| + odnes (~odnes@109-178-137-160.pat.ren.cosmote.net) | 00:05 | |
| + alex4nder (~alexander@c-73-240-236-237.hsd1.or.comcast.net) | 00:10 | |
| - odnes (QUIT: Remote host closed the connection) (~odnes@109-178-137-160.pat.ren.cosmote.net) | 00:11 | |
| + odnes (~odnes@109-178-137-160.pat.ren.cosmote.net) | 00:11 | |
| + ephase (~ephase@2a01:e0a:482:9c50:1c87:6977:e9f1:f5fb) | 00:33 | |
| ephase | hi | 00:34 |
|---|---|---|
| - Asmadeus (QUIT: Quit: breaking this machine again!) (~asmadeus@240b:13:8c80:d300:e:98c:8000:d300) | 00:38 | |
| alex4nder | hi | 00:38 |
| mntmn | hi | 00:38 |
| ephase | mntmn: have you tried to run qutebrowser on the Reform? | 00:40 |
| mntmn | ephase: hm, just installed it, and it works | 00:42 |
| ephase | Damned! I have a GL error here on sway | 00:43 |
| mntmn | do you have a non standard setup? | 00:43 |
| mntmn | i am using sway as well | 00:43 |
| ephase | I have only debian packages, no compilation | 00:44 |
| mntmn | ah, you are using debian's mesa? | 00:45 |
| mntmn | and/or debian's sway? | 00:45 |
| ephase | yes for all | 00:45 |
| mntmn | wow ok... and 3d works? | 00:46 |
| mntmn | is it debian testing or which channel? | 00:46 |
| ephase | I don't have tried it for mow | 00:46 |
| ephase | yes, Debian testing | 00:46 |
| mntmn | you could try mesa / libgl* from experimental | 00:46 |
| mntmn | i just saw they have a much newer version there | 00:47 |
| mntmn | the one in testing is too old | 00:47 |
| mntmn | qutebrowser is surprisingly fast | 00:50 |
| ephase | Yes, i like this simple browser | 00:52 |
| mntmn | as i remember it didn't perform this well 1 or 2 years ago when i last tried it | 00:52 |
| technomancy | I wish I could use a non-mainstream browser without capitulating to webkit/blink monoculture | 00:52 |
| mntmn | well, most of the web is made for those engines unfortunately | 00:53 |
| technomancy | using them just feels like giving up and admitting defeat | 00:53 |
| technomancy | I guess it's inevitable since Mozilla laid off the servo team | 00:54 |
| ephase | technomancy: I'm agree, Firefox is my main browser, I wish qutebrowser with Firefox engine | 00:54 |
| mntmn | hmm, idk. webkit is still open source, but to each their own battles | 00:55 |
| technomancy | actually what I really want is for netsurf to somehow become usable | 00:55 |
| technomancy | like ... for them to finally acknowledge that keyboard shortcuts are a thing | 00:56 |
| mntmn | haha | 00:56 |
| mntmn | what do you think about gemini and such? | 00:56 |
| technomancy | I'd rather bring back HTML 2.0 | 00:57 |
| technomancy | I hate the idea of excluding normies intentionally | 00:57 |
| technomancy | seems to attract a certain sort of smug and self-satisfied minimalist. (not everyone, but enough to make me stay away) | 00:58 |
| ephase | mntmn: qutebrowser run with xcb | 00:58 |
| mntmn | ephase: xcb? | 00:59 |
| mntmn | technomancy: i agree mostly | 00:59 |
| ephase | mntmn: with xwayland | 00:59 |
| mntmn | ephase: ah you mean qt with x backend, not with wayland backend? | 01:00 |
| ephase | yes | 01:00 |
| mntmn | ephase: for me it works with wayland backend | 01:00 |
| mntmn | oh, so that makes it the first wayland-native webkit browser i can use | 01:00 |
| technomancy | I think there's a sense in which a lot of hackers give certain technologies (like webkit) a pass because they're open source but they're not open in any meaningful way that promotes user freedom. | 01:02 |
| technomancy | like, you can read the source, but it's far too complex for you to modify; if you make changes are you just going to maintain your own fork and somehow keep on top of security updates? | 01:03 |
| mntmn | ephase: thanks for reminding me of qutebrowser, it's really cool | 01:03 |
| mntmn | youtube also very performant | 01:03 |
| mntmn | technomancy: hmm i don't agree, it is possible to modify it, but the question is if that makes a lot of sense to do | 01:04 |
| mntmn | technomancy: it's a huge burden to maintain a browser | 01:04 |
| mntmn | technomancy: i mean qutebrowser etc show that it's possible to work with these things | 01:04 |
| mntmn | so qutebrowser is def. much faster than firefox | 01:06 |
| - bluerise (QUIT: Changing host) (~bluerise@pc19f8619.dip0.t-ipconnect.de) | 01:12 | |
| + bluerise (~bluerise@user/bluerise) | 01:12 | |
| technomancy | yeah, I'm talking about browser engines, not browsers | 01:13 |
| technomancy | if you want to make a fork of gecko you have to go the palemoon route, and I feel like that's been pretty thoroughly shown as a dead end | 01:14 |
| mntmn | the qutebrowser version in debian is pretty ancient according to the changelog though | 01:15 |
| mntmn | i always forget that i want to try to get darker/dimmer backlight levels than currently possible | 01:16 |
| technomancy | I tried luakit a few weeks ago and it was markedly slower than firefox. I wonder what the difference is. | 01:18 |
| ephase | I think I don't have any 3d acceleration... minetest is soooooo sloooooow | 01:21 |
| mntmn | ephase: glxinfo | 01:22 |
| mntmn | glxinfo | grep Vendor | 01:22 |
| mntmn | OpenGL vendor string: etnaviv | 01:22 |
| mntmn | OpenGL renderer string: Vivante GC7000 rev 6214 | 01:22 |
| mntmn | OpenGL version string: 2.1 Mesa 21.2.0-devel (git-3c390e2eb6) | 01:22 |
| mntmn | you should see something like this | 01:23 |
| ephase | I don't : OpenGL Vendor String : Mesa/X.org | 01:23 |
| mntmn | ephase: and renderer, OpenGL version? | 01:24 |
| mntmn | llvmpipe? | 01:24 |
| khm | I regard the major browser engines as closed because it's impossible to influence them without working for whatever corporation controls them | 01:25 |
| ephase | OpenGL renderer string: llvmpipe (LLVM 12.0.1_ | 01:25 |
| khm | and I say this as someone who has tried in the past | 01:25 |
| mntmn | ephase: yeah that's software rendering | 01:25 |
| technomancy | khm: I think honestly the easiest way is to get a job working at one of those companies | 01:26 |
| ephase | I forgot to install a package... | 01:26 |
| khm | technomancy: I think that's the *only* way | 01:26 |
| technomancy | well, bribery, threats, etc... =) | 01:26 |
| mntmn | ephase: which one? | 01:27 |
| ephase | got it, libdrm-etnaviv | 01:27 |
| khm | after years of interacting with the firefox project, I can claim two concrete results: 1) they ignored everything I ever said, including rewriting patches so the committer was @mozilla.org, and 2) they put me in the credits anyway | 01:27 |
| mntmn | ephase: and that changed things? | 01:27 |
| khm | truly bizarre | 01:27 |
| mntmn | khm: huh | 01:28 |
| ephase | mntmn: I install it... | 01:28 |
| technomancy | khm: I opened a bug with chromium like a decade ago about their tab-switching key bindings on linux being inconsistent with everything else and I asked them if they had the fastest JS engine in the world why do they have something as simple as key bindings hard-coded in C++ and the response was they didn't want to deal with support issues from stupid users who changed their key bindings without realizing it. | 01:28 |
| technomancy | you could tell they had a real contempt for their users | 01:29 |
| mntmn | ephase: it is interesting that sway works for you but GLX doesn't. i guess you should also do a test that is wayland native... let me think | 01:29 |
| khm | mozilla has never been mean-spirited about it, but firefox is Extremely Owned By Mozilla | 01:29 |
| khm | technomancy: finally, something I can agree with google about | 01:30 |
| mntmn | https://github.com/dv1/eglinfo | 01:30 |
| technomancy | anyway I tried recompiling chromium to change the tab switching keys like they recommended, and my laptop overheated after two and a half hours and shut itself off. | 01:30 |
| technomancy | "open source" yall | 01:30 |
| mntmn | ./waf configure --platform=wayland | 01:31 |
| khm | technomancy: surely you can just use gce as your build host! | 01:31 |
| mntmn | technomancy: yeah chromium needs a lot of resources to build, but i built it several times and also changed stuff in it for experiments | 01:31 |
| mntmn | i also built firefox but it needs >4GB ram to link | 01:32 |
| mntmn | ephase: i made a binary if interested: http://dump.mntmn.com/eglinfo | 01:33 |
| mntmn | if that also shows llvmpipe then something is very wrong | 01:34 |
| ephase | mntmn: thanks, i try this | 01:35 |
| mntmn | ephase: the next step would be to try kmscube -D /dev/dri/card1 in console to see if you get a spinning cube | 01:35 |
| technomancy | mntmn: btw since you seemed to have narrowed the cause of the standby issues to the GPU, do you have any further suggestions for how to debug further? | 01:36 |
| mntmn | technomancy: not really narrowed, that is more like a hunch... i guess to confirm it one should try to run the desktop without GPU | 01:36 |
| mntmn | i can try that real quick | 01:37 |
| technomancy | I feel like I'm slowly climbing maslow's hierarchy of computing. first I got emacs installed, then I got wifi fixed, then I got firefox keybindings working, now I can finally start to think about problems like "what if I didn't leave my computer running 100% of the day" | 01:37 |
| mntmn | technomancy: (weston --use-pixman makes a software rendered weston desktop) | 01:38 |
| mntmn | technomancy: haha! do you have emacs with nativecomp? that makes it a lot faster | 01:38 |
| - odnes (QUIT: Read error: No route to host) (~odnes@109-178-137-160.pat.ren.cosmote.net) | 01:39 | |
| technomancy | mntmn: oh, it's plenty fast for me; I haven't bothered looking into that | 01:39 |
| mntmn | ok | 01:39 |
| technomancy | ACTION looks into what weston is | 01:39 |
| mntmn | notmuch-emacs was too slow for me without | 01:39 |
| mntmn | technomancy: weston is a wayland demo compositor | 01:39 |
| technomancy | I'm coming from a thinkpad X301... 1.6GHz core 2 duo with 4GB | 01:40 |
| - alex4nder (QUIT: Quit: leaving) (~alexander@c-73-240-236-237.hsd1.or.comcast.net) | 01:41 | |
| swivel | nice, i lived with an x61s until a year or two ago... c2duo is so slow compared to anything nehalem or faster though | 01:41 |
| ephase | with eglinfo Vivante is detected | 01:42 |
| technomancy | swivel: I liked everything about my X61 except for how dim the screen was. I can go retro to a certain extent, but not further back than LED backlighting. | 01:42 |
| mntmn | ephase: that's good, so your problem is in Xwayland then | 01:42 |
| mntmn | ephase: it is not using glamor or glamor is not working | 01:42 |
| technomancy | I guess there's aftermarket mods for that nowadays if you can track down the right shadowy hacker collective in a Shenzhen back alley | 01:43 |
| ephase | Haaa kmscube said permission error mntmn | 01:46 |
| mntmn | is your user in the video group? | 01:47 |
| swivel | technomancy: i had the sxga+ led mod, but kept cracking plastic bits in the frame and palmrest etc.. moved on to x220 and now x230, totally spoiled by 16G ram and i7 perf now | 01:48 |
| ephase | mntmn: yes, and brightnessctl work | 01:48 |
| + Asmadeus (~asmadeus@240b:13:8c80:d300:e:98c:8000:d300) | 01:49 | |
| mntmn | ephase: hmm maybe try something silly like chmod 777 /dev/dri/card1 | 01:49 |
| mntmn | technomancy: i tested with software rendered weston and the first like 6 or 7 times wake worked, but then it froze again | 01:50 |
| mntmn | so i think it's not gpu | 01:51 |
| ephase | mntmn: doesn't work | 01:52 |
| mntmn | ephase: humm. sudo kmscube -D /dev/dri/card1 ? | 01:52 |
| mntmn | but anyway, GL works for you in wayland, just not in Xwayland | 01:52 |
| ephase | idem... | 01:53 |
| mntmn | technomancy: no real progress regarding wake. the next step would be to debug via UART and put lots of debug prints in the kernel in the responsible routines | 02:00 |
| mntmn | it works more often than not, so there is a lot of hope for sure | 02:04 |
| ephase | I'm really dumb.. tried to launch kmscube while running sway... | 02:07 |
| mntmn | ah lol | 02:07 |
| mntmn | it happens! | 02:08 |
| mntmn | ephase: maybe try to get xwayland from experimental... just a shot in the dark though | 02:09 |
| mntmn | normally it should just work™ but maybe debian has weird compile options | 02:09 |
| technomancy | shoot... I think I misplaced my uart cable in my last move | 02:10 |
| mntmn | technomancy: time to get a pack of fresh ones! | 02:10 |
| ephase | mntmn: I have a little spot on my reform screen since this night, it doesn't seems to be dead pixels but juste the plastic film | 02:22 |
| mntmn | ephase: meh! | 02:22 |
| ephase | It seems not quite important for now, I see it on black screen, and it wasn'n here two hours ago | 02:24 |
| ephase | For GLES, firefox say : No GPU detected via PCI | 02:28 |
| _Bnu | And it's not some "render" group issue like on the Pi 4, I guess? | 02:34 |
| mntmn | ephase: that's normal | 02:35 |
| mntmn | ephase: because the GPU is not on PCI. | 02:35 |
| ephase | mntmn: ok | 02:35 |
| mntmn | ephase: firefox is kind of tricky because it has multiple levels of using the GPU or not | 02:36 |
| mntmn | for example, webrender cannot work on reform because it uses opengl 3.2+ or so | 02:36 |
| mntmn | layer compositing can kind of work but it had glitches for me | 02:37 |
| mntmn | webkit based stuff or chromium work best currently | 02:37 |
| mntmn | or qtwebengine like qutebrowser apparently | 02:38 |
| mntmn | ephase: anyways, your problem is xwayland. | 02:38 |
| mntmn | ephase: i would suggest to use the newest version available, for example from experimental | 02:39 |
| mntmn | or use the source/build from reform-system-image... | 02:39 |
| mntmn | ephase: btw i documented some browser performance stuff and settings here now i want a | 02:40 |
| mntmn | ah, copy paste fail lol | 02:40 |
| mntmn | ephase: link https://community.mnt.re/t/recordings-of-browsers-chromium-93-chromium-90-firefox-89/339 | 02:40 |
| ephase | mntmn: I saw your post | 02:45 |
| ephase | :) | 02:45 |
| mntmn | ok ^^ | 02:45 |
| mntmn | ok gonna watch some ds9 and call it a day | 02:48 |
| mntmn | see ya around! | 02:48 |
| ephase | good night | 02:50 |
| khm | mntmn: https://i.imgur.com/PdyxD7J.png | 02:52 |
| - ephase (QUIT: Quit: WeeChat 3.2) (~ephase@2a01:e0a:482:9c50:1c87:6977:e9f1:f5fb) | 03:42 | |
| technomancy | how far thru ds9 are you? | 05:08 |
| lastebil_ | I'm thru like 3 episodes (: or 4 (: | 05:54 |
| + odnes (~odnes@109-178-137-160.pat.ren.cosmote.net) | 09:43 | |
| - odnes (QUIT: Read error: Connection reset by peer) (~odnes@109-178-137-160.pat.ren.cosmote.net) | 09:44 | |
| + odnes (~odnes@109-178-137-160.pat.ren.cosmote.net) | 09:45 | |
| - odnes (QUIT: Ping timeout: 252 seconds) (~odnes@109-178-137-160.pat.ren.cosmote.net) | 10:29 | |
| + odnes (~odnes@109-178-211-148.pat.ren.cosmote.net) | 12:10 | |
| - odnes (QUIT: Ping timeout: 252 seconds) (~odnes@109-178-211-148.pat.ren.cosmote.net) | 12:29 | |
| + odnes (~odnes@109-178-211-148.pat.ren.cosmote.net) | 12:49 | |
| + odnes_ (~odnes@109-178-139-91.pat.ren.cosmote.net) | 13:01 | |
| - odnes (QUIT: Read error: Connection reset by peer) (~odnes@109-178-211-148.pat.ren.cosmote.net) | 13:01 | |
| - odnes_ (QUIT: Ping timeout: 272 seconds) (~odnes@109-178-139-91.pat.ren.cosmote.net) | 13:08 | |
| + odnes (~odnes@109-178-139-91.pat.ren.cosmote.net) | 13:29 | |
| lastebil_ | mntmn: so I'm back on "me time" and would like to do things with my reform2 again (: I _believe_ what I need is "just the kernel and modules" - which I can make via the sd card creation portions, correct? | 13:44 |
| lastebil_ | (or is there a completely separate 'kernel and modules' area; or is there some firmware etc. that I should also be building?) | 13:44 |
| lastebil_ | (the 'userland' (yes I know BSD term, Linux doesn't have that concept (: but you likely know what I mean.) I will be using is from Void, as that's what I use.) | 13:46 |
| lastebil_ | oh I need to re-authenticat | 13:46 |
| lastebil_ | ok. | 13:48 |
| - lastebil_ (PART: !!unknown attribute: msg!!) (truck@77.246.204.69) | 13:48 | |
| + lastebil_ (truck@77.246.204.69) | 13:48 | |
| * lastebil_ -> lastebil | 13:48 | |
| - odnes (QUIT: Read error: No route to host) (~odnes@109-178-139-91.pat.ren.cosmote.net) | 14:23 | |
| mntmn | lastebil: you need u-boot, the kernel and a dtb. all of these can be built by scripts in the reform-system-image repo | 14:25 |
| lastebil | mntmn: oh yes, u-boot also. And we have patches to our u-boot that aren't in mainline, yes? | 14:46 |
| lastebil | (I have to build u-boot for some other things also so (: ) | 14:46 |
| mntmn | yep currently this is still a fork from the vendor's u-boot. | 14:47 |
| mntmn | bluerise has a more mainline u-boot but i haven't tested it with linux myself | 14:47 |
| bluerise | sup | 14:48 |
| bluerise | https://patchwork.ozlabs.org/project/uboot/list/?series=252942 | 14:49 |
| bluerise | I should probably re-submit with USB support, now that mainline gained it like a week ago | 14:49 |
| bluerise | https://patchwork.kernel.org/project/linux-arm-kernel/patch/20210508121650.105864-3-dev@lynxeye.de/ | 14:50 |
| bluerise | I'm a bit surprised the device tree submit stuff got 'archived' | 14:50 |
| mntmn | wot | 14:50 |
| mntmn | doesn't it just mean it's in the mailing list archive? | 14:51 |
| bluerise | https://lore.kernel.org/linux-devicetree/20210714041313.GA4419@dragon/ | 14:52 |
| bluerise | ahhh | 14:52 |
| bluerise | nice | 14:52 |
| bluerise | ok so shawnguo has it and I guess it'll head into linux soon | 14:52 |
| mntmn | yeah! need to get the internal pci refclk thing in there. in the latest reform-system-image, the patch for this is already minimal | 14:57 |
| bluerise | I'll resubmit once it's in linux + added USB support | 14:57 |
| mntmn | ok cool | 14:57 |
| bluerise | yeah, that PCIe refclk thing is... | 14:57 |
| bluerise | sigh | 14:57 |
| mntmn | the clock part of it i moved to dts already | 14:58 |
| bluerise | I've been shipping an i.MX8MQ product for like two years already | 14:58 |
| bluerise | and BACK THEN there were already solutions/workarounds for PCIe | 14:58 |
| bluerise | and it's *still* not in mainline | 14:58 |
| bluerise | I mean wtf | 14:58 |
| mntmn | yeah | 14:58 |
| mntmn | and this funny little gem https://source.mnt.re/reform/reform-system-image/-/blob/main/reform2-imx8mq/template-kernel/patches/0001-nwl-dsi-fixup-mode-only-for-LCDIF-input-not-DCSS.patch | 15:08 |
| + odnes (~odnes@109-178-139-91.pat.ren.cosmote.net) | 15:12 | |
| bluerise | lol | 15:14 |
| - odnes (QUIT: Ping timeout: 272 seconds) (~odnes@109-178-139-91.pat.ren.cosmote.net) | 15:41 | |
| - artfwo (QUIT: Ping timeout: 250 seconds) (~artfwo@2a02:8109:8500:26d0:adf7:350e:766b:41ee) | 16:04 | |
| + artfwo (~artfwo@2a02:8109:8500:26d0:99c7:b083:5156:166a) | 16:16 | |
| - artfwo (QUIT: Ping timeout: 272 seconds) (~artfwo@2a02:8109:8500:26d0:99c7:b083:5156:166a) | 16:36 | |
| + artfwo (~artfwo@2a02:8109:8500:26d0:35f7:f3d8:2973:7308) | 16:48 | |
| + ephase (~ephase@2a01:e0a:482:9c50:eb5b:d9f4:3aab:d5a1) | 16:59 | |
| ephase | Hi | 17:02 |
| mntmn | hi | 17:02 |
| ephase | mntmn: I compiled glmark2, on wayland I have EGL acceleration, but not on Xwayland | 17:03 |
| mntmn | ephase: yeah, as i said, did you try a newer xwayland? | 17:03 |
| ephase | Xwayland isn't on experimental repository, and I can't compile ir dues to protoinput dependencies | 17:04 |
| mntmn | ah, sorry. you can easily get that missing dependency | 17:05 |
| mntmn | ephase: https://github.com/freedesktop/inputproto | 17:05 |
| ephase | haaa great | 17:05 |
| mntmn | sorry, correct link probably https://gitlab.freedesktop.org/xorg/proto/inputproto | 17:06 |
| mntmn | wait, maybe you can just install x11proto-dev ? | 17:07 |
| mntmn | (debian package) | 17:07 |
| mntmn | ephase: did you try apt install x11proto-dev ? | 17:07 |
| ephase | yes I tried | 17:08 |
| mntmn | ok | 17:09 |
| mntmn | hmm inputproto is deprecated | 17:09 |
| mntmn | how exactly do you build xwayland? | 17:10 |
| ephase | but version is too old: found 2.3.2 but need >= 2.3.99.1 | 17:10 |
| ephase | meson build -Dxorg=true -Dxwayland=true -Dglamor=true -Dxwayland_eglstream=false -Dxnest=false -Ddmx=false -Dxvfb=true -Dxwin=false -Dxephyr=false -Ddri3=true | 17:10 |
| mntmn | sorry i sent you down the wrong path with inputproto, this is now deprecated and it is now inside of https://cgit.freedesktop.org/xorg/proto/xorgproto/ | 17:11 |
| ephase | ok thanks | 17:11 |
| mntmn | https://gitlab.freedesktop.org/xorg/proto/xorgproto | 17:11 |
| mntmn | yeah this includes inputproto 2.3.99.2 | 17:11 |
| mntmn | ephase: if you manage to build xwayland and install it, it will be in /usr/local/bin, but sway expects it hardcoded in /usr/bin i think... so i deleted the one in /usr/bin und replaced it with a link | 17:13 |
| mntmn | i mean, if you use the sway from debian | 17:13 |
| ephase | Yes i saw tour build script :) thanks | 17:13 |
| ephase | -tour +your | 17:14 |
| mntmn | ok :) | 17:15 |
| ephase | Hoo great it compile | 17:15 |
| ephase | thank you so much mntmn | 17:16 |
| mntmn | no problem! i think there was a recent change in xserver where they did something with input, and this required the new dependency | 17:16 |
| ephase | mntmn: for the little stain in the screen, is it possible to order a replacement screen from you? | 17:20 |
| ephase | on your shop | 17:21 |
| mntmn | ephase: are you sure the screen itself is damaged, have you tried cleaning it off with a big of detergent or tiny bit of IPA? | 17:21 |
| + odnes (~odnes@109-178-139-91.pat.ren.cosmote.net) | 17:21 | |
| ephase | mntmn: yes with a bezel cloth | 17:23 |
| mntmn | ok | 17:24 |
| ephase | Xwayland work for now and found the vivante GPU \O/ | 17:24 |
| mntmn | ephase: awesome! | 17:24 |
| mntmn | ephase: note that for some older applications you might need the one line patch from my buildscript... with glFinish() | 17:24 |
| mntmn | https://www.ebay.de/itm/Lenovo-Thinkpad-X260-FRU-00NY418-12-5-N125HCE-GN1-REV-C2-LCD-FHD-IPS-Screen-/133499252389 | 17:25 |
| mntmn | we currently don't sell the displays because we have not yet tested all the displays in our inventory | 17:25 |
| mntmn | we need to make sure we can deliver everything for the campaign etc | 17:25 |
| mntmn | but then we will start selling them as well | 17:25 |
| ephase | ok | 17:26 |
| mntmn | omg i just searched for N125HCE-GN1 on ebay.fr and the prices are insane | 17:26 |
| ephase | mntmn: ^^ ebay... I'll wait a little to order from mnt :) Stain is not so important and is visible on black screen | 17:27 |
| mntmn | ok ^^ | 17:27 |
| ephase | So qutebrowser doesn't display any content, just white pages | 17:33 |
| ephase | I think I have to compile mesa | 17:34 |
| ephase | or DRM module | 17:34 |
| ephase | and on wayland It doesn't start at all | 17:37 |
| mntmn | old mesa probably yes | 17:38 |
| mntmn | btw qutebrowser works for me without xwayland | 17:38 |
| ephase | witch version of mesa have you mntmn? | 17:39 |
| mntmn | 21.2.0 | 17:40 |
| mntmn | with patch/MR 7603 | 17:41 |
| mntmn | ah i see it was merged in the meantime https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7603 | 17:42 |
| mntmn | ephase: you will want a version of mesa that has this patch | 17:42 |
| ephase | yes for sure, then I don't have to patch it :) | 17:44 |
| mntmn | also this sounds like xserver can theoretically work https://gitlab.freedesktop.org/mesa/mesa/-/issues/3721 | 17:49 |
| + rasmus (~rasmus@c83-253-223-217.bredband.tele2.se) | 18:12 | |
| ephase | So, I compile mesa 21.2.0-rc3 | 18:15 |
| ephase | I think a can have a cup of tea while waiting... | 18:16 |
| mntmn | haha yes it can take a while | 18:21 |
| ephase | So, I'll be back this night | 18:28 |
| ephase | see you | 18:28 |
| - ephase (QUIT: Quit: WeeChat 3.2) (~ephase@2a01:e0a:482:9c50:eb5b:d9f4:3aab:d5a1) | 18:28 | |
| - artfwo (QUIT: Quit: Leaving) (~artfwo@2a02:8109:8500:26d0:35f7:f3d8:2973:7308) | 18:31 | |
| + alex4nder (~alexander@c-73-164-208-201.hsd1.or.comcast.net) | 20:05 | |
| - rasmus (PART: Disconnected: timeout during receiving) (~rasmus@c83-253-223-217.bredband.tele2.se) | 20:12 | |
| + rasmus (~rasmus@c83-253-223-217.bredband.tele2.se) | 20:13 | |
| technomancy | my wifi is still cutting out randomly even when I'm less than 1m away from the base station | 20:18 |
| technomancy | any idea what could cause deauthenticating from fe:50:cd:b7:6d:85 by local choice (Reason: 3=DEAUTH_LEAVING) | 20:22 |
| mntmn | hmm try a different channel perhaps? | 20:31 |
| - rasmus (PART: Disconnected: timeout during receiving) (~rasmus@c83-253-223-217.bredband.tele2.se) | 22:13 | |
| + rasmus (~rasmus@c83-253-223-217.bredband.tele2.se) | 22:14 | |
| - odnes (QUIT: Read error: Connection reset by peer) (~odnes@109-178-139-91.pat.ren.cosmote.net) | 22:27 | |
| + odnes (~odnes@109-178-139-91.pat.ren.cosmote.net) | 22:28 | |
| + ephase (~ephase@2a01:e0a:482:9c50:d3e2:f3b3:515b:52c5) | 22:55 | |
| - odnes (QUIT: Ping timeout: 240 seconds) (~odnes@109-178-139-91.pat.ren.cosmote.net) | 22:59 | |
| ephase | mntmn:so I tried to compile drm, mesa, sway, wlroots, xwayland but qutebrowser refuse to start | 23:10 |
| mntmn | ephase: what does it say? | 23:11 |
| ephase | Failed to create OpenGL context for format QSurfaceFormat | 23:14 |
| ephase | and Warning before : WARNING: QWaylandContext: failed to create EGLContext, error=3009 | 23:15 |
| - rasmus (PART: Disconnected: closed) (~rasmus@c83-253-223-217.bredband.tele2.se) | 23:15 | |
| mntmn | ephase: strange | 23:17 |
| mntmn | ephase: are you sure that your own GL libraries are being picked up? | 23:18 |
| ephase | Yes Mesa 21.2.0-rc3 | 23:19 |
| mntmn | ephase: qutebrowser 2.0.2-2? | 23:22 |
| mntmn | ephase: ah, and qutebrowser-qtwebengine, yes? | 23:22 |
| ephase | yes to all | 23:23 |
| mntmn | ephase: probably unrelated, but do you have ETNA_MESA_DEBUG=nir ? | 23:24 |
| ephase | no, what does this env var? | 23:25 |
| mntmn | maybe it is the default now, but if not, you need to set it, best before launching sway | 23:25 |
| mntmn | this enables the NIR shader compiler | 23:25 |
| mntmn | without it, a lot of OpenGL stuff doesn't work right | 23:25 |
| ephase | I try... | 23:25 |
| mntmn | yeah that is the reason | 23:26 |
| mntmn | when i unset that variable, qutebrowser doesn't work | 23:26 |
| ephase | IT WORK! | 23:26 |
| mntmn | ah! good :) sorry i forgot about this earlier | 23:26 |
| mntmn | this is in the default bash profile in the reform-system-image | 23:27 |
| mntmn | i wonder why it is not the default still | 23:27 |
| ephase | mntmn: no importance, now I became a great padaway learning how to compile stuff related to graphic stack on Linux ^^ | 23:28 |
| mntmn | awesome! | 23:28 |
| ephase | -padaway +padawan | 23:28 |
| mntmn | i also learned this only because i had to fix stuff for reform ^^ | 23:28 |
| ephase | webgl sample work better now :D | 23:29 |
| ephase | 50 fps on metaballs | 23:29 |
| mntmn | in qutebrowser? | 23:29 |
| ephase | yes mntmn | 23:30 |
| ephase | with default settings | 23:30 |
| mntmn | oh yeah it is also very fast here | 23:30 |
| mntmn | so qtwebengine seems to work well | 23:30 |
| ephase | yes, advantage of webkit (sandbox...) without google stuff | 23:31 |
| mntmn | there's also falkon | 23:32 |
| mntmn | same performance and a more familiar interface maybe for people who don't want vim style controls | 23:33 |
| mntmn | cc scops | 23:34 |
| scops | 👍 | 23:35 |
| mntmn | there's also viper and otter browsers but those are not in debian. falkon is | 23:35 |
| ephase | nyxt seems great too | 23:36 |
| mntmn | ephase: is there an aarch64 build? | 23:38 |
| mntmn | hmm but this is good, finally alternatives to chromium with decent performance | 23:39 |
| ephase | mntmn: I dont think so, but I'm thinking about trying to compile it | 23:40 |
| mntmn | ephase: cool | 23:40 |
| mntmn | now building openxcom ;3 | 23:59 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
