- S0rin (QUIT: Ping timeout: 255 seconds) (~S0rin@user/s0rin) | 00:13 | |
+ S0rin (~S0rin@user/s0rin) | 00:16 | |
- jacobk (QUIT: Ping timeout: 255 seconds) (~quassel@129.110.242.224) | 00:21 | |
- mjw (QUIT: Quit: Leaving) (~mjw@gnu.wildebeest.org) | 00:29 | |
eery | The fact the A311D is so much more power efficient is pretty sweet, I seem to be getting similar battery life to the imx8 on 1100 mah cells | 00:30 |
---|---|---|
- vagrantc (QUIT: Ping timeout: 245 seconds) (~vagrant@2a00:79e1:abc:133:77d:8d2b:ed76:33d) | 00:36 | |
+ jacobk (~quassel@utdpat241106.utdallas.edu) | 01:02 | |
- S0rin (QUIT: Ping timeout: 252 seconds) (~S0rin@user/s0rin) | 01:24 | |
+ S0rin (~S0rin@user/s0rin) | 01:25 | |
- S0rin (QUIT: Ping timeout: 255 seconds) (~S0rin@user/s0rin) | 02:09 | |
- jacobk (QUIT: Ping timeout: 255 seconds) (~quassel@utdpat241106.utdallas.edu) | 02:13 | |
+ S0rin (~S0rin@user/s0rin) | 02:15 | |
+ jacobk (~quassel@utdpat242016.utdallas.edu) | 02:20 | |
- ajr (QUIT: Quit: Connection closed for inactivity) (uid609314@user/ajr) | 03:36 | |
- jacobk (QUIT: Ping timeout: 255 seconds) (~quassel@utdpat242016.utdallas.edu) | 05:16 | |
+ jacobk (~quassel@129.110.242.173) | 05:37 | |
- jacobk (QUIT: Ping timeout: 255 seconds) (~quassel@129.110.242.173) | 05:44 | |
+ jacobk (~quassel@64.189.201.150) | 05:58 | |
+ Boostisbetter (4a410829d7@irc.cheogram.com) | 06:26 | |
Boostisbetter | Hmmm I have at most 4 things on a workspace. And for that the manual horizontal and vertical tiling controls are enough for me, but i'll check that out | 06:27 |
+ vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 07:13 | |
vkoskiv | josch: so do I just time the last two commands in the README in reform-debian-packages? | 07:58 |
vkoskiv | Seems like the README is out of date, I don't see a build.sh in here | 08:04 |
vkoskiv | Oh, chdir, it's in linux/ | 08:04 |
vkoskiv | Hm, it errors out because there's no sources in sources.list. I added them in, but it's still not happy | 08:09 |
vkoskiv | Ah, need to run the chdist again probably | 08:10 |
vkoskiv | Now it's going, I think | 08:10 |
vkoskiv | unpacking | 08:10 |
vkoskiv | It blew up, ModuleNotFoundError: No module named 'jinja2' | 08:12 |
- pandora (QUIT: Quit: Connection closed for inactivity) (uid585533@id-585533.ilkley.irccloud.com) | 08:13 | |
+ deadjakk (~deadjakk@68.183.17.6) | 08:41 | |
- vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 09:19 | |
- buckket (QUIT: Remote host closed the connection) (~buckket@vps.buckket.org) | 09:28 | |
+ buckket (~buckket@vps.buckket.org) | 09:29 | |
josch | vkoskiv: you at least found one bug but maybe more | 09:52 |
josch | vkoskiv: do you have some time to help me identify them? | 09:53 |
- laumann__ (QUIT: Changing host) (~quassel@2a0a-e5c0-2-2-0-c8ff-fe68-bef1.loves.ipv6.at.ungleich.ch) | 10:00 | |
+ laumann__ (~quassel@user/laumann) | 10:00 | |
* laumann__ -> laumann | 10:00 | |
vkoskiv | Sure thing | 10:50 |
vkoskiv | Might not reply right away until after work though | 10:50 |
josch | vkoskiv: did you directly copypaste the commands from the README while being in the same directory as the README? | 10:54 |
vkoskiv | Typed manually, but I think they were the exact same | 10:55 |
vkoskiv | Let me pull up my cmd history, sec | 10:55 |
josch | ah there is indeed a bug | 10:56 |
vkoskiv | First I ran 'chdist create base', then I realised it put it into ~, so I deleted that, ran | 10:57 |
vkoskiv | 'chdist --data-dir=./chdist create base' | 10:57 |
vkoskiv | Then 'chdist --data-dir=./chdist apt-get base update | 10:58 |
vkoskiv | ' | 10:58 |
vkoskiv | Then I tried the build, I think it said I needed deb-src entries in the source list | 10:58 |
vkoskiv | I tried the build again, then realised I need to run that update command again | 10:59 |
josch | vkoskiv: it's a bit simpler | 10:59 |
josch | vkoskiv: can you just run git clean -fdx to get back to a clean state? | 10:59 |
vkoskiv | Then ran 'vkoskiv@reform:~$ time env --chdir=./linux DEBEMAIL='robot <reform@reform.repo>' BUILD_ARCH=arm64 HOST_ARCH=arm64 BASESUITE=unstable OURSUITE=reform VERSUFFIX=reform ./build.sh | 10:59 |
vkoskiv | ' | 10:59 |
vkoskiv | once more, and got that final module not found error | 10:59 |
vkoskiv | Sure | 10:59 |
vkoskiv | Done | 11:00 |
josch | vkoskiv: try this: | 11:00 |
josch | sh -xc '. ./setup.sh; cd linux; . ./build.sh' | 11:00 |
vkoskiv | I'll prepend 'time' to measure how long it takes? | 11:00 |
josch | you don't need to because sbuild tracks this for you in the log | 11:01 |
vkoskiv | Ah | 11:01 |
vkoskiv | sh: 24: ./setup.sh: reprepro: not found | 11:02 |
josch | you need to install that | 11:02 |
josch | that should probably also be part of the README | 11:02 |
josch | sec... | 11:02 |
vkoskiv | Yeah :D | 11:02 |
vkoskiv | cleaning once more, then trying again | 11:02 |
vkoskiv | Now it's doing stuff | 11:02 |
vkoskiv | Slow network, let me hook up ethernet and try again | 11:03 |
vkoskiv | there we go | 11:03 |
vkoskiv | unpacking linux | 11:04 |
vkoskiv | Same ModuleNotFoundError | 11:04 |
josch | vkoskiv: here is the list of packages you should have installed | 11:04 |
josch | curl debhelper debian-archive-keyring debian-keyring devscripts faketime git mmdebstrap pristine-tar python3 python3-debian python3-jinja2 quilt reprepro rsync sbuild uidmap | 11:04 |
josch | i'm putting those in the readme now | 11:05 |
vkoskiv | LOL @ this apt-bugs warning: | 11:06 |
vkoskiv | grave bugs of faketime (→ 0.9.10-2.1) <Forwarded> | 11:06 |
vkoskiv | b1 - #1032177 - faketime doesn't fake time (on i386) | 11:06 |
vkoskiv | Are you *sure* you want to install this? | 11:06 |
vkoskiv | Mm, given it's not 1988, I think I do. | 11:07 |
vkoskiv | Started a new run | 11:08 |
josch | vkoskiv: 1988? | 11:08 |
vkoskiv | (i386 was introduced in 1985) | 11:08 |
josch | and what happened in 1988? | 11:08 |
vkoskiv | Eh, I might have reasonably had that CPU by then? | 11:08 |
vkoskiv | If I wasn't what, -8 years old :D | 11:08 |
vkoskiv | More info for README: | 11:09 |
vkoskiv | User vkoskiv is not currently an effective member of group sbuild. Please run: | 11:09 |
josch | vkoskiv: these are the changes to README.md that i distilled from our conversation so far: https://paste.debian.net/hidden/548f089a/ | 11:09 |
vkoskiv | sudo sbuild-adduser vkoskiv | 11:09 |
josch | nono | 11:09 |
josch | vkoskiv: have you ever used sbuild before? | 11:09 |
vkoskiv | I don't think so | 11:09 |
vkoskiv | But the build stopped, sh: 33: No such process | 11:09 |
josch | vkoskiv: i have put in the README diff above a very small section on how to set it up initially | 11:09 |
vkoskiv | Okay, I'll set that up | 11:10 |
josch | should be very quick :) | 11:11 |
vkoskiv | I don't think I'll ever get used to linux copypaste | 11:12 |
josch | what is linux copypaste? | 11:12 |
vkoskiv | ctrl+shift+c/v in terminal, but nowhere else | 11:12 |
vkoskiv | I always just do it accidentally in firefox, and it then slowly brings up the debuffer instead of copying text | 11:12 |
josch | i never use that | 11:12 |
vkoskiv | Second command is missing ~? | 11:13 |
josch | i just use selecting to copy and middle mouse button to paste | 11:13 |
vkoskiv | That works better, thanks | 11:13 |
josch | vkoskiv: yes, a tilda is missing, thank you! fixed locally | 11:13 |
vkoskiv | What's it actually building under ~/.cache ? | 11:14 |
vkoskiv | Something related to sbuild? | 11:14 |
josch | yes, a chroot where it can build packages in | 11:14 |
josch | that way, the package build will run in a chroot environment | 11:14 |
josch | and not on your actual machine | 11:15 |
vkoskiv | Alright, I'll run one more git clean and then try running the build | 11:15 |
josch | nice, thank you! | 11:15 |
vkoskiv | https://pastebin.com/hU7iq6vw | 11:16 |
vkoskiv | It's still doing stuff though | 11:16 |
vkoskiv | Patching now | 11:16 |
vkoskiv | Now it's building linux, it says | 11:16 |
josch | vkoskiv: that message just means that the last time you ran it, the server process that was started in the background was not killed and still running | 11:17 |
josch | just make sure to manually kill it in ~6 hours | 11:17 |
vkoskiv | Curious, what does the build need a server for? | 11:17 |
josch | vkoskiv: it does not -- this is for the other packages that get built | 11:18 |
josch | this is why the error is not fatal | 11:18 |
josch | for the other packages, that http server is needed to serve the packages that were already built to the new builds | 11:18 |
josch | but to build linux you don't need that | 11:18 |
josch | it would be possible to start this without starting the http server but that would make the command longer as the default is to start it | 11:19 |
vkoskiv | unpacking linux again now :D | 11:19 |
vkoskiv | Applying patches again | 11:19 |
josch | it will work now | 11:19 |
josch | you can go back to doing other stuff :) | 11:19 |
vkoskiv | All this packaging infrastructure is wildly impressive to me, but I feel like I know 0.0001% of how it all works | 11:19 |
josch | vkoskiv: no worries, there is not a single person who knows more than 90% of it | 11:20 |
josch | it's a big mess thanks to the mentality in debian that everybody is free to do packaging the way they like | 11:20 |
vkoskiv | So far it's mostly been single threaded, only using all cores when decompressing stuff | 11:20 |
josch | it will use all cores once it does the actual compilation | 11:21 |
vkoskiv | My brain always goes into a twist when there are tons of shell scripts calling each other, and depending on state in env vars | 11:22 |
josch | you are not the only one. i think it's one of the reasons why we have 20 different ways of doing things | 11:22 |
vkoskiv | And now also cmake, meson, and a dozen others! | 11:22 |
josch | because a lot of people think they can do a better job making this whole thing much simpler | 11:23 |
vkoskiv | Is there some 'best practices' guide for how to make a .deb from a small C project that just has 'make install/remove'? | 11:23 |
josch | and then they get faced with reality and their software becomes the same noodle soup as the others :D | 11:23 |
josch | vkoskiv: yes | 11:23 |
vkoskiv | I see a lot of things that I could do, but not what I *should* do | 11:23 |
josch | vkoskiv: but do you really want to do it yourself? You can also throw it to me and I do it in a few minutes for you. | 11:24 |
vkoskiv | Sure, I'll just have to polish it up a bit more, it will be fully available once I'm happy with it | 11:24 |
josch | vkoskiv: what build system do you use? | 11:24 |
vkoskiv | make | 11:25 |
vkoskiv | (I think it depends on GNU extensions, though) | 11:25 |
josch | the problem with just using "make" is, that you then have to implement everything yourself | 11:25 |
josch | once somebody wants to use a different compiler or wants to cross compile or wants to change compiler flags or wants to change installation DEST, your makefile will either break or you have to re-implement things in your makefile that other tools already do for you | 11:26 |
vkoskiv | I think one can just do 'make CC=mycoolcc', no? | 11:26 |
vkoskiv | But I get what you're saying | 11:26 |
vkoskiv | Can't get very far with just a makefile | 11:27 |
vkoskiv | I even added PREFIX and DESTDIR, to appear professional! | 11:28 |
josch | nice :) | 11:28 |
vkoskiv | Since I'm specifically targeting the Reform system image, it's quite easy. | 11:28 |
vkoskiv | I depend only on packags that come with the system image | 11:29 |
vkoskiv | Now it's actually building, I see a bunch of cc1 in htp[ | 11:29 |
vkoskiv | htop* | 11:29 |
josch | true, but now think of the following: since this is reform-specific it might become part of the reform-tools package | 11:29 |
josch | and i was thinking about putting the reform-tools package into Debian proper in some shape or form -- once that happens, your code will get through all the QA tools that developers let loose on all the other 20k packages as well :) | 11:30 |
josch | what i'm saying is: once you publish code, people *will* start using it in weird and unintended ways :) | 11:30 |
vkoskiv | Sure, I get that, and I'm not opposed to switching to another build system when the makefile becomes a pain | 11:31 |
josch | yes, but no rush | 11:31 |
josch | thank you for your input | 11:31 |
josch | i'll update the README of reform-debian-packages now | 11:31 |
vkoskiv | Thank you for all the stuff you do! | 11:31 |
josch | nah i'm just scratching my own itch | 11:32 |
josch | i wouldn't do any of this if it would not benefit me ;) | 11:32 |
vkoskiv | We'll find out which one finishes up first - The linux build, or my work week (~16.30) | 11:32 |
josch | given the other a311d benchmarks i've seen it might be the linux build! | 11:33 |
josch | i've recently posted some linux build times... sec... | 11:33 |
josch | imx8mq: 347 minutes | 11:33 |
josch | ls1028a: 412 minutes | 11:33 |
josch | a311d is absolutely faster than imx8mq, so i think it will be done first :) | 11:34 |
josch | vkoskiv: does it utilize all six cores? | 11:34 |
vkoskiv | It does, yeah | 11:38 |
vkoskiv | 68C, the temps remain very reasonable | 11:38 |
- Boostisbetter (QUIT: Ping timeout: 258 seconds) (4a410829d7@irc.cheogram.com) | 12:02 | |
+ Boostisbetter (4a410829d7@irc.cheogram.com) | 12:03 | |
* mark_ -> mjw | 12:18 | |
vkoskiv | with the lid open, temps seem to have stabilized to around 74-76C | 13:09 |
vkoskiv | Though it is sitting on top of my desktop tower, that might skew it up a bit | 13:09 |
vkoskiv | cmake is nice because it shows targets built and total targets. Sort of a progress indication. | 13:10 |
vkoskiv | Linux builds are just a wall of text until it's done :D | 13:10 |
josch | vkoskiv: with cmake consider using the ninja generator instead of the default makefile genetaror | 13:11 |
josch | it is signifiantly faster | 13:11 |
vkoskiv | Yeah, I'm familiar with that | 13:17 |
vkoskiv | Though does the linux kernel have cmake? | 13:18 |
vkoskiv | I guess it's makefiles all the way down :D | 13:18 |
vkoskiv | Or rather, config scripts + make | 13:18 |
+ murphnj (~murph@user/murphnj) | 13:24 | |
josch | vkoskiv: yes, linux is only make and there is no progress bar | 13:29 |
vkoskiv | I wouldn't be surprised if there exists a perl script somewhere that tries to show a progress indication :D | 13:30 |
vkoskiv | josch: Finished at 2023-10-13T11:59:50Z | 14:14 |
vkoskiv | Build needed 02:41:55, 5624864k disk space | 14:14 |
vkoskiv | So 161 minutes, 55 seconds? | 14:17 |
josch | wow, that's a massive difference! | 14:22 |
vkoskiv | Yeah, it's so satisfying to see so many CPUs busy doing stuff | 14:22 |
vkoskiv | In my raytracer, c-ray, it can optionally show the image being rendered in real-time, with each CPU working on a small portion of the image. I can actually spot the two slightly slower A53 cores :D | 14:28 |
erle | josch https://mister-muffin.de/paste is 500 on upload? | 14:37 |
josch | echo hello |mm | 14:37 |
josch | works for me | 14:37 |
josch | erle: you are not exceeding the maximum size limit, right? | 14:38 |
erle | ; du -hs -- google-ci-cd-fail.png && pastebin <google-ci-cd-fail.png >[2] /dev/null |grep title | 14:39 |
erle | 48Kgoogle-ci-cd-fail.png | 14:39 |
erle | <head><title>500 Internal Server Error</title></head> | 14:39 |
erle | josch i believe not | 14:39 |
erle | ; whatis pastebin | 14:40 |
erle | fn pastebin {curl -F 'arg=<-' https://mister-muffin.de/paste} | 14:40 |
erle | josch do you need the image to reproduce? | 14:47 |
- Boostisbetter (QUIT: Ping timeout: 260 seconds) (4a410829d7@irc.cheogram.com) | 15:36 | |
+ Boostisbetter (4a410829d7@irc.cheogram.com) | 17:05 | |
+ mark_ (~mjw@gnu.wildebeest.org) | 17:08 | |
josch | erle: i'm unable to reproduce it, so yes please | 17:16 |
erle | josch http://daten.dieweltistgarnichtso.net/pics/screenshots/fail/google-ci-cd-fail.png | 17:20 |
sknebel | wat | 17:22 |
erle | well, i tried to educate myself on computing terms and google failed me | 17:24 |
erle | josch it works now, maybe look into your logs? | 17:25 |
erle | proof: https://mister-muffin.de/p/E-ES.png | 17:25 |
josch | erle: thanks for reporting! :) | 17:50 |
josch | vkoskiv: i'd like to post your linux compile-time benchmark as a follow-up to https://floss.social/@josch/111220297029781835 | 17:50 |
josch | vkoskiv: if you have a mastodon profile i can @-mention you if you like | 17:51 |
erle | josch so what was it? | 18:07 |
erle | (i mean, did you find out?) | 18:07 |
erle | josch also are you taking patches for your pastebin? i have ideas | 18:07 |
erle | (mostly mime type related) | 18:07 |
vkoskiv | josch: I don't have one yet, maybe it's time to make one? | 18:10 |
vkoskiv | I miss MNT posts | 18:10 |
josch | erle: yes, i'm taking patches | 18:11 |
josch | vkoskiv: then welcome to the club of people who started mastodon only because of MNT ;) | 18:11 |
- mark_ (QUIT: Ping timeout: 258 seconds) (~mjw@gnu.wildebeest.org) | 18:14 | |
erle | josch consider the following: https://mister-muffin.de/p/xSBn.diff | 18:19 |
- stites (QUIT: Ping timeout: 255 seconds) (~stites@130.44.147.204) | 18:49 | |
+ stites (~stites@2607:fb91:de1:5076:f63e:26cf:4622:d9ad) | 18:49 | |
- stites (QUIT: Read error: Connection reset by peer) (~stites@2607:fb91:de1:5076:f63e:26cf:4622:d9ad) | 18:57 | |
+ stites (~stites@130.44.147.204) | 18:57 | |
minute | currently building GZDoom | 19:00 |
- stites (QUIT: Ping timeout: 245 seconds) (~stites@130.44.147.204) | 19:03 | |
+ stites (~stites@2607:fb91:de1:5076:f63e:26cf:4622:d9ad) | 19:05 | |
- jacobk (QUIT: Ping timeout: 255 seconds) (~quassel@64.189.201.150) | 19:10 | |
+ jacobk (~quassel@utdpat242016.utdallas.edu) | 19:28 | |
- jacobk (QUIT: Ping timeout: 240 seconds) (~quassel@utdpat242016.utdallas.edu) | 19:34 | |
+ jacobk (~quassel@129.110.242.224) | 19:39 | |
- stites (QUIT: Read error: Connection reset by peer) (~stites@2607:fb91:de1:5076:f63e:26cf:4622:d9ad) | 19:40 | |
+ stites (~stites@130.44.147.204) | 19:40 | |
+ mark_ (~mjw@gnu.wildebeest.org) | 19:46 | |
- mark_ (QUIT: Ping timeout: 255 seconds) (~mjw@gnu.wildebeest.org) | 19:58 | |
sevan | minute: great, is this a benchmark? :) | 19:59 |
- cwebber (QUIT: Ping timeout: 258 seconds) (~user@user/cwebber) | 20:40 | |
+ cryptix (~cryptix@scuttlebutt/cryptix) | 20:48 | |
sevan | heh, regarding UPS, return parcel to Berlin is sitting in the Herne-Boernig warehouse, it gets a date bump every day around the same time (about 15 mins ago) | 20:48 |
+ mark_ (~mjw@gnu.wildebeest.org) | 21:36 | |
erle | minute if you are playing doom, you might enjoy this WAD: https://www.doomworld.com/files/file/1038-aliens-total-conversion/ | 21:42 |
Boostisbetter | erle: seems that link leads to a 404 | 21:45 |
Boostisbetter | minute: btw, so you think a future news update could mention that suspend is working on the Reform? I have practically perfect use of it on mine with 6.5. | 21:55 |
erle | Boostisbetter for me it does not, using firefox hmmm | 22:00 |
Boostisbetter | erle: just used the link on the phone and it worked | 22:02 |
- mjw (QUIT: Killed (NickServ (GHOST command used by mark_!~mjw@gnu.wildebeest.org))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 22:08 | |
* mark_ -> mjw | 22:09 | |
+ mark_ (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 22:09 | |
Boostisbetter | Also has anyone found a way to get Netflix working on the reform? | 22:31 |
josch | i doubt that drm works with hardware decoding, no? | 22:32 |
- jacobk (QUIT: Ping timeout: 240 seconds) (~quassel@129.110.242.224) | 22:41 | |
Boostisbetter | josch: no idea. I think the imx8 has enough computing power to handle it without decoding. | 22:45 |
+ jacobk (~quassel@utdpat241106.utdallas.edu) | 22:46 | |
josch | Boostisbetter: software decoding h264 on the imx8mq only works up to 720p though | 22:47 |
Boostisbetter | Would be fine with me. | 22:47 |
- stites (QUIT: Ping timeout: 252 seconds) (~stites@130.44.147.204) | 23:05 | |
+ stites (~stites@2607:fb91:ded:48e3:83d8:c2d:8ba4:feb4) | 23:06 | |
- jacobk (QUIT: Ping timeout: 272 seconds) (~quassel@utdpat241106.utdallas.edu) | 23:39 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!