| minute | Twodisbetter: yeah, i'm also very interested in hibernate | 00:09 |
|---|---|---|
| * theesm1_ -> theesm1 | 00:12 | |
| - NanoCodeBug (QUIT: Quit: Client closed) (~NanoCodeB@c-73-35-191-67.hsd1.wa.comcast.net) | 00:14 | |
| minute | Twodisbetter: which cpu archtitecture is the steamcmd binary for? | 00:14 |
| minute | Twodisbetter: if steamcmd is /usr/bin/steamcmd, you can do: file /usr/bin/steamcmd | 00:15 |
| minute | Twodisbetter: that should tell you if it is intel x86 32 bit or 64 bit for example | 00:15 |
| minute | Twodisbetter: do you have these two files: /etc/binfmt.d/box64.conf and /etc/binfmt.d/box86.conf ? | 00:18 |
| * andypiper -> andypiper[afk] | 00:20 | |
| - andypiper[afk] (QUIT: Quit: My device has gone to sleep. Zzzz…) (~andypiper@45.146.10.36) | 00:20 | |
| mrbcmorris | Do we have any info on what the hibernate problem/feature gap is? I've heard folks talking about suspend/hibernate but i'm not sure if there's a known problem or if it's something that needs a lot more research | 00:52 |
| minute | mrbcmorris: i haven't had time to investigate it in depth | 01:00 |
| - staticbunny (QUIT: Ping timeout: 252 seconds) (~halloy193@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 01:31 | |
| - mark_ (QUIT: Ping timeout: 248 seconds) (~mjw@gnu.wildebeest.org) | 01:38 | |
| - mtm (QUIT: Ping timeout: 248 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:04 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:05 | |
| mrbcmorris | Not a worry, just had heard a few people talking about it, didn't know if we had anything to read to catch up on the issue. | 03:03 |
| - nsc_ (QUIT: Ping timeout: 260 seconds) (~nicolas@i5C74DF27.versanet.de) | 03:19 | |
| + nsc (~nicolas@i5C74DCFB.versanet.de) | 03:20 | |
| - cobra (QUIT: Quit: ZNC 1.8.2 - https://znc.in) (~cobra@user/Cobra) | 03:22 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 03:24 | |
| + cobra (~cobra@user/Cobra) | 03:31 | |
| - kop316 (QUIT: Remote host closed the connection) (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 03:56 | |
| + kop316 (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 03:57 | |
| - cobra (QUIT: Quit: ZNC 1.8.2 - https://znc.in) (~cobra@user/Cobra) | 04:02 | |
| - kop316 (QUIT: Remote host closed the connection) (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 04:03 | |
| + kop316 (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 04:04 | |
| - kop316 (QUIT: Remote host closed the connection) (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 04:04 | |
| + kop316 (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 04:04 | |
| + cobra (~cobra@user/Cobra) | 04:07 | |
| - staticbunny (QUIT: Ping timeout: 246 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 04:19 | |
| + midfavila (midfavila@freeshell.org) | 04:24 | |
| midfavila | hey, question for those with experience putting together system images for embedded machines: where would i start if i wanted to learn? i've done some tinkering on x86 and friends but not really on anything else. i'd like to port KISS Linux to both the Reform and the uConsole R-01, an open hardware handheld based on risc-v from clockwork | 04:26 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 04:27 | |
| midfavila | ive been fiddling with debian on the uconsole, trying to get it to play nice with what i want, and it's just impossible. maybe i'm getting filtered but there's way too many ways to do things and not enough documentation that's actually up to date and i'm just fed up | 04:27 |
| midfavila | admittedly clockwork has a similar issue to pine but *still* | 04:27 |
| midfavila | (wrt: software support) | 04:28 |
| staticbunny | damn that sucks, im looking forward to my uconsole | 04:31 |
| midfavila | i mean to be clear the uconsole rocks | 04:31 |
| staticbunny | maybe when i get it next year they will have better documentation ;-) | 04:31 |
| midfavila | hardware wise i only have a few complaints | 04:31 |
| midfavila | oh no i'm bitching about debian here | 04:31 |
| staticbunny | oh gotcha | 04:31 |
| midfavila | i've never liked debian | 04:31 |
| midfavila | i can *tolerate* it, but, geez, dude, who decides they need like five friggin package managers for their distro? | 04:32 |
| staticbunny | yeah its a mess distro wise, when you see how many newer distros surpassed it | 04:33 |
| midfavila | for context i was fiddling with my uconsole to set it up as a hotspot. idea being i use iptables to nat together wwan and wlan interfaces and then expose wlan to clients using hostapd | 04:33 |
| midfavila | ive done similar things before with laptops so like, mmmmmnmhmmm | 04:33 |
| midfavila | but no. they've just had to go and change it all | 04:34 |
| midfavila | maybe there's a valid reason for it but it's a pain in the ass. | 04:34 |
| midfavila | i've always preferred slackware to debian. | 04:34 |
| midfavila | like, i hate to be That Guy:tm:, but slackware... just... it just works. it just does the thing. | 04:34 |
| staticbunny | i feel ya, not everyone loves linux for wasting time troubleshooting some basic feature | 04:35 |
| midfavila | i mean fwiw my hobby is troubleshooting xorg driver stacktraces | 04:35 |
| midfavila | i run my own fork of kiss linux that's a stitched-together abomination of linux, openbsd, netbsd, plan 9, and a bunch of crap i found on github | 04:36 |
| staticbunny | that sounds crazy lol | 04:36 |
| midfavila | getting it to all work and play nice and neat together is fun, and ive learned a fair amount from slowly replacing the userland with my own programs | 04:36 |
| midfavila | while ago i was fussing about with -ck and -hardened patchsets, getting those to run together | 04:36 |
| midfavila | whatever | 04:36 |
| midfavila | no biggie | 04:36 |
| midfavila | but then i install mfing debian and it's just *suuuuuuuch* a pita to do anything | 04:37 |
| midfavila | i mean maybe im just an adminlet but gosh | 04:38 |
| staticbunny | im considering switching to cm4 or rk3588. I think i might be able to run kali on cm4 easier | 04:41 |
| staticbunny | trying to find the wiki page that said which chipset supported which boot method | 04:42 |
| midfavila | cm4 has a kali port ye | 04:42 |
| midfavila | but ngl i dont know how discreet you would be walking around with a uconsole | 04:42 |
| midfavila | unless it's for educational purposes and not work :p | 04:42 |
| staticbunny | im thinking of using uconsole with meshtastic and emulation. I feel like the pocket has way more options for expansion | 04:45 |
| staticbunny | i noticed it has an i2c connector and another internal usb-c port | 04:46 |
| midfavila | the pocket prolly does but i prefer my handhelds vertical | 04:47 |
| midfavila | :c | 04:47 |
| midfavila | im used to blackberries and the old linux zaurus | 04:48 |
| staticbunny | i dont have mine yet, but you might be right | 04:48 |
| midfavila | overall the pocket is almost certainly better though | 04:58 |
| staticbunny | even with my charge board being faulty, i'm loving it so far | 05:00 |
| staticbunny | my only complaint is documentation. i wish there was a wiki or confluence page that community could help maintain. | 05:03 |
| staticbunny | this is what i was looking for btw. https://reform.debian.net/d-i/ | 05:16 |
| - aloo_shu (QUIT: Ping timeout: 264 seconds) (~aloo_shu@90.166.98.132) | 05:31 | |
| + aloo_shu (~aloo_shu@90.166.98.132) | 05:40 | |
| - staticbunny (QUIT: Ping timeout: 260 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 06:01 | |
| josch | midfavila: i'm the person who is mainly maintaining the debian system images for the reform and i'm the upstream author of mmdebstrap and the maintainer of sbuild in debian. If you have any specific questions, i might be able to help. | 06:32 |
| midfavila | awesome, i'll keep it in mind. i'm mostly just completely lost on where to start porting a new system. i have a rootfs and a kernel that i built from the reform's source, but getting a device tree and the bootloader configured is a bit daunting since i don't have a backup computer and don't have much experience (read: any) with embedded platforms | 06:33 |
| midfavila | i would assume it's as simple as dumping the rootfs on a partition with a kernel and pointing the bootloader at it and the DTC, but... i'm not sure | 06:34 |
| josch | midfavila: the biggest pain with non x86 is the bootloader part i find | 06:35 |
| josch | once you have kernel and initramfs loaded, it's mostly just your normal distro from that point onward | 06:35 |
| midfavila | yeah the initramfs is another thing i was iffy about | 06:35 |
| midfavila | i uhhhhh normally use a fat kernel | 06:36 |
| josch | sure, you can do that too | 06:36 |
| midfavila | so i'm not sure how that would work on the reform | 06:36 |
| midfavila | ig im mostly just nervous about bricking it lmao | 06:36 |
| josch | midfavila: which reform do you have? | 06:36 |
| josch | without a second computer, doing this kind of tinkering is extremely challenging i'd say | 06:36 |
| josch | i did most of my work when i still had my thinkpad | 06:37 |
| midfavila | the imx8mq | 06:37 |
| midfavila | first revision | 06:37 |
| midfavila | got it off crowd supply | 06:37 |
| josch | nice! | 06:37 |
| josch | then you have the big advantage that you even got display support with u-boot | 06:37 |
| josch | so you can nice perks like a functioning boot menu that lets you choose what system or kernel to load | 06:38 |
| josch | the other SoMs the reform supports do not have that feature sadly | 06:38 |
| midfavila | yeah that's why i got it off crowd supply | 06:38 |
| midfavila | the other modules are neat and it's amazing the choice is there but in terms of support and power usage the imx8mq seems the best | 06:38 |
| midfavila | i'm also not a fan of integrated wireless because that's just a peeve of mine | 06:39 |
| midfavila | >.< | 06:39 |
| josch | the imx8mq was my main computer for nearly a year -- it certainly worked for me | 06:39 |
| midfavila | yeah it's been great so far | 06:39 |
| midfavila | i got an x server running bare on it a while ago and it's p comfy | 06:39 |
| midfavila | tinyx runs nicely with a quick edit to the source to let the preproc match aarch against arm32 | 06:39 |
| midfavila | now and again firefox locks the system up but that's because the modern web is a disaster | 06:40 |
| josch | yup :( | 06:40 |
| midfavila | i don't intend to buy another laptop for personal use again | 06:40 |
| midfavila | unless there ends up being a plus-size reform | 06:40 |
| midfavila | with sata and stuff | 06:40 |
| midfavila | wink wink hint hint nudge nudge | 06:41 |
| midfavila | :3 | 06:41 |
| josch | if you have multiple sd-cards, maybe start by setting the dip switch on the som to load u-boot from there | 06:41 |
| midfavila | yeah | 06:41 |
| midfavila | i need to get spares | 06:41 |
| josch | i get my sd-cards second-hand | 06:41 |
| midfavila | a dangerous game | 06:41 |
| josch | you get a box of them for very few euros | 06:41 |
| josch | and i do not need more than 4 GB anyways | 06:42 |
| midfavila | hmm | 06:42 |
| midfavila | where do you get these boxes? | 06:42 |
| midfavila | 4gb is plenty for my testing image | 06:42 |
| josch | in germany there are a couple of second-hand websites | 06:42 |
| midfavila | the entire system is like under a gig | 06:42 |
| josch | i'm sure things like that exist elsewhere as well | 06:42 |
| midfavila | i used to tinker with my fork and once i got a complete system including toolchain, debugger, X server, everything, in under a hundred megs | 06:43 |
| midfavila | i think compressed it was 24mb? | 06:43 |
| josch | squashfs? | 06:43 |
| midfavila | but yeah i'll have to look around | 06:43 |
| midfavila | nope | 06:43 |
| midfavila | that's just regular gzip | 06:43 |
| josch | ah kk | 06:43 |
| midfavila | mind you it was a *very* stripped down system but for daily use it was sufficient | 06:43 |
| josch | oh, this page might be interesting to you: | 06:43 |
| midfavila | had a browser and stuff | 06:44 |
| midfavila | oh? | 06:44 |
| josch | https://source.mnt.re/reform/reform-handbook/-/issues/2 | 06:44 |
| midfavila | oh yeah | 06:44 |
| midfavila | reminds me | 06:44 |
| midfavila | i need to attack crowd supply | 06:44 |
| midfavila | i ordered a handbook and they still havent sent one to me | 06:44 |
| midfavila | like three months down the line | 06:44 |
| midfavila | kneecaps *will* be broken | 06:44 |
| josch | please don't break people's kneecaps | 06:44 |
| midfavila | >:3 | 06:45 |
| dozens | hide your knees hide your caps | 06:45 |
| midfavila | the kneecappening commences at 0000EST | 06:45 |
| midfavila | tomorrow | 06:45 |
| josch | so once, your imx8mq loads u-boot from sd-card, create a extlinux/extlinux.conf file in your /boot and use that to load your kernel and dtb | 06:45 |
| midfavila | hmmm | 06:46 |
| josch | i'd recommend avoiding writing a boot.scr | 06:46 |
| midfavila | if it's really just that simple i should be able to get that done | 06:46 |
| midfavila | the scr was something i was nervous about yeah | 06:46 |
| josch | it should be just that simple | 06:46 |
| midfavila | okay cool we'll see tomorrow then :3 | 06:46 |
| josch | extlinux.conf is just a machine readable, declarative way of telling u-boot where your kernel, dtb and initramfs are | 06:46 |
| midfavila | yeah i've used extlinux lots before | 06:47 |
| josch | and what your kernel boot arguments are | 06:47 |
| midfavila | it's my usual bootloader | 06:47 |
| midfavila | on desktops anyways | 06:47 |
| josch | cool -- it's just like that | 06:47 |
| midfavila | when i was on x86 i had a little 200b assembly program that would boot my laptops | 06:47 |
| josch | for imx8mq, there are some bootargs you might want to pass | 06:47 |
| midfavila | oh? | 06:47 |
| josch | https://source.mnt.re/reform/reform-tools/-/blob/main/machines/MNT%20Reform%202.conf#L19 | 06:48 |
| josch | this is the line we use by default | 06:48 |
| josch | you might not need all of that but it's a starting point and then you can tinker and see what works for you | 06:48 |
| midfavila | will do | 06:49 |
| midfavila | it's something i hafta get over the hump of anyway | 06:49 |
| midfavila | i'm hoping to drop x86 | 06:49 |
| henesy | Twodisbetter: is this running with box64? i never had any exec issues | 07:25 |
| josch | henesy: do you have box86 installed in addition to box64? | 07:33 |
| henesy | i believe so | 08:11 |
| henesy | i thought box64 dep'd on box86 | 08:11 |
| josch | it does not | 08:16 |
| josch | i packaged box64 for debian but not box86 | 08:17 |
| josch | and box64 works completely independently of box86 | 08:17 |
| josch | did you build it from source? | 08:17 |
| henesy | ohhh i used the github box64 repos | 08:38 |
| henesy | https://paste.sr.ht/~henesy/d84b969cc066bfee95c3098e03cb534f4a424076 | 08:40 |
| + mark_ (~mjw@gnu.wildebeest.org) | 08:54 | |
| josch | Twodisbetter: this is the information you wanted ^ | 09:01 |
| Twodisbetter | josch: Thanks! | 09:03 |
| Twodisbetter | Left the pocket at home today so I'll sort this out this evening when I get home. | 09:03 |
| henesy | so yes i have both installed | 09:09 |
| henesy | tho there | 09:09 |
| henesy | 's not much box86 alone can run | 09:09 |
| - mark_ (QUIT: Ping timeout: 246 seconds) (~mjw@gnu.wildebeest.org) | 09:31 | |
| * Guest7933 -> mjw | 09:44 | |
| minute | midfavila: crowd supply doesn't have any reform handbooks/didn't order any from us. | 09:55 |
| + _hramrach (~hramrach@193.86.92.181) | 09:57 | |
| _hramrach | The mnt pocket screen is glossy :\ | 09:58 |
| vkoskiv | https://mister-muffin.de/p/xfhP.txt | 10:06 |
| vkoskiv | Known issue? | 10:06 |
| vkoskiv | (linux pkg having issues) | 10:08 |
| ch | https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1077678 | 10:08 |
| ch | xtrx package is the problem, not linux | 10:08 |
| vkoskiv | Looking at /boot, I think I'm still good to go, there are a bunch of kernels in there | 10:09 |
| vkoskiv | Unusual that linux would ship with an SDR driver built in like that | 10:13 |
| vkoskiv | Well, as a module anyway. | 10:13 |
| _hramrach | Also the setup wizard has broken control properties like default button, cannot be easily operated with keyboard | 10:13 |
| josch | _hramrach: can you report that issue here so that it does not get lost? https://source.mnt.re/reform/mnt-reform-setup-wizard/-/issues/ | 10:16 |
| + NanoCodeBug (~NanoCodeB@c-73-35-191-67.hsd1.wa.comcast.net) | 10:32 | |
| _hramrach | charging is very slow, went from 22% to 52% overnight, and the current dropped from 1A at 20V to 0.3A | 10:38 |
| josch | _hramrach: i don't have a pocket but since nobody else answered: there is a firmware update. Did you install that? https://source.mnt.re/reform/pocket-reform/-/releases/2024-08-26B Also for the charging percentage to be correct you apparently first need to let the battery drain all the way down to zero (so that it switches off) and then charge it up again and only then is it calibrated. | 10:48 |
| _hramrach | so it can be misreported, too. It says the voltage is 3.8/4.0 V | 10:58 |
| _hramrach | did not do any update yet | 10:58 |
| _hramrach | I will try to do the calibration, thanks | 11:08 |
| + andypiper (~andypiper@89.36.117.58) | 11:17 | |
| _hramrach | hm, I see the sway config that is shipped sets some gsettings every time sway starts. I find this rather unfriendly. There are other tools fro managing these settings, and with this every time the user logs in the changes get reverted. Reminds of GNOME reverting keyboard and screen settings on every GNOME start | 11:36 |
| josch | _hramrach: good point. Do you know how to replace gsettings with something that works even in a chroot without xsettings daemon running? | 11:38 |
| _hramrach | Also the background is hardcoded while there seems to be a variable for setting it https://github.com/manjaro-sway/manjaro-sway/commit/afa36 | 11:41 |
| _hramrach | I suppose you the setup wizard runs before the user session is started, and gsettings needs the user session so it cannot be used in the wizard | 11:43 |
| josch | _hramrach: how is what you linked applicable to the default sway config in the mnt repo? | 11:43 |
| josch | _hramrach: yes, if we want to use gsettings, a mechanism would be needed which runs it on first login into sway and then never again | 11:44 |
| _hramrach | josch: the config sets hardcoded background with swaybg rather than setting the designeated config variable for it | 11:44 |
| _hramrach | josch: The gsettings are backed by some database that can be distributed already pre-made in /etc/skel. It's not like the setting depend dynamically on user input, they are hardcoded | 11:45 |
| josch | _hramrach: yes, but i'd imagine that that database looks different depending on the settings version? | 11:46 |
| josch | thus using gsettings avoids having to ship different databases for different settings versions | 11:46 |
| _hramrach | does it? | 11:49 |
| _hramrach | That would mean settings resetting all the time | 11:49 |
| josch | if i understand it correctly, then settings are stored in ~/.config/dconf/user which is a binary file in GVariant Database format | 11:50 |
| josch | how do i initialize and set that up without gsettings? | 11:50 |
| _hramrach | there are alos some override files, apparently: https://github.com/manjaro-sway/manjaro-sway/commit/afa36 | 11:53 |
| _hramrach | wrong link | 11:53 |
| _hramrach | https://askubuntu.com/questions/73096/how-to-make-applications-pinned-in-launcher-appear-for-all-users | 11:53 |
| josch | oh that looks useful! | 11:55 |
| josch | _hramrach: can you provide a patch? | 11:55 |
| _hramrach | not sure I can provide a patch | 11:57 |
| josch | oh i think i found it | 11:57 |
| _hramrach | here is the doc about overrides https://help.gnome.org/admin/system-admin-guide/unstable/overrides.html.en | 11:57 |
| josch | this should do the trick: | 12:01 |
| josch | https://paste.debian.net/1328289/ | 12:01 |
| josch | _hramrach: could you place that into /usr/share/glib-2.0/schemas/20_reform.gschema.override create a new user and see if that does the trick? | 12:02 |
| minute | _hramrach: the keyboard controls in reform-setup-wizard are indeed bad, sorry for that. | 12:06 |
| minute | cc plomlompom lets fix this | 12:06 |
| minute | josch: a fixed pocket-reform-manual with images is just waiting for some debian upload to go through, right? or do i still need to provide sth? | 12:07 |
| minute | i have a list of homework to polish the pocket reform experience... i guess i should create issues for everything in pocket-reform issue tracker | 12:07 |
| josch | minute: i didn't touch the packages built by the MNT gitlab. I created new packages from scratch but only uploaded the big-reform handbook so far | 12:08 |
| _hramrach | josch: looks like the $background setting does not work. Not sure if they have downstream patch for that or I did not set it correctly | 12:13 |
| josch | _hramrach: the $background variable is consumed by their sway configuration | 12:14 |
| josch | interestingly, the manjaro image you linked is doing a very similar stunt as we do with gsettings: https://paste.debian.net/hidden/768ec5dc/ | 12:14 |
| josch | (that's from their /etc/sway/config.d/90-enable-theme.conf) | 12:14 |
| minute | i think back in the day it was the recommended way to do it from arch or sway wiki, i think | 12:17 |
| josch | _hramrach: the $background variable works for them, because they ship a file /etc/sway/autostart which has: | 12:17 |
| minute | sorry, double "i think"!! | 12:17 |
| josch | set $apply_background swaymsg 'output * bg $background fill' | 12:17 |
| josch | now triple! :D | 12:17 |
| minute | haha | 12:17 |
| josch | though i see the issue with always resetting this every time sway starts even if the user changed it with gsetting or the gui tool | 12:18 |
| josch | and then gets surprised that it gets reset and doesn't realize that their sway config is the culprit | 12:18 |
| minute | sway isn't a full DE in terms of there are no centralized settings except for its config file and random state poked into the remains of other DEs like gnome | 12:18 |
| josch | the only problem i see with shipping the defaults in /usr/share/glib-2.0/schemas/20_reform.gschema.override is, that then those defaults would also be applied for other desktop environments than sway | 12:19 |
| josch | so if the user would install gnome on their machine, then those would be the defaults as well and they'd have to manually change them | 12:19 |
| minute | i've been thinking about a "indie" settings app that would at least provide a central place for all this stuff and does the poking for you | 12:19 |
| minute | i guess it is relatively easy to discover things in the sway config | 12:20 |
| _hramrach | josch: yes ,suspected something like that might be the case. That said it does make the configuration more data-driven and less reliant on code which might be generally a good idea | 12:20 |
| josch | _hramrach: i agree with moving from turing complete scripts to declarative machine readable data, yes | 12:20 |
| minute | did you know i made a NIH logo back in the day? http://dump.mntmn.com/nih.png | 12:21 |
| josch | minute: do you think using a /usr/share/glib-2.0/schemas/20_reform.gschema.override instead of calling gsettings each time sway starts would be an improvement? | 12:21 |
| josch | minute: the existing NIH logo wasn't good enough so you had to make your own? ;) | 12:21 |
| minute | josch: i don't know how this override would work | 12:21 |
| minute | josch: ahaha | 12:22 |
| _hramrach | I think it would be an improvement if it turns out to work | 12:22 |
| josch | minute: as far as i understand the docs, the override would apply these settings the as the default the first time the user initializes their settings database | 12:22 |
| minute | josch: so the user can change them later with gnome-tweaks or sth...? | 12:22 |
| _hramrach | and some tool needs to be called to compile those settings into the binary database | 12:23 |
| josch | or by running gsettings | 12:23 |
| josch | _hramrach: that tool would be called automatically if the file gets shipped by the reform-tools package as a trigger on the /usr/share/glib-2.0/schemas/ directory in glib2.0 | 12:23 |
| _hramrach | yes, it may be automated by the packaging or it may need some poking | 12:25 |
| _hramrach | I used ot use lxappearance for theme settings. Not sure it still works | 12:28 |
| - andypiper (QUIT: Ping timeout: 246 seconds) (~andypiper@89.36.117.58) | 12:35 | |
| minute | ah i forgot about lxappearance | 12:46 |
| minute | josch: ok so about pocket-reform-handbook, i installed the package like 2 days ago after watching deBear's unboxing video, because i wanted to suggest to install that package. but it isn't the build that i tagged here i think? because all the images were broken https://source.mnt.re/reform/pocket-reform-handbook/-/commits/main?ref_type=heads | 12:50 |
| minute | so i'm not sure what i installed there :D | 12:51 |
| minute | not sure how this works https://source.mnt.re/reform/reform-debian-packages/-/blob/main/build_custom.sh?ref_type=heads#L71 | 12:52 |
| josch | minute: line 78 will always get the latest git commit | 12:53 |
| josch | build_custom.sh are not sticking to a specific tag/release for either handbook | 12:54 |
| josch | they just build the top commit | 12:54 |
| minute | ok cool... so i will try to build locally from a fresh git to see what's up | 12:54 |
| josch | yes, that should reproduce the problem | 12:54 |
| josch | here is the debian/rules i uploaded to debian: https://salsa.debian.org/reform-team/reform-handbook/-/blob/master/debian/rules?ref_type=heads | 12:58 |
| josch | one notable difference to the one on source.mnt.re is the compilation step of the *-icon.png files which are the images that are missing from the reform-handbook in the mntre.com repo | 12:58 |
| minute | aha | 12:58 |
| josch | for the pocket-reform-handbook, i think the ./make-pngs.sh is the one missing | 12:58 |
| minute | the debian build script for the web version is wrong. it calls make directly | 12:59 |
| minute | but it should call build-web.sh | 12:59 |
| minute | there are several workarounds in build-web.sh that need to be applied before calling make | 12:59 |
| minute | i.e. this https://source.mnt.re/reform/pocket-reform-handbook/-/blob/main/src/build-web.sh?ref_type=heads | 12:59 |
| minute | josch: should i try messing with the debian install script or do you know how to do that in a few seconds? :D | 13:01 |
| josch | minute: as you can see, my d/rules replicates big parts of what ./src/build-web.sh does. It does not call ./src/build-web.sh because i want to ln instead of cp | 13:02 |
| josch | minute: i'll give it a try | 13:03 |
| josch | specifically, calling cp instead of ln creates duplicates which eat up space unnecessarily | 13:03 |
| minute | josch: ok thanks, also the sed is important to fix a table that would otherwise disappear. unfortunately sphinx and pandoc disagree on the syntax | 13:04 |
| josch | aha i was wondering about that | 13:04 |
| NanoCodeBug | i may have gotten a bit silly with it | 13:04 |
| NanoCodeBug | but i have refactored the pocket firmware and added toggle between usb/pd | 13:04 |
| NanoCodeBug | i have only tested it with my previously unworking anker charger, swapping between it and a sd card reader | 13:05 |
| NanoCodeBug | the important logic is in here https://source.mnt.re/NanoCodeBug/pocket-reform/-/blob/main/pocket-reform-sysctl-fw/src/pd_com.c?ref_type=heads | 13:05 |
| NanoCodeBug | but i hope the refactor makes the firmware a bit cleaner by moving each communication peripheral into its own function | 13:05 |
| josch | minute: ah we were talking past each other -- now i see what you mean! fixing it... :) | 13:07 |
| minute | NanoCodeBug: oh wow | 13:08 |
| NanoCodeBug | i initially tried using the fusb302b internal interrupt and automatic toggle feature, but it was strange - it takes control of the CC pins so you can't poll them manually anymore, which would be fine, if that didn't interfere with the pd communication. but disabling the fusb302b automatic toggle feature breaks the interrupts, would be nice if that | 13:08 |
| NanoCodeBug | worked so there doesn't have to be a poll loop within the rp2040 | 13:08 |
| minute | ah i see | 13:08 |
| minute | i wish the successor of the rp2040 had already been out at that time | 13:09 |
| minute | we had beta silicon of it here since a long time... | 13:09 |
| minute | (saying that because it has real low power modes) | 13:09 |
| minute | NanoCodeBug: very nice refactor, thanks a ton! could you make a merge request to the main repo? then i would see to testing it | 13:11 |
| NanoCodeBug | ah that would be nice, the lack of low power is what kept me from using the rp2040 for my hobby projects, i'll have to take a look at the new one | 13:11 |
| minute | at some point we can make spins of the keyboard + motherboard with the new one, but would like to gather more data, problems, etc from everyone first ^^ | 13:12 |
| NanoCodeBug | minute: i can! i have not given it a full once over after getting it working so it might be a bit rough still. pretty sure my change of the build script will have broken the build pipeline as well. | 13:13 |
| minute | NanoCodeBug: no problem | 13:13 |
| minute | i can assign a builder to your project as well | 13:14 |
| NanoCodeBug | https://source.mnt.re/reform/pocket-reform/-/merge_requests/8 | 13:14 |
| minute | a "runner", that is | 13:14 |
| NanoCodeBug | i would love to chat more but it is big 4 am on the pacific coast so I shall get some sleep now, feel free to leave feedback or build on top of the PR | 13:15 |
| minute | NanoCodeBug: for sure! good night and thanks again! | 13:16 |
| minute | i wonder... if something like background image path is available in gsettings as well? :0 | 13:19 |
| + mark_ (~mjw@gnu.wildebeest.org) | 13:22 | |
| josch | minute: you mean so that in the sway config you call $(gsettings get org.gnome.desktop.background picture-uri) to retrieve it? | 13:25 |
| * mjw -> Guest555 | 13:26 | |
| - Guest555 (QUIT: Killed (iridium.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 13:26 | |
| * mark_ -> mjw | 13:26 | |
| + Guest555 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 13:26 | |
| josch | minute: and yes, it's possible -- look at /usr/share/glib-2.0/schemas/10_desktop-base.gschema.override | 13:27 |
| josch | this is what i have so far: https://source.mnt.re/reform/reform-tools/-/blob/65347cdb6e5ecf02dd8b6782a22d069d7e075643/schemas/20_reform.gschema.override | 13:29 |
| minute | josch: exactly @ background-uri, then one could use some settings tool to change it (i don't know which one though), and also it wouldn't solve the problem that the swaybg command would need to be re-run. but maybe there's a better alternative to swaybg that could auto refresh | 13:37 |
| minute | josch: i've even seen someone from our team struggle with setting a background image, that's why i'm asking | 13:38 |
| minute | josch: that's really cool @ override | 13:38 |
| josch | nice, lets try it out then :) | 13:38 |
| josch | i had my partner curse at the wallpaper settings in gnome flashback yesterday (no option to change the wallpaper mode between zoom, stretch, fit etc) -- so i'd say we are not *that* far worse :) | 13:40 |
| minute | haha ok :D | 13:40 |
| minute | wow: | 13:41 |
| minute | minute@minute-i9:~$ lxappearance | 13:41 |
| minute | Segmentation fault (core dumped) | 13:41 |
| minute | (i just installed it ooc on my i9 debian pc) | 13:41 |
| minute | wow, crash is in > #1 0x00007ffff7338e31 in XInternAtom () at /lib/x86_64-linux-gnu/libX11.so.6 | 13:42 |
| minute | just wow | 13:42 |
| minute | old bug https://github.com/NixOS/nixpkgs/issues/67697#issuecomment-699679481 | 13:44 |
| minute | ok, lets close the lxappearance chapter here | 13:44 |
| minute | a shame because it's a actually a solid looking app (runs with GDK_BACKEND=x11 lxappearance) | 13:46 |
| minute | i wonder why noone has fixed that xwayland bug... | 13:46 |
| minute | i'm now using github code search to look for UIs that set org.gnome.desktop.background | 13:50 |
| minute | there's lots of interesting stuff :D https://github.com/search?q=org.gnome.desktop.background&type=code | 13:51 |
| - buckket (QUIT: Quit: buckket) (~buckket@vps.buckket.org) | 13:53 | |
| minute | this mentions an "X background" https://github.com/derf/feh/issues/225 | 13:53 |
| + buckket (~buckket@vps.buckket.org) | 13:53 | |
| _hramrach | waybg does not exit, it keeps running. Then it could be patched to listen to gsettings changes | 14:01 |
| _hramrach | of course, it would balso possible to write a simple script that does the listening and pokes swaybg accordingly | 14:02 |
| - mtm (QUIT: Ping timeout: 246 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:03 | |
| josch | minute: this should fix the missing images issue: https://source.mnt.re/reform/pocket-reform-handbook/-/merge_requests/2 | 14:04 |
| josch | minute: do you have an image that is known to be missing that i could check to be present now? | 14:04 |
| josch | casual browsing through the html looks good | 14:04 |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:05 | |
| minute | _hramrach: neat idea about the script. i was thinking that maybe there are other background tools that had this listening already built in, but don't know | 14:13 |
| + bkeys (~Thunderbi@45.134.140.153) | 14:25 | |
| minute | josch: do you have a deb for me to quickly inspect? | 14:39 |
| minute | i mean, it should be fine | 14:39 |
| minute | i can try the .deb of the next reform-debian-packages when we build it | 14:39 |
| - nsc (QUIT: Ping timeout: 246 seconds) (~nicolas@i5C74DCFB.versanet.de) | 14:43 | |
| + nsc (~nicolas@i5C74DCFB.versanet.de) | 14:45 | |
| josch | minute: https://mister-muffin.de/reform/pocket-reform-handbook_1.0_all.deb | 14:48 |
| minute | i've sent an email to the openmoko project to apply for 10 usb product IDs | 15:08 |
| minute | lets see... | 15:08 |
| minute | josch: thanks!! | 15:08 |
| minute | josch: the images are broken unfortunately, for example file:///usr/share/doc/pocket-reform-handbook/html/quickstart.html | 15:09 |
| minute | and there is no logo :0 | 15:09 |
| minute | i guess the logo would be the wrong one in any case, will check that now | 15:10 |
| minute | i just fixed various css fixes | 15:32 |
| minute | josch: there were a bunch of bugs left in pocket-reform-handbook web build, now pushed fixes | 15:55 |
| minute | there's still the odd weird-sized graphic, but better already https://mntre.com/documentation/pocket-reform-handbook/safety.html | 15:57 |
| minute | still pretty broken on iphone, but that's for another day perhaps | 15:59 |
| * mjw -> Guest8314 | 16:05 | |
| - Guest8314 (QUIT: Killed (calcium.libera.chat (Nickname regained by services))) (~mjw@gnu.wildebeest.org) | 16:05 | |
| * Guest555 -> mjw | 16:05 | |
| + Guest8314 (~mjw@gnu.wildebeest.org) | 16:05 | |
| minute | ok anyway did the fix for mobile | 16:14 |
| josch | ah the problem was in quickstart.html | 17:10 |
| josch | for the big reform handbook the problems were in hardware.html | 17:10 |
| minute | fixed the graphics etc etc | 17:27 |
| minute | https://mntre.com/documentation/pocket-reform-handbook | 17:27 |
| minute | and the reform handbook is now at https://mntre.com/documentation/reform-handbook | 17:28 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 17:28 | |
| josch | ah i see you made a redirect from https://mntre.com/reform2/handbook to https://mntre.com/documentation/reform-handbook -- nice! :) | 17:31 |
| minute | yeah :3 | 17:32 |
| minute | uh oh: | 17:33 |
| minute | minute@minute-i9:~$ gsettings get org.gnome.desktop.background picture-uri | 17:33 |
| minute | 'file:///usr/share/images/desktop-base/desktop-background.xml' | 17:34 |
| minute | lol | 17:34 |
| josch | minute: that's from /usr/share/glib-2.0/schemas/10_desktop-base.gschema.override | 17:34 |
| josch | it looks like that xml file can do even more magic if tags like <duration>8640000.0</duration> can be any indication | 17:35 |
| minute | that's... something | 17:35 |
| josch | speaking about wallpapers my brain tells me that there used to be this tool which was essentially a wrapper around tons of tools that are able to show a wallpaper (imagemagick can do it too) and it would essentially check what is installed on the system and then use whatever is there... but maybe that was back in Xorg-only days... | 17:36 |
| minute | there's another surprise | 17:36 |
| minute | there's also picture-uri-dark | 17:37 |
| josch | minute: you removed src/source/_inkscape/4-oled-menu.svg but then the latex build fails because it cannot find _static/illustrations/4-oled-menu.eps | 17:37 |
| minute | josch: oh no | 17:38 |
| minute | josch: i'll restore it | 17:38 |
| josch | maybe there is more | 17:38 |
| josch | i can fish it out of the git history and create a MR | 17:38 |
| minute | josch: i deleted some svgs that were wrongly in there, leftovers from reform handbook | 17:38 |
| minute | i now have a background image in sway where i can do this and it changes: gsettings set org.gnome.desktop.background picture-uri-dark file:///usr/share/backgrounds/reform-mountains.jpg | 17:39 |
| minute | because: https://mastodon.social/@AdrianVovk@fosstodon.org/113074115803042636 | 17:39 |
| josch | nice, vala :) | 17:41 |
| minute | unfortunately thunar crashes when selecting set as wallpaper ;______________; | 17:42 |
| minute | i wonder what nautilus does... | 17:42 |
| sigrid | josch: fbsetbg? :) | 17:42 |
| sigrid | it's of fluxbox heritage | 17:42 |
| minute | nautilus "set as background" also does nothing | 17:44 |
| minute | sigh | 17:44 |
| minute | so both do some other proprietary stuff i guess | 17:44 |
| minute | maybe they send a lonely dbus message to nowhere | 17:44 |
| josch | sigrid: yeeeeees that was it! :D | 17:45 |
| minute | aha, one needs XDG_CURRENT_DESKTOP https://github.com/xfce-mirror/thunar/blob/7f468882b9580b1ce994c924efd3fa7be148d3cc/plugins/thunar-wallpaper/twp-provider.c#L194 | 17:45 |
| minute | > Segmentation fault (core dumped) | 17:46 |
| minute | makes sense | 17:46 |
| josch | oh no XD | 17:47 |
| ch | the only correct answer | 17:47 |
| josch | seems 2024 is not the year of the linux desktop after all :) | 17:47 |
| minute | > 0x00007ffff6c58889 in XNextRequest () from /lib/x86_64-linux-gnu/libX11.so.6 | 17:50 |
| minute | can't make this up | 17:50 |
| minute | all segfaults today are in libX11 | 17:50 |
| sigrid | why does it even go to x11? | 17:50 |
| sigrid | is thunar using xwayland? | 17:51 |
| minute | legacy stuff i guess... lets see the code | 17:51 |
| minute | sigrid: it shouldn't, because it's gtk3 | 17:51 |
| minute | well, probably thunar doesn't actually support gnome settings, because it's part of xfce4 normally | 17:51 |
| minute | hmm no it should do exactly what we want https://github.com/xfce-mirror/thunar/blob/7f468882b9580b1ce994c924efd3fa7be148d3cc/plugins/thunar-wallpaper/twp-provider.c#L302 | 17:52 |
| minute | ah there are no debug symbols for #2 0x00007ffff5713701 in () at /usr/lib/x86_64-linux-gnu/thunarx-3/thunar-wallpaper-plugin.so | 17:54 |
| josch | minute: is https://mntre.com/documentation/pocket-reform-handbook supposed to contain the files produced by the pocket-reform-handbook git? Because there is https://mntre.com/documentation/pocket-reform-handbook/_images/4-oled-battery-status.png and the svg for that is missing | 17:55 |
| minute | josch: yep, i'll reinstate it | 17:56 |
| minute | also 4-oled-battery-status.svg | 17:59 |
| minute | josch: pushed | 18:00 |
| josch | minute: i went through all the files you removed and the only files i did *not* see being used are 4-keyboard-v1-special-keys and 9-trackball-sensor-v1 | 18:00 |
| minute | oh ok one moment | 18:00 |
| minute | josch: that can't be | 18:01 |
| josch | oh? | 18:02 |
| minute | josch: maybe you looked at reform-handbook sources? | 18:02 |
| josch | i did XD | 18:02 |
| josch | argh | 18:02 |
| josch | sorry | 18:02 |
| minute | ah :D | 18:02 |
| josch | :( | 18:02 |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 18:02 | |
| minute | josch: i think only 2 files needed restoring, the 4-oled*.svg | 18:02 |
| josch | lets do this again... | 18:03 |
| minute | josch: don't all debian packages have debug symbols? | 18:04 |
| minute | (available i mean) | 18:04 |
| josch | okay, only 4-oled-battery-status and 4-oled-menu | 18:05 |
| josch | minute: no, it's possible to upload packages to the Debian archive without them producing dbgsym packages | 18:05 |
| minute | ok... asking for thunar ;__; | 18:07 |
| minute | ah, there's an upgrade available | 18:07 |
| josch | thunar has thunar-dbgsym and libthunarx-3-0-dbgsym packages | 18:08 |
| josch | so that should work | 18:08 |
| minute | josch: yes, it was my mistake, one needs an up to date package | 18:08 |
| minute | crash is in twp_get_active_workspace_number | 18:09 |
| minute | gasp https://github.com/xfce-mirror/thunar/blob/7f468882b9580b1ce994c924efd3fa7be148d3cc/plugins/thunar-wallpaper/twp-provider.c#L328 | 18:09 |
| + andreas-e (~Andreas@2001:861:c4:f2f0::c64) | 18:10 | |
| minute | sigrid: that's the X11 right there | 18:11 |
| midfavila | minute: rip fair enough | 18:12 |
| midfavila | i'll have to ask them for a partial refund then | 18:12 |
| midfavila | i did pay out for a handbook | 18:12 |
| minute | midfavila: yeah, i'm sorry for that | 18:13 |
| midfavila | i don't strictly need one but i like the idea of it | 18:13 |
| midfavila | nah it's just a mixup :p | 18:13 |
| minute | we offer it in our shop but i guess that's too high a shipping cost | 18:13 |
| midfavila | i'll probably order one straight from you in the future | 18:13 |
| midfavila | whatever happened to the spiral-bind btw? | 18:13 |
| minute | anyway the thunar thing won't work because it doesn't set picture-uri-dark | 18:14 |
| minute | and i guess we won't ship a fork of thunar... | 18:14 |
| minute | but we could ship gde-background and let it use only picture-uri. | 18:15 |
| minute | i will just raise an issue about the x11 related crash in the xfce gitlab | 18:16 |
| + andypiper (~andypiper@89.36.117.58) | 18:17 | |
| minute | ok i just patched gde-background a bit and i can set the wallaper with gnome tweaks | 18:19 |
| * mjw -> Guest55 | 18:25 | |
| - Guest55 (QUIT: Killed (tungsten.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 18:25 | |
| * Guest8314 -> mjw | 18:25 | |
| + Guest55 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 18:25 | |
| minute | filed https://gitlab.xfce.org/xfce/thunar/-/issues/1437 | 18:34 |
| - andreas-e (QUIT: Quit: Leaving) (~Andreas@2001:861:c4:f2f0::c64) | 18:35 | |
| minute | of course we could put a filepicker for the background in reform-waypanel... | 18:37 |
| _hramrach | MATE has its own settings for wallpaper ... because wallpaper settings must be desktop-specific there is no way for the settings to be general | 18:57 |
| _hramrach | of course, it does not help that gsettings has every setting prefixed with the project URL | 18:58 |
| minute | _hramrach: not sure if sarcasm :D | 18:59 |
| minute | wow https://gitlab.xfce.org/xfce/thunar/-/issues/1437 | 19:03 |
| minute | they already fixed the bug | 19:03 |
| minute | this is the fastest bugfix i've ever seen | 19:03 |
| andypiper | the power of the reform community! :-) | 19:04 |
| minute | if this gets into debian, and we package gde- | 19:05 |
| minute | ...-background, we can have nice things | 19:05 |
| minute | (setting wallpaper in sway/wayfire via thunar) | 19:06 |
| * andypiper -> andypiper[afk] | 19:07 | |
| - andypiper[afk] (QUIT: Quit: My device has gone to sleep. Zzzz…) (~andypiper@89.36.117.58) | 19:07 | |
| minute | so, the recipe for getting nice wallpapers will be: 1. replace swaybg with gde-background (with background-uri-dark patched out, 1 line fix) 2. set XDG_CURRENT_DESKTOP=GNOME 3. ship thunar 4.18.12 when it comes out | 19:08 |
| _hramrach | cool | 19:33 |
| _hramrach | hm, somhow the wifi died, and I can't even figure out the module name of the driver to reload it | 19:34 |
| _hramrach | hm, and it can be unloaded but not loaded. Something fishy going on | 19:39 |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 19:44 | |
| - laumann_ (QUIT: Changing host) (~quassel@2a0a-e5c0-2-2-0-c8ff-fe68-bef1.loves.ipv6.at.ungleich.ch) | 19:46 | |
| + laumann_ (~quassel@user/laumann) | 19:46 | |
| * laumann_ -> laumann | 19:46 | |
| _hramrach | it plays like 20s of video and then breaks down, previously package download worked but that was with different kernel version. It complains about missing firmware and regdb but it does start initially so that should not be the problem | 19:47 |
| minute | hmm, do we have a regression in qcacld2? | 19:54 |
| _hramrach | the module probably comes from somewhere where it is cached in ramdisk | 19:56 |
| _hramrach | I have the package installed but it puts the module in /opt, not available to the kernel. Linking it to the module directory it still cannot be loaded: qcacld2: no symbol version for module_layout | 19:57 |
| _hramrach | Looks like I should get the special Ethernet cable | 19:59 |
| _hramrach | aargh, now the wifi does not come up at all | 20:01 |
| _hramrach | no, it only takes quite a while | 20:02 |
| Twodisbetter | I have found that the further from the AP and the hotter the device the more frequent the drops. If you move closer to the AP it should improve. | 20:02 |
| + gustav28 (~gustav@c-4135524e.019-141-67626730.bbcust.telenor.se) | 20:02 | |
| minute | _hramrach: that's not how it is intended. it's loaded by /sbin/reform-hw-setup from /opt | 20:04 |
| minute | (because it's an out-of-tree module and we don't want it loaded on any platform) | 20:05 |
| _hramrach | why it can't be installed as standard kernel module? | 20:05 |
| _hramrach | there are hardware IDs for determining which module to load for which device | 20:05 |
| minute | _hramrach: for sdio? | 20:06 |
| _hramrach | surely for SDIO as well, there are people raving how it's discoverable bus | 20:06 |
| minute | _hramrach: how exactly should that work? | 20:07 |
| _hramrach | how does PCI discovery work? | 20:07 |
| _hramrach | you have IDs that are reported by the bus, some platform-specific quirks, .. | 20:08 |
| minute | and how is the module loaded on demand? | 20:08 |
| _hramrach | when it matches the detected ID it is loaded | 20:08 |
| minute | by who? (genuine question, i don't know) | 20:09 |
| _hramrach | modprobe, the kernel passes it some 'module alias' that it tries to resolve to a specific module | 20:09 |
| minute | ah, i see | 20:10 |
| minute | i don't know how to integrate a binary out-of-tree driver into debian's kernel infra though... we had a lot of discussions here that didn't lead to a good solution | 20:11 |
| minute | there's dkms but we only understood how to use it for built-from-source modules. the driver is too big to be compiled from source on imx8mp (it takes 15 minutes or so, don't remember) | 20:12 |
| _hramrach | but also you have device tree so you can add 'this-is-very-special-driver' as the note compatible, and match that | 20:12 |
| minute | btw there's also a mainline driver for this card but it doesn't work and noone knows why | 20:12 |
| minute | (ath10k) | 20:12 |
| _hramrach | since the module is kernel version specific you can ship a hardcoded symlink or directly install to the module directory, and run depmon in postinst/postrm | 20:13 |
| _hramrach | *depmod | 20:13 |
| _hramrach | I understand that Debian with reliance on dkms does not support this very well but I don't see why it would not work | 20:14 |
| minute | if you want to work on this, you're welcome to tinker and send MR | 20:14 |
| _hramrach | adding the symlinkk and script should not be too difficult. Auditing that the hardware IDs are correct is another story | 20:16 |
| _hramrach | that's not exactly my area of expertise | 20:17 |
| Twodisbetter | It is amazing that the Pocket is so good at playing all my favorite games. | 20:41 |
| _hramrach | is this the Ethernet cable https://eu.mouser.com/ProductDetail/HARTING/33480653830002?qs=Li%252BoUPsLEnt9QBrDSqfKbw%3D%3D ? | 20:50 |
| minute | _hramrach: yeah but that one is only 20cm | 20:51 |
| minute | _hramrach: useful only with a coupler | 20:51 |
| minute | 2m by amphenol https://www.mouser.de/ProductDetail/Amphenol-Commercial-Products/ND9-ACN2-B0A?qs=T%252BzbugeAwjiDhhAAwf6CKQ%3D%3D | 20:53 |
| minute | sorry, no. that's a socket on one end | 20:53 |
| minute | we're in the process of buying a bunch of cheap iX to rj45 socket adapters | 20:53 |
| minute | _hramrach: 3m cable https://eu.mouser.com/ProductDetail/HARTING/33480147826030?qs=W%2FMpXkg%252BdQ4Dl9lSDo1Ldg%3D%3D | 20:55 |
| minute | they're filed wrongly under "USB Type A" on one end at mouser | 20:55 |
| _hramrach | I am looking for a cable that I can connect to the normal cable with a cube | 21:01 |
| minute | ah | 21:01 |
| _hramrach | I would need 5m+ cable otherwise | 21:01 |
| minute | makes sense | 21:02 |
| _hramrach | how is the module even loaded when it is not loadable? | 21:27 |
| josch | minute: if i understand this correctly, then the graphite desktop environment and with it gde-background are archived and not developed anymore. So am I assuming correctly that you intend to fork the project and become its new upstream for the lack of a better alternative? | 21:30 |
| erle | yay, i am in the situation that i want to flash my second MNT reform keyboard. | 21:57 |
| erle | \o/ | 21:57 |
| NanoCodeBug | i'm happy with the current state of the revised pocket firmware i've been working on - it works on my battery packs and chargers i've tested against. | 22:20 |
| NanoCodeBug | it has some debug logging if it should encounter a error and folks can capture the serial output of the rp2040. | 22:20 |
| NanoCodeBug | if anyone is feeling adventurous, and knows how to flash their firmware, feel free to try | 22:20 |
| NanoCodeBug | https://source.mnt.re/NanoCodeBug/pocket-reform/-/jobs/5821/artifacts/browse/pocket-reform-sysctl-fw/build/ | 22:20 |
| NanoCodeBug | and let me know how it goes. | 22:20 |
| NanoCodeBug | ah i take that back | 22:24 |
| NanoCodeBug | don't try it just yet | 22:24 |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 22:32 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 22:44 | |
| Twodisbetter | NanoCodeBug: thanks for your efforts! | 22:48 |
| Twodisbetter | Any chance you wanna help out with hibernation? | 22:48 |
| - staticbunny (QUIT: Ping timeout: 260 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 22:48 | |
| Twodisbetter | minute: must a charging source for the pocket be a PD cable charger? If it is a dumb source that can do 5V at up to 4 amps max, would this work for charging the pocket? | 22:56 |
| NanoCodeBug | Twodisbetter: hibernation is going to be tricky. a lot of kernel debugging I imagine, that OS-SoC interaction is not something I know a lot about. | 22:58 |
| ^alex | NanoCodeBug, oh, good, you're doing the sysctl refactor that we were lowkey dreading | 22:58 |
| Twodisbetter | minute: I have a Jackery battery unit that does 5V up to 2.5 amps and it does not seem to work, which is why I am asking. | 22:58 |
| NanoCodeBug | Twodisbetter: the internal charge chip will not accept a voltage below 6 volts | 22:59 |
| Twodisbetter | NanoCodeBug: thank you! That answers my questions nicely. I wonder if there is any chance to improve that, to allow more types of chargers? Or is 6V the minimum voltage to charge both cells? | 23:00 |
| NanoCodeBug | functionally, this means a PD charger is required. | 23:03 |
| NanoCodeBug | that requirement is there to prevent the pocket from feeding back into the charger while powering a USB device. | 23:03 |
| NanoCodeBug | I think it could be changed to allow that, but that would break the usb-pd/usb-device detection. | 23:03 |
| NanoCodeBug | a manually enabled mode could be added - toggled via the keyboard or within the OS - that keeps vbus off, drops the threshold to 5volts, and sets a current limit to 3 amps (its 2 amps by default since its expecting 9-20 volts on the PD rail). | 23:03 |
| NanoCodeBug | ^alex: yeah I saw everything in one file and my brain could not handle that, https://source.mnt.re/NanoCodeBug/pocket-reform has the current code i'm using. | 23:05 |
| ^alex | yeah that's what we were looking at | 23:07 |
| Twodisbetter | NanoCodeBug: That manually activated 5V charging mode would be awesome and improve the ability of the Pocket to charge. | 23:11 |
| NanoCodeBug | Twodisbetter: my current understanding of the charger circuitry is that should be possible, but I have not tested that, not sure if the charge controller would be happy with that either. ideally a external battery pack would bypass the charging circuitry entirely to avoid energy loss charging the batteries. | 23:16 |
| NanoCodeBug | ah, the charger chip refuses to report discharge current and voltage unless vbus is powered, fascinating. | 23:27 |
| ch | NanoCodeBug: any reason you're not using the fusb's toggle logic? | 23:28 |
| ch | the resulting interrupt flags make it very clear what role one -should- be in | 23:29 |
| ch | i need to do some more debugging with my 'interesting' chargers, but so far using the toggle logic was quite easy to do | 23:30 |
| NanoCodeBug | ch: i could not get it work reliably, it would sometimes settle into the wrong state, and it would block PD communication until it was disabled even after exiting - according to the interrupt flags. | 23:30 |
| NanoCodeBug | renabling toggle after disabling it to do PD communication would cause it to refuse to send interrupt flags (so detach would not be shown) until a full reset | 23:30 |
| ch | hmm | 23:31 |
| ch | at least the detach was very reliable for me so far | 23:31 |
| NanoCodeBug | TOGGLE and the interrupts are the ideal solution by far, but I could not get them to behave reliably across the chargers i was testing with. | 23:32 |
| NanoCodeBug | its certainly possible i was dealing with other bugs at the same time that might have been affecting this. | 23:34 |
| ch | ack. i'll hack some more on my fw but not today | 23:37 |
| ch | one thing to keep in mind is testing with non-chargers | 23:37 |
| NanoCodeBug | i have a handful of regular usb-c devices and cables i've plugged in from my menagerie of devices to see what happens, which has helped me find a few issues | 23:38 |
| minute | _hramrach: it is loaded by /sbin/reform-hw-setup | 23:39 |
| minute | _hramrach: https://source.mnt.re/reform/reform-tools/-/blob/main/sbin/reform-hw-setup#L93 | 23:39 |
| josch | minute: gde-background is not working for me -- did you have to change something in your sway config first? | 23:40 |
| minute | josch: you need to patch one line | 23:41 |
| josch | currently i have a line in it that says: output "*" bg ~/.config/sway/wallpaper.jpg fit | 23:41 |
| minute | josch: well you need to disable swaybg | 23:41 |
| josch | hah i got it! | 23:42 |
| josch | thank you :) | 23:42 |
| minute | josch: http://dump.mntmn.com/gde-background.patch.txt | 23:42 |
| minute | (most frontends don't support the dark uri) | 23:42 |
| josch | minute: but gde-background is abandoned upstream, no? I also searched for a similarly minimal alternative but could not find one | 23:43 |
| minute | josch: well, the author recommended it to me on fedi today. i think we could fork & maintain it (and keep it minimal) | 23:45 |
| minute | (just not add any features ^^^^) | 23:45 |
| josch | minute: instead of forking and you maintaining it, how about replacing it by this 70 line python script i just wrote which does the same thing and would be architecture independent and could thus be shipped by reform-tools without needing to compile it with vala? | 23:47 |
| josch | plus, i would maintain it and not you :) | 23:47 |
| minute | josch: oh? | 23:48 |
| minute | josch: well, sure! | 23:48 |
| minute | josch: is it something that talks to swaybg? | 23:48 |
| minute | like, an adapter? | 23:48 |
| josch | it does not support these wallpaper slideshow xmls though | 23:48 |
| ch | .oO( clearly it should be in rust ) | 23:48 |
| minute | well, we don't need to support those | 23:48 |
| minute | ch: lol | 23:48 |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-4135524e.019-141-67626730.bbcust.telenor.se) | 23:49 | |
| josch | minute: no, it doesn't talk to swaybg (or otherwise it would not support wayfire) but instead uses the same mechanism as gde-background and just creates a window and then uses GtkLayerShell to make it the background | 23:49 |
| josch | and in the window it draws an image in a gdk pixbuf | 23:49 |
| minute | josch: nice | 23:56 |
| josch | we were talking about NIH, so i thought i'd contribute to the topic | 23:58 |
| josch | clearly standard nr. 15 from xkcd#927 | 23:58 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
