to be | 00:45 |
| m2 | 2) wicd doesn't find any wifi networks even when the interface is showing up w/ifconfig. so i configure the network manually in /etc/network/interfaces as described on debian.org, linked to from the page you sent | 00:46 |
| m2 | i wouldn't really notice the interface name, except that i had to type it out a few times in /etc/network/interfaces :) | 00:46 |
| mntmn | here are some ideas https://unix.stackexchange.com/a/400898 | 00:46 |
| mntmn | zakx hmm makes some sense | 00:46 |
| zakx | fwiw you can just boot with net.ifnames=1 | 00:46 |
| zakx | or # ln -s /dev/null /etc/udev/rules.d/80-net-setup-link.rules | 00:47 |
| m2 | when configured like this, wicd shows no connection, & theres a message on i3 statusbar that says "No WIFI" even when i'm connected | 00:48 |
| zakx | oh sorry, it's net.ifnames=0 of course | 00:48 |
| mntmn | m2, it’s possible that networkmanager works better for you than wicd | 00:48 |
| zakx | here's some docs https://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInterfaceNames/ | 00:48 |
| mntmn | so, you can try to install that and remove wicd | 00:49 |
| m2 | zakx thanks! yea especially if I was using multiple NICs at once on a machine i'd appreciate that feature | 00:49 |
| m2 | but for this laptop where I'll probably be grabbing whatever wifi stick i find laying around, i'd almost prefer if it treats every nic the same and connects to same network | 00:50 |
| m2 | mntmn, will give that a try | 00:51 |
| mntmn | wicd also has different backends but i’m not an expert regarding those | 00:52 |
| mntmn | called “drivers” iirc | 00:53 |
| mntmn | anyways i think NM will work. i had a similar problem with another user who was at my place and with NM it could actually find SSIDs | 00:55 |
| erlehmann | the predictable interface naming is a boon i think | 01:01 |
| erlehmann | it actually describes where to find the device | 01:01 |
| mntmn | in the best case yeah | 01:02 |
| swivel | did erlehmann just express appreciation for a feature of systemd? | 01:13 |
| erlehmann | swivel, the main problem with systemd is not features, but the way the developers behave. | 01:15 |
| m2 | i would really like the feature *if* i could remap the predictable naming to a shorter alias | 01:15 |
| erlehmann | i.e. saying that udev will be usable without systemd and then suddenly, 3 years later, claiming that it was always clear that this would be deprecated. | 01:15 |
| erlehmann | also ignoring user consent. | 01:15 |
| mntmn | yeah i guess it works better on pci etc buses | 01:15 |
| mntmn | m2, is it now named after the mac address? | 01:16 |
| m2 | nope, its more than just hex chars: wlx000f6008285b | 01:16 |
| zakx | yeah, that's the MAC | 01:17 |
| mntmn | yeah but it’s the mac | 01:17 |
| swivel | erlehmann: lennart can be annoying, but I wouldn't group all the systemd developers so generally. | 01:17 |
| zakx | >4. Names incorporating the interfaces's MAC address (example: enx78e7d1ea46da) | 01:17 |
| zakx | see the link to freedesktop.org I posted above | 01:17 |
| mntmn | i use a minipcie wifi card in reform and it’s named wlp0s1 or similar | 01:18 |
| mntmn | i guess they don’t have a short scheme for usb net devices | 01:18 |
| zakx | it wouldn't make sense to have one as usb is highly volatile | 01:19 |
| mntmn | yeah i guess so | 01:19 |
| zakx | the whole case of this is to make it predictable; USB isn't | 01:19 |
| mntmn | m2 any other pressing questions before i go to sleep? | 01:24 |
| m2 | no, thank you so much! none of this is pressing, it's totally usable at this point and there's actually a lot I want to explore, so even if I hit a few problems i can just play with something else | 01:25 |
| swivel | m2: I'm pretty sure you can control the renaming if you wish to | 01:25 |
| mntmn | m2, glad to help. in case you haven’t seen, there are some known issues listed here https://source.mntmn.com/MNT/reform/issues and here https://source.mntmn.com/MNT/reform/wiki/HardwareTricks | 01:29 |
| mntmn | n8n8 | 01:30 |
| m2 | hah, so far everything i've noticed is already on HardwareTricks page :) | 01:30 |
| m2 | well i haven't seen the usb port 0 issue, and i don't think the fans whiny compared to any other laptop | 01:31 |
| erlehmann | swivel, oh, it's not only lennart. but i do not want to talk about this. | 01:33 |
| swivel | erlehmann: well, I'm a systemd developer, and do not consider myself part of what you're complaining about. | 01:34 |
| swivel | I became involved to improve components of software that was shipping in my distro of choice, not to ignore users or otherwise upset the community | 01:35 |
| erlehmann | swivel, but you are aware of the “if you don't say no, we assume yes” dynamic that is going on, right? | 01:36 |
| erlehmann | i mean, kay sievers for example | 01:36 |
| swivel | kay sievers isn't even involved for years | 01:36 |
| erlehmann | that may be. but the point is that this (and lennarts) behaviour seems to be tolerated (because they produce usable code, i assume?) by the other devs. i personally would not work with such a team. but really, right now, i want to talk about something else. | 01:37 |
| swivel | i am often upset by lennart's arrogance and general personality, and despised systemd and how it seemed to be forced onto the community | 01:38 |
| swivel | *shrug* | 01:38 |
| erlehmann | it's probably not hard to be upset by that ^^ | 01:38 |
| erlehmann | so i noticed a side channel yesterday. apparently the USB port leaks computation | 01:38 |
| erlehmann | i already informed mntmn and he told me that REFORM is probably not the only machine where that happens | 01:39 |
| erlehmann | any pointers regarding exploitation? | 01:39 |
| specing | "leaks computation"?!? | 01:39 |
| erlehmann | i used a headset with a USB-powered bass amplifier. even when the amplifier was powered off, i could hear a faint noise. | 01:40 |
| erlehmann | if i turn the amplifier on, this is more noticeable | 01:40 |
| erlehmann | reads from /dev/urandom sound differently than reads from /dev/zero | 01:41 |
| erlehmann | the first is noice | 01:41 |
| erlehmann | the second one is a more-or-less uniform tone | 01:41 |
| erlehmann | also if i start xclock and give it a period of 1 second, i can hear a tick every one second | 01:41 |
| erlehmann | like i can hear X11 rendering it or something | 01:41 |
| erlehmann | noise | 01:42 |
| specing | swivel: you could produce lfsystemd | 01:42 |
| specing | for lennart-free systemd, like EFNet was for IRC :) | 01:42 |
| erlehmann | lennart is not the problem | 01:44 |
| erlehmann | seriously | 01:44 |
| erlehmann | i get the hate, he is not very nice to others. but a single person can rarely upset things so much. there is a whole project with specific priorities and a lack of respect for user choice. | 01:45 |
| erlehmann | that, much more than lennarts utterings, is my problem with sd. | 01:45 |
| swivel | erlehmann: I think you underappreciate how much of that resulted from the leadership | 01:46 |
| swivel | erlehmann: but also a lot of these things like the monorepo for everything, swallowing udev, appearing to disrespect choice while consuming everything, was just optimizing the development process with a complete awareness that they were attempting to do too much with too little resources and couldn't have inefficiencies in the development process like having myriad public interfaces to define and iterate | 01:48 |
| swivel | across disparate projects | 01:48 |
| swivel | the intention was not to disrespect the users, it was kind of a crisis mode development | 01:49 |
| specing | reads from /dev/urandom sound differently than reads from /dev/zero | 01:49 |
| specing | < erlehmann> reads from /dev/urandom sound differently than reads from /dev/zero | 01:50 |
| specing | well... one is pseudarandom and the other is a stream of zeroes, what did you expect? | 01:50 |
| erlehmann | swivel, one problem is telling people that everything will be fine and interoperable and then 2 or 3 years later saying “we always told you we would migrate this stuff to systemd and couple it tightly” | 01:54 |
| erlehmann | happened too often for my taste | 01:54 |
| erlehmann | with, like, every project sd swallowed | 01:54 |
| erlehmann | there are nicer ways of noticing people that you want to break compatibility | 01:55 |
| erlehmann | specing, regarding the reads. the problem is that i do “dd if=/dev/urandom of=/dev/null bs=1M” and hear noise OVER THE USB PORT | 01:55 |
| erlehmann | i can start a program and hear a faint sound. or press a key and hear a sound. | 01:56 |
| specing | by noise you mean power supply noise? | 01:56 |
| specing | what is making the sounds? | 01:56 |
| erlehmann | the headphones | 01:57 |
| erlehmann | i think i hear the fluctuations in power supply caused by computation | 01:57 |
| specing | someone forgot to add fast capacitors? i.e. only electrolytes present? | 01:57 |
| erlehmann | no idea? | 01:59 |
| specing | buy some tantalums and add them yourself? | 01:59 |
| specing | see if that fixes it | 02:00 |
| swivel | that'd be annoying without even considering any information leak / security risk potential | 02:01 |
| swivel | in the old days we had to take care to not put the sound board near the video board to prevent noise in the audio output | 02:02 |
| erlehmann | it would be great if anyone could help fixing this | 02:02 |
| erlehmann | i mean, any program having USB access is currently being able to spy :/ | 02:03 |
| m2 | erlehmann, i'm not sure if what you're hearing is a signal that's available digitally on the USB data lines. my guess is theres audio frequency noise on the 5v bus, which is used as the power supply for your amp too | 02:12 |
| m2 | it is a information leak, but requires physical access to the device, not programmatic access | 02:13 |
| swivel | try digitally recording it from the host | 02:13 |
| erlehmann | if i record it from the host, i have a feedback loop | 02:13 |
| erlehmann | because i hear the programs that execute on the host | 02:13 |
| erlehmann | i have to record externally, probably | 02:14 |
| swivel | uh | 02:14 |
| swivel | if you can successfully create that loop, you've demonstrated the leaked information is available digitally to the host | 02:14 |
| swivel | which is the point of the test | 02:14 |
| erlehmann | i see | 02:14 |
| erlehmann | i'm not a very clever man | 02:15 |
| m2 | and i'm only here because I couldn't set up my wifi :P | 02:15 |
| swivel | erlehmann: it's probably an interesting test to do: just record audio from the USB audio device, while doing things like cat random vs. null, etc. then play the recording back on something without this problem | 02:18 |
| erlehmann | :/ | 02:18 |
| - erlehmann (QUIT: Remote host closed the connection) (~erlehmann@x59cc89d2.dyn.telefonica.de) | 02:30 | |
| + erlehmann (~erlehmann@x59cc89d2.dyn.telefonica.de) | 02:30 | |
| - erlehmann (QUIT: Client Quit) (~erlehmann@x59cc89d2.dyn.telefonica.de) | 02:31 | |
| + erlehmann (~erlehmann@x59cc89d2.dyn.telefonica.de) | 02:31 | |
| - erlehmann (QUIT: Read error: Connection reset by peer) (~erlehmann@x59cc89d2.dyn.telefonica.de) | 03:01 | |
| + erlehmann (~erlehmann@x59cc89d2.dyn.telefonica.de) | 03:01 | |
| - erlehmann (QUIT: Remote host closed the connection) (~erlehmann@x59cc89d2.dyn.telefonica.de) | 03:05 | |
| + erlehmann (~erlehmann@x59cc89d2.dyn.telefonica.de) | 03:05 | |
| - erlehmann (QUIT: Read error: Connection reset by peer) (~erlehmann@x59cc89d2.dyn.telefonica.de) | 03:19 | |
| + Guest8203 (~erlehmann@x59cc89d2.dyn.telefonica.de) | 03:29 | |
| - Guest8203 (QUIT: Quit: Leaving) (~erlehmann@x59cc89d2.dyn.telefonica.de) | 03:36 | |
| + erlehmann_ (~erlehmann@x59cc89d2.dyn.telefonica.de) | 03:37 | |
| - erlehmann_ (QUIT: Client Quit) (~erlehmann@x59cc89d2.dyn.telefonica.de) | 03:39 | |
| + erlehmann_ (~erlehmann@x59cc89d2.dyn.telefonica.de) | 03:40 | |
| - erlehmann_ (QUIT: Remote host closed the connection) (~erlehmann@x59cc89d2.dyn.telefonica.de) | 03:49 | |
| + erlehmann_ (~erlehmann@x59cc89d2.dyn.telefonica.de) | 03:50 | |
| - erlehmann_ (QUIT: Client Quit) (~erlehmann@x59cc89d2.dyn.telefonica.de) | 03:54 | |
| + erlehmann_ (~erlehmann@x59cc89d2.dyn.telefonica.de) | 04:14 | |
| * erlehmann_ -> erlehmann | 04:21 | |
| - erlehmann (QUIT: Remote host closed the connection) (~erlehmann@x59cc89d2.dyn.telefonica.de) | 04:39 | |
| + erlehmann (~erlehmann@x59cc89d2.dyn.telefonica.de) | 04:39 | |
| m2 | got the wifi working with NetworkManager. per the debian wiki, i had to add wifi.scan-rand-mac-address=no to NetworkManager.conf | 04:44 |
| - erlehmann (QUIT: Read error: Connection reset by peer) (~erlehmann@x59cc89d2.dyn.telefonica.de) | 04:44 | |
| m2 | after rebooting the interface name finally matched the MAC address (previously even though the name looked like "wlc"+macaddress, it was not the same as the mac reported by ifconfig | 04:47 |
| mntmn | m2, cool! | 09:38 |
| mntmn | erlehmann, to be able to “spy” you would also need a microphone or other ADC connected to the amplified audio. the information is analog in the 5V rail, not available digitally | 09:41 |
| mntmn | it is modulated onto the audio of your headphones by the amplifier which you feed with this 5V rail | 09:42 |
| mntmn | if you use an external power source to amplify the audio, you will not hear the noise | 09:43 |
| mntmn | https://www.maximintegrated.com/en/app-notes/index.mvp/id/4713 | 10:02 |
| - plomlompom (QUIT: Quit: WeeChat 1.0.1) (~plom@plomlompom.com) | 20:36 | |
| + plomlompom (~plom@play.plomlompom.com) | 20:36 | |
| + erlehmann (~erlehmann@x59cc8ba3.dyn.telefonica.de) | 21:03 | |
| - erlehmann (QUIT: Remote host closed the connection) (~erlehmann@x59cc8ba3.dyn.telefonica.de) | 22:00 | |
| + erlehmann (~erlehmann@x59cc8ba3.dyn.telefonica.de) | 22:00 | |
| - erlehmann (QUIT: Client Quit) (~erlehmann@x59cc8ba3.dyn.telefonica.de) | 22:03 | |
| + erlehmann (~erlehmann@x59cc894d.dyn.telefonica.de) | 23:32 | |
| - erlehmann (QUIT: Remote host closed the connection) (~erlehmann@x59cc894d.dyn.telefonica.de) | 23:44 | |
| + Guest78909 (~erlehmann@x59cc894d.dyn.telefonica.de) | 23:45 | |
| * Guest78909 -> erlehmann | 23:49 | |
| + esad (~textual@84-113-242-212.cable.dynamic.surfer.at) | 23:55 | |
| esad | hello everyone! I just got my reform today! | 23:56 |
| plomlompom | congratulations esad | 23:56 |
| mntmn | esad awesome | 23:56 |
| mntmn | welcome! | 23:56 |
| esad | it was so fun to unbox it and it really reminded me of unboxing my first 8-bit computer, with schematics in the user manual and everything :) | 23:56 |
| mntmn | haha cool! what was your first one? | 23:57 |
| esad | oh, it's quite obscure, it was called ORIC Nova | 23:57 |
| esad | I think it was made in slovenia under license | 23:58 |
| mntmn | ah special | 23:58 |
| esad | https://en.wikipedia.org/wiki/Oric | 23:58 |
| esad | the keycaps feel soooo nice! | 23:58 |
| esad | on reform, not on oric :) | 23:59 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!