| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 00:11 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 00:11 | |
| - bkeys (QUIT: Remote host closed the connection) (~Thunderbi@static-198-54-130-101.cust.tzulo.com) | 00:54 | |
| + jacobk (~quassel@2603-8080-b200-7b02-77cb-6304-f9db-dda1.res6.spectrum.com) | 01:07 | |
| - jacobk (QUIT: Ping timeout: 246 seconds) (~quassel@2603-8080-b200-7b02-77cb-6304-f9db-dda1.res6.spectrum.com) | 01:14 | |
| - S0rin (QUIT: Ping timeout: 252 seconds) (~S0rin@user/s0rin) | 01:20 | |
| + S0rin (~S0rin@user/s0rin) | 01:22 | |
| vagrantc | ACTION eyes https://reform.debian.net | 01:47 |
|---|---|---|
| vagrantc | :) | 01:47 |
| dozens | ACTION eyes vagrantc | 01:48 |
| vagrantc | the gpg signed repository is a nice improvement :) | 01:49 |
| - mtm (QUIT: Ping timeout: 240 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 02:03 | |
| - mjw (QUIT: Ping timeout: 246 seconds) (~mjw@gnu.wildebeest.org) | 02:04 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 02:36 | |
| + jacobk (~quassel@47-186-122-163.dlls.tx.frontiernet.net) | 03:04 | |
| - nsc (QUIT: Ping timeout: 245 seconds) (~nicolas@95-96-142-46.pool.kielnet.net) | 03:08 | |
| + bkeys (~Thunderbi@static-198-54-130-101.cust.tzulo.com) | 03:08 | |
| + nsc (~nicolas@131-98-142-46.pool.kielnet.net) | 03:10 | |
| - jacobk (QUIT: Quit: No Ping reply in 180 seconds.) (~quassel@47-186-122-163.dlls.tx.frontiernet.net) | 03:18 | |
| + jacobk (~quassel@47-186-122-163.dlls.tx.frontiernet.net) | 03:20 | |
| - lifning (QUIT: Ping timeout: 246 seconds) (~lifning@user/lifning) | 03:23 | |
| + lifning (~lifning@user/lifning) | 03:25 | |
| - lifning (QUIT: Ping timeout: 246 seconds) (~lifning@user/lifning) | 03:36 | |
| + lifning (~lifning@user/lifning) | 03:36 | |
| - lifning (QUIT: Ping timeout: 240 seconds) (~lifning@user/lifning) | 03:52 | |
| + lifning (~lifning@user/lifning) | 03:54 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 04:09 | |
| josch | minute: also good news: since i've fixed the bending of my keyboard i have not yet experienced another problem with it. No more disconnects for a few seconds and all keys function perfectly. :) | 06:10 |
| - GNUmoon (QUIT: Remote host closed the connection) (~GNUmoon@gateway/tor-sasl/gnumoon) | 06:13 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 06:13 | |
| + natalie (~natalie@user/natalie) | 06:36 | |
| josch | sevan: thank you for your input yesterday. I am much more happy with this version: https://reform.debian.net/repo/ | 07:03 |
| - klardotsh (QUIT: Ping timeout: 246 seconds) (~klardotsh@98.97.115.75) | 07:47 | |
| josch | minute: the only additional module you added to /etc/initramfs-tools/modules for a311d was display_connector, right? | 07:49 |
| josch | minute: can you do an experiment for me and run | 07:50 |
| josch | mv /etc/initramfs-tools/modules /usr/share/initramfs-tools/modules.d/reform.conf | 07:50 |
| josch | if that works, then we can ship and update the module list via the reform-tools package instead of using the reform-system-image script | 07:50 |
| josch | rebuild the initramfs after this move and reboot | 07:51 |
| josch | i'm afraid of doing that because if it goes wrong i'll loose my display output :D | 07:51 |
| - GNUmoon (QUIT: Remote host closed the connection) (~GNUmoon@gateway/tor-sasl/gnumoon) | 07:58 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 07:58 | |
| - GNUmoon (QUIT: Remote host closed the connection) (~GNUmoon@gateway/tor-sasl/gnumoon) | 08:18 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 08:21 | |
| - Boostisbetter (QUIT: Ping timeout: 260 seconds) (4a410829d7@irc.cheogram.com) | 08:44 | |
| + mjw (~mjw@gnu.wildebeest.org) | 09:20 | |
| violet | is there a particular place the wifi antenna is supposed to go | 09:35 |
| violet | the one in mine is just kinda chilling on the usb ports | 09:35 |
| violet | but it has this peel (i think?) thats not peeled off that i assume has adhesive underneath it to stick it somewher or something | 09:36 |
| - ndufresne5 (QUIT: Quit: The Lounge - https://thelounge.chat) (~ndufresne@souvlakia.collabora.co.uk) | 09:45 | |
| + ndufresne5 (~ndufresne@souvlakia.collabora.co.uk) | 09:46 | |
| minute | violet: normally there is a piece of acrylic included to mount them | 09:47 |
| minute | violet: seen here in step 6 https://www.ifixit.com/Guide/MNT+Reform+WiFi+Card+Replacement/147329 | 09:48 |
| minute | josch: sure, i can try that later today! | 09:48 |
| - mjw (QUIT: Ping timeout: 245 seconds) (~mjw@gnu.wildebeest.org) | 10:00 | |
| * mark_ -> mjw | 10:00 | |
| - ajr (QUIT: Quit: Connection closed for inactivity) (uid609314@user/ajr) | 10:12 | |
| violet | oh i dont think i have that | 10:49 |
| violet | its just the mimo antenna kinda in there, no separate acrylic thing | 10:50 |
| violet | i could probably get someone to print one for me | 11:18 |
| + mark_ (~mjw@gnu.wildebeest.org) | 11:26 | |
| violet | gods i love the btter life ofthe reform | 12:22 |
| violet | and how the power suppply doesnt get super hot the way my x220 psu does | 12:22 |
| violet | need to go pre order the banana pi cm3 | 12:23 |
| violet | cm4 | 12:23 |
| minute | violet: thanks for the feedback! | 12:37 |
| minute | violet: did you just order and did you get a confirmation mail? just checking if our systems are in order | 12:38 |
| + Boostisbetter (4a410829d7@irc.cheogram.com) | 12:49 | |
| - mtm (QUIT: Ping timeout: 245 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 14:03 | |
| josch | minute: i tested the initramfs-tools/modules.d solution and it works \o/ | 14:21 |
| josch | MR here: https://source.mnt.re/reform/reform-tools/-/merge_requests/42 | 14:21 |
| minute | josch: oh nice! i have some modules to request to add, after thinning it out a bit more | 14:22 |
| josch | minute: okay, then maybe wait with merging until you have the final list | 14:26 |
| josch | just message me once you have it and i amend that commit | 14:26 |
| minute | josch: will totally do that | 14:41 |
| - mark_ (QUIT: Ping timeout: 264 seconds) (~mjw@gnu.wildebeest.org) | 14:43 | |
| minute | intel abandons NUC https://www.servethehome.com/intel-exiting-the-pc-business-as-it-stops-investment-in-the-intel-nuc/ | 15:02 |
| josch | we used to use intel nucs in our robots at work but now there are so many alternatives from other vendors with a very similar form factor (that we switched to because they are cheaper than intel nucs) that i don't think this news should worry anybody, no? | 15:05 |
| minute | no, it's just interesting to me how companies abandon these things abruptly | 15:09 |
| minute | remember intel edison etc? | 15:09 |
| - GNUmoon (QUIT: Remote host closed the connection) (~GNUmoon@gateway/tor-sasl/gnumoon) | 15:09 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 15:10 | |
| josch | hm... i guess by now i saw so many abandoned projects that i liked and thought were successful that i'm not even surprised anymore | 15:10 |
| amospalla[m] | This small form factor came to stay, now without Intel, but there are other brands. | 15:11 |
| amospalla[m] | Oh, I mean, not exactly this form factor, but similar sizes. | 15:13 |
| josch | unfortunately those other brands did not copy the exact sizes and screw positions of the NUC... :/ | 15:15 |
| minute | yeah, i was going to ask, no sign of a standard for that stuff yet, no? | 15:15 |
| amospalla[m] | I think the only standard was the NUC. Lot of products based on these will disappear | 15:17 |
| amospalla[m] | Like lots of fanless cases for NUCs. | 15:17 |
| minute | reform mentioned https://www.wired.co.uk/article/open-source-your-blender-to-fight-electronic-waste | 15:18 |
| - ec0 (QUIT: Ping timeout: 260 seconds) (~ec0@vps-446f4f39.vps.ovh.ca) | 15:18 | |
| + ec0 (~ec0@vps-446f4f39.vps.ovh.ca) | 15:19 | |
| minute | ah, typo in a311d cmdline, pci=pci_bus_perf but should be pci=pcie_bus_perf | 15:34 |
| sevan | josch: that's really good. Thanks for laying out the details. | 15:43 |
| sevan | josch: thanks for the default sources.list snippet too, I can now restore sources.list and retry the instructions :) | 15:44 |
| sevan | minute: re NUC, that's interesting since many folks said that the PCengines/APU going away doesn't matter since they rely on NUCs now | 15:45 |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 16:09 | |
| josch | TIL: there a standard for glass jar twist-off threads in europe of 82 mm o0 | 16:11 |
| josch | also, i found my next mixer ;) | 16:12 |
| josch | nice to see the reform mentioned in the same list with the fairphone | 16:12 |
| minute | ha | 16:14 |
| minute | on monday i met luca weiss from fairphone (in paris). to my surprise he showed me that fairphone actually publishes schematics for the FP4 | 16:14 |
| josch | after being disappointed by purism my partner got a fairphone and is 100% satisfied with it -- we'll see how repairable it is once it breaks :) | 16:16 |
| minute | i have an fp3+ (that i don't use anymore though), it was pretty repairable | 16:18 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 16:20 | |
| josch | vagrantc: after your idea from yesterday i read more initramfs-tools code and found out about /usr/share/initramfs-tools/modules.d/ | 16:21 |
| josch | vagrantc: it does exactly what you think it does -- it's just not used at all in Debian XD | 16:21 |
| vagrantc | oh nice | 16:22 |
| vagrantc | can it also be accessed from /etc/initramfs-tools/modules.d ? | 16:23 |
| vagrantc | although obviously, for packaging, /usr/share is better ... | 16:23 |
| josch | vagrantc: it cannot https://sources.debian.org/src/initramfs-tools/0.142/mkinitramfs/#L329 | 16:30 |
| + mtm- (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 16:31 | |
| josch | vagrantc: it is documented in initramfs.conf(5) | 16:31 |
| - mtm (QUIT: Ping timeout: 245 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 16:32 | |
| jacobk | Is there a good way to compare my current computer's performance to the MNT Pocket Reform's performance? My current laptop is from 2013 so I expected the Pocket Reform to be more performant, but some benchmarks online suggest it's not: | 17:11 |
| jacobk | current CPU: <https://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i5-3210M+%40+2.50GHz> | 17:11 |
| jacobk | Pocket Reform CPU(?): <https://www.cpubenchmark.net/cpu.php?cpu=ARM%20Cortex-A53%204%20core%2018> | 17:11 |
| josch | jacobk: I do not have a pocket but a normal reform. Is there a benchmark you'd like me to run? | 17:15 |
| jacobk | I don't know which benchmark | 17:17 |
| jacobk | Passmark itself is nonfree I think, so I haven't actually run it myself | 17:17 |
| josch | i think it's more useful to use at benchmarks that run exactly what one wants to run on the machine in the end | 17:18 |
| jacobk | "(sysbench single thread speed: 1474.79 vs 738.25, stress-ng matrixprod test, 10s: 410 bogo ops vs 270)" <https://mntre.com/media/reform_md/2022-07-25-ls1028a-status-update.html> | 17:18 |
| josch | for example "time it takes to compile linux" | 17:18 |
| jacobk | That's an interesting idea | 17:18 |
| sevan | https://beagleboard.org/beaglev-ahead | 17:20 |
| jacobk | Problably the place I most often run into slowdowns and freezes is when I have too many tabs open in Abrowser/Firefox | 17:21 |
| josch | how can i find out how many tabs i have open in my firefox? | 17:22 |
| jacobk | The problem with that is that it's hard to measure; I have 35 tabs open now, but most of those are pinned and I haven't looked at them in a long time, so the browser probably unloaded them. | 17:22 |
| jacobk | Also, each tab might use a different amount of processing, though I have JavaScript disabled in most places so that might make it more even. | 17:22 |
| josch | so i can tell you that with firefox it's not much fun to browse twitch or youtube -- so i use clapper to view videos from these platforms | 17:23 |
| jacobk | I use Invidious with JavaScript for YouTube currently | 17:23 |
| jacobk | What is clapper? | 17:24 |
| josch | a video player that can use the hardware decoder that the imx8mq and imx8mp come with | 17:25 |
| josch | the processor is too slow for 1080p@60 software decoding | 17:25 |
| jacobk | oh, interesting; so that's not really an indication that the processor is slow overall right, I mean software decoding is something I thought was usually pretty slow | 17:26 |
| josch | that's one of the reason why i think those benchmarks don't say much | 17:26 |
| josch | in the end i want to have answers to questions like "will i be able to watch the movies i have on the thing?" | 17:27 |
| jacobk | josch: What limitations do you run into when using the Reform? Is the answer to "Will I be able to...?" 'no' in some cases? | 17:34 |
| jacobk | I found a game once that I thought I wouldn't be able to run, but admittedly I didn't even try it because it recommended a dedicated GPU and I don't have that. <https://codeparade.itch.io/marblemarcher> | 17:37 |
| jacobk | I would guess that the Pocket Reform cannot run that game well, but then, if it did, that wouldn't really tell me much about its capabilities elsewhere. | 17:37 |
| jacobk | I'm worried that if I switch to a weaker computer I might notice limitations I wouldn't have even thought of. | 17:38 |
| jacobk | josch: What happens if you run `sysbench --test=cpu run` (while you have other programs open is fine)? | 17:53 |
| jacobk | I got `events per second: 761.02` | 17:53 |
| minute | jacobk: we have some benchmarks online that you can compare with, "7z b" and "glmark2-es2" | 18:05 |
| minute | jacobk: the fastest reform that you can currently get is the one with a311d, but it has max 4GB ram | 18:06 |
| minute | jacobk: the benchmark results are linked here under "benchmarks" https://shop.mntre.com/products/mnt-reform-cm4-processor-module-adapter?taxon_id=13 | 18:06 |
| minute | oh nice, bluetooth seems to work @ a311d | 18:11 |
| jacobk | ok cool, I ran `7z b`, and `glmark2-es2` is in progress | 18:21 |
| minute | jacobk: what's your 7z total score? | 18:21 |
| jacobk | `Tot: 299 2224 6634` | 18:22 |
| jacobk | not sure which of those is the right number | 18:22 |
| jacobk | but it does seem like my compression is about twice as fast as the Q | 18:22 |
| jacobk | the IMX8MQ, I mean | 18:22 |
| jacobk | I don't see any benchmarks for the IMX8MPlus though | 18:23 |
| jacobk | Not sure how different the NXP i.MX8M Plus is from the Q | 18:26 |
| jacobk | Ooh, glmark score is higher on Q than on my current laptop | 18:29 |
| jacobk | 7z b full result: https://bpa.st/3C3LE | 18:29 |
| jacobk | glmark2-es2 full result: https://bpa.st/BP56Q | 18:29 |
| jacobk | Also decompression is not that much slower on Q than on mine, like ~90% | 18:30 |
| jacobk | or, more like ~80% actually | 18:30 |
| josch | jacobk: I get: events per second: 738.85 | 18:31 |
| jacobk | Thank you, that's slower but not that much slower I guess; you have the NXP i.MX8MQ right josch? | 18:31 |
| josch | correct | 18:32 |
| jfred | The NUC news is disappointing but not entirely surprising given Intel's history - relying on them for anything besides CPUs feels risky | 18:35 |
| jfred | (As someone with a compute card and a few docks for it...) | 18:35 |
| josch | jacobk: the reform is my only personal computer (i only have another at work) so i do absolutely all personal computing things like web, email, videos, chat, software development and so on on the reform. With my own requirements I see the following limitations the the reform has: | 18:38 |
| josch | - no reliable suspend | 18:38 |
| josch | - too slow to emulate most proprietary x86 games | 18:39 |
| josch | sometimes a bit more speed would be nice (compiling the Debian linux kernel takes 6 hours so i do that over night) but the limitation of 4 GB ram never created any issues for me, for example | 18:40 |
| josch | oh another items is: | 18:40 |
| josch | - x265 playback with the hardware decoder shows some artifacts | 18:40 |
| minute | jacobk: very interesting scores, that's with an i5? | 18:40 |
| minute | a311d total 7z b score: 531 1715 8981 | 18:40 |
| josch | (so essentially i'm just watching h264 videos instead) | 18:40 |
| josch | jacobk: so my list of problems with the reform is short and the problems that do exist are not deal breakers for me | 18:42 |
| sigrid | not to compare, but 6h is linux kernel's own issue here | 18:42 |
| sigrid | 9front's kernel builds in less than 40s | 18:43 |
| minute | hehe | 18:43 |
| vkoskiv | Would it be possible to turn off a lot of unneeded drivers to speed up the linux compile time? | 18:43 |
| minute | vkoskiv: totally | 18:43 |
| vkoskiv | Since it's a reform-specific kernel either way, right? | 18:44 |
| josch | sigrid: it is also a Debian issue as Debian enables nearly everything that can be enabled | 18:44 |
| minute | well, it's not supposed to stay that way vkoskiv | 18:44 |
| josch | vkoskiv: yes | 18:44 |
| minute | but yeah, with debian you get support for all kinds of hardware | 18:44 |
| josch | so if you want to build a custom kernel, of course you can do so :) | 18:44 |
| vkoskiv | Anything you could reasonably plug in via USB should have drivers, but maybe stuff like SAS cards and other esoteric things could be disabled :D | 18:44 |
| sigrid | josch: even if there is no way to make it work with the reform? | 18:44 |
| josch | sigrid: no way to make what work with the reform? | 18:45 |
| jacobk | josch: What does "no reliable suspend" mean? That sounds like it could be a problem for me, although it seems like something that could be fixed with a software update | 18:45 |
| sigrid | what vkoskiv mentioned, as an example | 18:45 |
| josch | so we are re-using the Debian config for arm64 and many things that are probably never used on arm64 are already disabled | 18:46 |
| josch | this lead to the problem I had recently that I plugged in my xbox controller and it didn't work out-of-the box | 18:46 |
| jacobk | minute: I've never heard of a311d until now, and I don't see it in the shop? Would I have to buy that in addition to the base Pocket Reform? | 18:46 |
| josch | but since Debian is a binary distribution and thus the compilation happens on somebody else's computer, build time is usually not a problem | 18:47 |
| vkoskiv | Doesn't chromium ship their own xbox controller driver? :D | 18:47 |
| minute | jacobk: sorry, i didn't get that you were looking for pocket reform | 18:47 |
| josch | even more so, since building the kernel happens on amd64 boxes within 20 minutes via cross compilation | 18:47 |
| vkoskiv | jacobk: You can get it as a bundle with the rcm4 adapter | 18:47 |
| minute | vkoskiv: not pocket, only reform | 18:47 |
| sigrid | josch: ah ok, this makes sense. it reminded me that all usb drivers in 9front are userspace programs | 18:47 |
| vkoskiv | Ah, gotcha | 18:47 |
| minute | jacobk: sorry, a311d should work with pocket but is not really tested yet | 18:48 |
| minute | jacobk: for a bit more background, these are the module options we have atm (table in the middle of the page) https://mntre.com/modularity.html | 18:50 |
| jacobk | minute: Thank you, I was looking for that table earlier today | 18:50 |
| jacobk | Does anyone have the i.MX8MPlus yet? | 19:01 |
| minute | jacobk: only me as the developer of pocket, i guess | 19:02 |
| jacobk | oh, cool | 19:04 |
| jacobk | minute: Have you run the 7z b and glmark2-es2 benchmarks on it? | 19:05 |
| minute | jacobk: yep | 19:06 |
| jacobk | Are the results public? | 19:07 |
| minute | i can't find the numbers right now, but i can rerun the test later | 19:07 |
| minute | at 2ghz it's maybe 30% faster than the imx8mq | 19:07 |
| minute | but the gpu is slower! | 19:07 |
| jacobk | that seems odd | 19:08 |
| minute | yeah unfortunately nxp put fewer shader cores into this model | 19:08 |
| minute | running 7z b now on pocket w/ 2ghz | 19:10 |
| minute | jacobk: so the "Tot:" scores are 356 1469 5379 | 19:12 |
| minute | multicore for compressing is 3508, for decompressing 7249 | 19:12 |
| + ajr (uid609314@user/ajr) | 19:25 | |
| minute | hmm strange, the rtw88 driver from our new kernel build still has issues | 19:34 |
| + mark_ (~mjw@gnu.wildebeest.org) | 19:35 | |
| + klardotsh (~klardotsh@98.97.113.255) | 19:47 | |
| - mjw (QUIT: Killed (NickServ (GHOST command used by mark_!~mjw@gnu.wildebeest.org))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:51 | |
| * mark_ -> mjw | 19:51 | |
| josch | minute: for the 6.4 build i dropped the rtw88 patches because all of them seemed to have been merged | 19:51 |
| + mark_ (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:52 | |
| josch | minute: maybe i checked wrongly and not all of those are actually included? | 19:52 |
| minute | josch: omg | 19:52 |
| minute | let me check what was dropped | 19:52 |
| josch | i thought i went through everything but maybe i made a mistake | 19:53 |
| minute | hmmm | 19:53 |
| minute | josch: what about 0121-rtw88-rxbuffer.patch ? | 19:55 |
| minute | and 0120-20230514_martin_blumenstingl_wifi_rtw88_sdio_always_use_two_consecutive_bytes_for_word_operations.patch | 19:55 |
| josch | the latter is in 6.4 as cb0ddaaa5db09d7d216fcbf0e68779be223a1128 | 19:56 |
| josch | the former is missing! | 19:58 |
| minute | ahaaa! | 19:58 |
| minute | extremely needed :3 | 19:58 |
| minute | josch: could you add it back and push, please? | 19:59 |
| josch | will do! | 19:59 |
| minute | thx!! | 19:59 |
| josch | first getting baby to bed | 19:59 |
| josch | then i'll get to it :) | 19:59 |
| minute | oh sure | 19:59 |
| josch | my other baby never goes to bed | 20:00 |
| josch | (because suspend doesn't work :D) | 20:00 |
| minute | :D | 20:00 |
| minute | haha | 20:00 |
| minute | in the meantime i will attempt a local build | 20:01 |
| minute | argh > E: Package build dependencies not satisfied; skipping | 20:11 |
| minute | there's some kind of conflict between libgcc-s1 arm64 vs amd64 | 20:16 |
| minute | unsat-conflict: libgcc-s1:arm64 (!= 13.1.0-8) | 20:16 |
| minute | version: 13.1.0-8 vs version: 13.1.0-7 | 20:16 |
| minute | ah, this can be fixed by using those env vars from common.sh | 20:24 |
| minute | no, i jinxed it | 20:25 |
| minute | i think the issue is that experimental is now at 6.4.1 | 20:27 |
| josch | minute: no, the issue is that unstable is broken half the time | 20:29 |
| minute | josch: ah | 20:30 |
| josch | minute: what you see there is what is called a multi-arch version skew | 20:30 |
| vagrantc | josch: the subkey used to sign the reform.debian.net repository is not actually present in the debian-keyring package in bookworm | 20:30 |
| josch | it happens when you cross compile | 20:30 |
| minute | josch: oooh ok | 20:30 |
| josch | it does not happen when you build natively | 20:30 |
| josch | vagrantc: that is correct | 20:30 |
| vagrantc | josch: also the reform_bookworm.sources contains a revoked key as well | 20:30 |
| vagrantc | josch: made this all a little confusing to verify based on the instructions | 20:30 |
| josch | vagrantc: do you know the right export options not have it contain useless material? | 20:31 |
| vagrantc | josch: no | 20:31 |
| vagrantc | josch: i mean, it contains a revoked primary key ... not worrying about subkeys or whatnot | 20:31 |
| minute | josch: would it be possible to cross-build the experimental kernel on a stable chroot then? | 20:31 |
| josch | minute: yes | 20:32 |
| josch | minute: or use a testing chroot | 20:32 |
| minute | ok thanks, gonna try that. | 20:33 |
| josch | vagrantc: so the exclamation mark at the end of the gpg --export is supposed to only export that specific key and not the rest | 20:33 |
| josch | vagrantc: and it does export less than without the exclamation mark but still too much | 20:33 |
| vagrantc | seems wild that it exports an unrelated key ... or is the subkey in both primary keys? | 20:34 |
| josch | it should not | 20:35 |
| vagrantc | in, er ... tied to | 20:35 |
| josch | the revoked keys are my identities before i changed my name | 20:35 |
| vagrantc | well, importing the key from the reform_bookworm.sources includes two primary keys, one rsa4096 and one ed25519 | 20:36 |
| vagrantc | the latter revoked | 20:36 |
| + vagrantc_reform (~vagrant@97-120-4-51.ptld.qwest.net) | 20:36 | |
| vagrantc_reform | ACTION waves | 20:36 |
| josch | vagrantc: an ed25519 key? i have to investigate... | 20:37 |
| vagrantc_reform | specifically BF4F7448B23221A99784CAB2D6A87BAC5EB24CA9 | 20:37 |
| vagrantc_reform | had some small hopes that i could test drive the shiny new kernel from this repository, but appears not to be the case :) | 20:39 |
| josch | vagrantc: that repository does not have the shiny new kernel but the debian stable kernel (6.1) instead | 20:39 |
| vagrantc_reform | right, which makes sense. | 20:39 |
| vagrantc_reform | at any rate, with some complication, i managed to get a key that i am reasonably confident is tied to josch@debian.org ... and while not present in the debian-keyring package, updating it from the reform_bookworm.sources seems corect ... so am glad to have a repository that is gpg-signed with a trust path i can verify :) | 20:42 |
| josch | vagrantc: i uploaded the subkey to some keyservers as well -- that one should also have more signatures | 20:43 |
| vagrantc_reform | josch: maybe update the documentation on the website to at least reference the version of debian-keyring that contains the subkey? | 20:43 |
| vagrantc_reform | i could figure it out, but i know what i am doing somewhat :) | 20:44 |
| josch | vagrantc_reform: you are right but there is a reason i didn't announce reform.debian.net yet ;) | 20:44 |
| josch | i'll announce it after the next upload of debian-keyring which will include the right subkey | 20:44 |
| vagrantc_reform | josch: well happy to do some prerelease testing and bug reporting :) | 20:45 |
| josch | it's very much appreciated! | 20:45 |
| josch | minute: in ~20 minutes this pipeline will give you a kernel with rtw88-rxbuffer.patch enabled: https://source.mnt.re/reform/reform-debian-packages/-/pipelines/1092 | 20:46 |
| josch | i'm currently writing a script going through the other 16 patches and making sure that they are included in 6.4 | 20:46 |
| josch | i just found another patch that only got partially included... | 20:46 |
| minute | josch: ahhhh thanks... craving for working wifi | 20:51 |
| minute | josch: which one was that patch? | 20:51 |
| minute | josch: your job failed for the same reason that i had locally with unstable https://source.mnt.re/reform/reform-debian-packages/-/jobs/1572#L2269 | 20:52 |
| minute | josch: can confirm that the build works locally with "testing" as a chroot | 20:56 |
| josch | oh no i'm stupid XD | 20:58 |
| josch | of course it would fail the same way :) | 20:59 |
| minute | it's a trap | 20:59 |
| josch | re "which one was that patch?": I added this https://source.mnt.re/reform/reform-debian-packages/-/blob/rtw88/linux/patches/meson-g12b-bananapi-cm4-mnt-reform2/0026-rtw88-rxbuffer.patch | 21:00 |
| - vagrantc_reform (QUIT: Quit: leaving) (~vagrant@97-120-4-51.ptld.qwest.net) | 21:03 | |
| minute | josch: hmm but that was exactly the one that i was complaining about (that was missing) ^^ | 21:07 |
| minute | josch: which one was the other, partial one you mentioned? | 21:07 |
| josch | minute: found it! meson-venc-sync.patch.patch was already partially applied in 6.4 and thus the patch became shorter | 21:20 |
| josch | I checked all patches that used to be named 0*-rtw88-sdio-* and they were all included in 6.4 | 21:22 |
| josch | so the only rtw88 patch that i wrongly dropped was the one that did not follow that naming convention | 21:22 |
| minute | ahhh, thanks, i will look into it | 21:23 |
| minute | ok, wifi works now with the package from local build | 21:36 |
| minute | audio/wm8960 is broken but i'll figure that out another day (looks like i2c issue) | 21:37 |
| josch | vagrantc: thank you for pointing out that reform_bookworm.sources contained a revoked ed25519 key from 2021 | 22:15 |
| josch | vagrantc: I now refreshed reform_bookworm.sources with a shorter key that only contains the subkey and not the revoked key by importing it into a new gpghome and then removing the revoked key with --delete-key and then exporting it again... | 22:15 |
| josch | now i wonder if there is an easier way to do this... | 22:16 |
| - bkeys (QUIT: Remote host closed the connection) (~Thunderbi@static-198-54-130-101.cust.tzulo.com) | 22:19 | |
| vagrantc | josch: oh wow. heh. yeah. | 22:23 |
| vagrantc | :) | 22:23 |
| vagrantc | ACTION wanders off for an afternoon hike | 22:24 |
| vagrantc | ACTION waves | 22:24 |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 22:24 | |
| - mjw (QUIT: Remote host closed the connection) (~mjw@gnu.wildebeest.org) | 22:31 | |
| + mjw (~mjw@gnu.wildebeest.org) | 22:31 | |
| + yewscion (~yewscion@2601:547:a01:2900:5c15:78d2:b913:fc63) | 23:02 | |
| violet | minute: yeah confirmation came in fine | 23:38 |
| violet | i sorry for the delay i think i fell asleep after | 23:39 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!