| - jjbliss (QUIT: Ping timeout: 264 seconds) (~jjbliss@1464766-static.elnsmiaa.metronetinc.net) | 00:50 | |
| + jjbliss (~jjbliss@1464766-static.elnsmiaa.metronetinc.net) | 00:57 | |
| - mtm (QUIT: Ping timeout: 268 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 01:03 | |
| sevan | first time I got the screen glitch on cold boot. There is a wide gap between the wifi aerial and display connector but I can remove the wireless card and aerial to rule that out next. | 01:12 |
|---|---|---|
| sevan | I'm using ethernet atm anyway | 01:12 |
| - vagrantc (QUIT: Ping timeout: 248 seconds) (~vagrant@2600:3c01:e000:21:7:77:0:40) | 02:21 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:40) | 02:26 | |
| flowy | sevan: if i could see the glitch, i could tell you if it was what i was experiencing. they totally went away after shielding the display cable | 02:47 |
| flowy | it was also quite easy to correlate with wifi activity in my case. though certain networks would not trigger it at all. | 02:48 |
| sevan | for me it only happens during initial startup at u-boot, the kernel blanks the screen when it starts and everything is ok after that. | 02:48 |
| flowy | ah. you're running u-boot w/ patches for the reform display? | 02:50 |
| sevan | yeah, from last week (haven't got the recent changes) | 02:50 |
| sevan | https://geeklan.co.uk/files/tmp/reform-screenglitch.mov | 02:50 |
| flowy | ooo nice glitches | 02:52 |
| sevan | :D | 02:53 |
| flowy | yeah, mine were a bit similar in the sense that they flashed like that | 02:54 |
| flowy | like i resorted to shielding because it was actually kind of disturbing/borderline epileptic causing while trying to focus on reading or whatever. | 02:55 |
| flowy | but i'm not sure it was quite the same actually | 02:55 |
| sevan | mine doesn't happen everytime, and it was mostly on rebooting after a session. Today was the first time it did it on cold boot. | 02:56 |
| flowy | so during regular use you don't get that? | 02:57 |
| sevan | correct | 02:57 |
| flowy | i would get it in sway | 02:57 |
| sevan | ah, I'm running Gnome | 02:57 |
| flowy | but just during heavy network/wifi usage and on certain networks | 02:57 |
| flowy | i guess networks that required high signal strength or something | 02:57 |
| sevan | wifi is off, I've been using ethernet the whole time. | 02:58 |
| sevan | are you using the atheros card? | 02:58 |
| flowy | yep, w/ the antenna that shipped with it | 02:59 |
| sevan | me too. | 02:59 |
| flowy | you should try using your wifi on a sort of weak network and then doing a transfer | 02:59 |
| sevan | in sway | 03:00 |
| flowy | i guess | 03:00 |
| sevan | ack | 03:00 |
| sevan | will add it to the list to try. | 03:00 |
| flowy | i'd imagine you could also just route the display cable differently | 03:01 |
| flowy | rather than apply shielding | 03:01 |
| flowy | if your issue is wifi interference | 03:01 |
| flowy | but i guess it's not a big deal if it doesn't happen during normal usage | 03:02 |
| flowy | i'm looking forward to trying this u-boot display support | 03:02 |
| flowy | very nice | 03:02 |
| sevan | clone https://source.mnt.re/reform/reform-boundary-uboot and build | 03:04 |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 03:10 | |
| sevan | flowy: thanks for the tips, I'll give them a try and report back. | 03:32 |
| + bgs (~bgs@212-85-160-171.dynamic.telemach.net) | 06:39 | |
| - bgs (QUIT: Remote host closed the connection) (~bgs@212-85-160-171.dynamic.telemach.net) | 07:41 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:40) | 08:57 | |
| - mtm (QUIT: Ping timeout: 260 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 09:00 | |
| + MajorBiscuit (~MajorBisc@145.94.153.3) | 10:32 | |
| josch | minute: i'd like the situation with the unsigned/untrusted reform apt repo see solve for when the next Debian stable release happens in a few months. Are you planning to set up a system that you can trust with a GPG key such that the repo can be signed? If you do not, then I could set up a builder that only builds stable with security updates and publishes a repo signed with my Debian key on | 12:57 |
| josch | https://people.debian.org/~josch/reform or similar. | 12:57 |
| minute | josch: i think i can set up signing, just need to think about how to do it in the most practical way | 13:00 |
| josch | okay, cool! | 13:00 |
| minute | josch: are there best practices for automating this? | 13:00 |
| josch | if we manage to do so within the next few weeks, then the repository could be included in the extrepo package and thus be easy to enable from Debian | 13:00 |
| josch | some of the software packages that create repositories (like reprepro) can also do signing | 13:01 |
| josch | i'm not aware of a write-up of best-practices | 13:01 |
| josch | essentially, in the end it runs gpg --sign and that's it | 13:02 |
| josch | obviously, the key must not have a passphrase | 13:02 |
| josch | ah, one best-practice i can think of might be to use a sub-key | 13:03 |
| josch | that way, even if the key gets compromised, you do not have to throw away your main key | 13:03 |
| josch | https://wiki.debian.org/Subkeys | 13:03 |
| josch | i have my main gpg key on a usb stick buried in the garden somewhere and use a sub key on my machines, for example | 13:03 |
| minute | ok | 13:25 |
| sevan | you can use Yubikeys as a GPG security token and they support 4096 bit gpg keys, if you're looking for cheap hardware | 13:34 |
| sevan | you can also reflash a SM32 controller with firmware if you want a really cheap security token :) | 13:34 |
| sevan | (for use with GPG) | 13:35 |
| q66 | ACTION keeps it simple by having no gpg key :) | 13:35 |
| q66 | minute: btw, ever considered offering some kind of standalone board so that one can do development/test things without having a full reform? | 13:36 |
| q66 | would be useful for me since I'd like to support the hardware but have no need for an extra laptop | 13:36 |
| + bgs (~bgs@212-85-160-171.dynamic.telemach.net) | 13:37 | |
| sevan | s/SM32/STM32 | 13:37 |
| sevan | q66: :) | 13:37 |
| josch | q66: creating a carrier board in itx factor came up a couple of times already, also see here: https://twitter.com/minut_e/status/1618226875167567873 | 13:44 |
| q66 | itx is perhaps a little large | 13:45 |
| q66 | but would work too | 13:45 |
| q66 | itx also only really makes sense if you plan to mount it in a case | 13:45 |
| q66 | and then you also want atx power supply | 13:45 |
| q66 | while something like usb-c or barrel plug would be much more convenient | 13:46 |
| q66 | but yeah if something gets made i would be interested in buying one, i have enough laptops but i am also deep enough into the devboards rabbit hole to disregard how many of them i have already | 13:48 |
| flowy | i think i've seen itx/matx motherboards with multiple power options incl barrel jack? also there's picopsu, i dunno how great that solution is though | 14:18 |
| flowy | i guess itx makes a selling point for attracting those who don't want DIY enclosures | 14:19 |
| flowy | i'm excited to build a proper system out of my old reform soc, probably as a router. i wonder if anyone has tried openwrt on the reform yet | 14:20 |
| flowy | i'm tempted to order reform MB v2 and also use the old reform MB for such a system. i would be into the opportunity for designing a custom case, would be fun to learn some cad | 14:23 |
| q66 | i wouldn't put mine in a case because that would be missing the point of having one, like, i want the som and all I/O directly accessible without having to reach into a case | 14:25 |
| flowy | i'm imagining a case w/o keyboard and display but still has room for the battery boards. i figure, it might actually be nice to keep the LIPO chemistry for power backup, vs putting it on my nasty lead acid UPS | 14:25 |
| flowy | q66: which I/O are you thinking of? couldn't everything be conveniently exposed w the right carrier/case? | 14:27 |
| q66 | how do you want to conveniently expose mipi dsi for example | 14:29 |
| q66 | besides, with a devboard that you do not transport case adds bulk that is mostly pointless | 14:31 |
| flowy | i see | 14:31 |
| q66 | uart tends to be onboard pins and it's best that way | 14:32 |
| q66 | some boards put it into a trs jack or a usb and that's just annoying | 14:33 |
| sevan | I have a galileo board that does that. horrible | 14:57 |
| q66 | pine64 boards use the trs jack | 14:58 |
| q66 | i made my own adapter for that | 14:58 |
| q66 | the most annoying part is that it overlaps with audio so you gotta flip a dip switch every time | 14:59 |
| q66 | especially on like a pinebook pro it involves taking off the bottom panel | 14:59 |
| sevan | makes me think of the raspberry pi cases which would enclose the entire thing without access to the gpio pins :) | 15:01 |
| sevan | so if you want to tinker you have to take the case apart | 15:02 |
| q66 | in this way i like the rpi400 | 15:02 |
| q66 | it's enclosed so you can take it to places, has a keyboard builtin and you have the whole gpio panel in the back under a rubber cover | 15:03 |
| sevan | nice | 15:03 |
| q66 | and it's still lightweight and compact | 15:03 |
| q66 | but yeah these rpi cases are not very good for tinkering in general, i guess their main intent is for like, if somebody uses rpi's in prod and is not going to touch them | 15:05 |
| q66 | for home automation nonsense and whatnot | 15:05 |
| sevan | yeah, it is very confusing since even the official red and white cases from rpi foundation were sealed (wasn't tinkering on the agenda?) | 15:08 |
| q66 | a lot of people incl companies now use rpi's as prod devices despite them being terrible for that | 15:09 |
| q66 | i do consultancy and I came across stuff like digital signage companies using them to drive displays and expecting wonders out of it | 15:10 |
| sigrid | re development/testing: I know cinap has his embedded stuff mounted in a rack, powered on/off remotely http://felloff.net/usr/cinap_lenrek/rack2.jpg | 15:12 |
| sigrid | and this, too: https://www.linux-automation.com/en/products/usb-sd-mux.html | 15:13 |
| sevan | the board is on top of the PoE switch in the middle of the photo? | 15:14 |
| sigrid | it's a lot of stuff on the top shelf | 15:14 |
| sevan | ah :) | 15:14 |
| sigrid | all kinds of raspis, aijuboard, idk where reform is | 15:14 |
| sigrid | josch: btw regarding that (some time ago) discussion about nvme reporting "unsafe shutdowns" | 15:16 |
| sigrid | if it matters I confirmed the theory by adding a proper nvme shutdown on 9front when fshalt (basically "sudo poweroff") is triggered and it got rid of unsafe shutdowns | 15:17 |
| sigrid | the fix involved adding code to nvme driver logic (which is definitely not missing in linux) and telling the kernel to disable nvme controller from userspace after filesystems are unmounted | 15:18 |
| sigrid | maybe in linux there is also a way to trigger controller shutdown from userspace, somehow? that could help | 15:19 |
| josch | sigrid: thanks a lot for confirming! | 15:42 |
| josch | it would be great to get rid of this problem | 15:43 |
| josch | but i'd first research how to do that in linux :/ | 15:43 |
| minute | q66: josch: we put a funding/grant proposal for reform mITX board in with nlnet, lets see if it works out | 15:54 |
| minute | we put in two proposals for the deadline that was a few days ago | 15:54 |
| minute | the idea for the mITX board is also to be able to take up to 4 modules, internally networked | 15:55 |
| minute | i hear you when you say bulky power supply etc... one could make USB-PD an alternative power input with the ports, for example | 15:55 |
| minute | i'm very open for ideas. but we need to see if we get some funding | 15:55 |
| minute | the other project is rk3588 adapter and "thin" reform (a mix of pocket and reform parts and a new enclosure | 15:56 |
| minute | q66: OTOH there is a tiny dev board from jacqueline, but it does not have all the peripherals of the reform motherboard | 16:08 |
| - qbit (QUIT: Quit: WeeChat 3.7.1) (~qbit@h.suah.dev) | 17:32 | |
| + qbit (~qbit@h.suah.dev) | 17:33 | |
| - MajorBiscuit (QUIT: Quit: WeeChat 3.6) (~MajorBisc@145.94.153.3) | 17:34 | |
| ysionneau | ACTION on his way to fosdem o/ | 18:42 |
| ysionneau | minute: will you have some prototypes with you ? :) | 18:43 |
| sevan | speaking of boards, racks, and modules, this went by on social media https://www.kickstarter.com/projects/uptimelab/compute-blade | 19:14 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:40) | 19:24 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 19:37 | |
| qbit | someone say boards, racks and modules?! I am here now! | 21:28 |
| minute | ysionneau: yep, pocket reform prototype | 22:30 |
| violet | maybe a weird question but what are the chances I could find a 1366x768 panel that'd I could succesfully replace the reform's 1080p panel with | 22:35 |
| violet | idk how standard these things are | 22:36 |
| violet | (these things being LCD panels in general) | 22:36 |
| klardotsh | most LCD panels use ribbon cables with a different termination if I recall correctly, but if you could find one that uses the almost-IDE-like termination Reform's panel uses there's no reason I can think of it shouldn't work. lemme see if I can find the panel site I once looked at looking for relatively similar things (I'd considered trying to find something just slightly lower res once to | 22:37 |
| klardotsh | avoid 125% weird fractional scaling) | 22:37 |
| violet | yeah avoiding fractional scaling is exactly the reason I want to do it lol | 22:38 |
| klardotsh | ah, yeah, this looks like the site I once used. it's... deeply technical, but also has basically everything. https://www.panelook.com/sizmodlist.php?sizes[]=1330 | 22:38 |
| violet | much appreciated | 22:39 |
| klardotsh | there's some neat 1600x900 panels that are probably a reasonable res at that size, many discontinued tho. the LP133WD2-SPB1 for example runs about $120 USD on eBay (not certain it'd actually work on Reform either) | 22:40 |
| josch | Linux reform 6.1.0-reform2-arm64 #1 SMP Debian 6.1.8-1+reform20230131T054557Z1 (2023-01-31) aarch64 GNU/Linux | 22:46 |
| josch | linux 6.1.8 seems to work fine | 22:46 |
| josch | just did a full unstable upgrade and things seem to keep working fine | 22:46 |
| q66 | 1080p seems like it'd be ideal for 12.5" without scaling | 22:51 |
| q66 | i have 3840x2400 on 13.4" and scale it at 150% | 22:52 |
| q66 | 1080p would be bigger than that | 22:52 |
| q66 | 768p is just uncomfortable to use imo | 22:53 |
| q66 | not enough space | 22:53 |
| violet | that's most people's opinions | 23:00 |
| violet | but 1366x768 is my preference | 23:00 |
| - cwebber (QUIT: Read error: Connection reset by peer) (~user@user/cwebber) | 23:02 | |
| violet | extra pixels do me little good when everything is too small for my feeble eyes to comprehend | 23:04 |
| JC[m] | once we discover compatible SKUs, it would be good to document them in a hardware compatibility list | 23:04 |
| violet | that'd be wonderful yes | 23:05 |
| + cwebber (~user@user/cwebber) | 23:06 | |
| JC[m] | I created the reform wiki a while back but need to find more free time to curate it | 23:06 |
| JC[m] | I'll need to sync with bnys and minute to establish what's appropriate for the forum vs wiki vs handbook so that information is available and maintainable over time | 23:11 |
| JC[m] | but I don't intend to hijack the conversation so I'll shut up now :) | 23:11 |
| violet | klardotsh: i still dont know what im doing but im looking at the list of panels with the same connector type. im finding some 12.5" panels from some other manufacturer, but they're slightly wider so idk if it'd fit. but im also finding some 11.6" panels on here, including some from innolux, and of those there are people selling replacements for that innolux model for like $30 on ebay. i think i wouldnt | 23:42 |
| violet | mind just adding a bit more bezel | 23:42 |
| violet | so maybe I'll give it a go | 23:42 |
| - bgs (QUIT: Remote host closed the connection) (~bgs@212-85-160-171.dynamic.telemach.net) | 23:53 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
