| + B[] (~Thunderbi@122-61-176-31-fibre.sparkbb.co.nz) | 02:41 | |
| - Jookia (QUIT: Remote host closed the connection) (~Jookia@gateway/tor-sasl/jookia) | 03:15 | |
| + Jookia (~Jookia@gateway/tor-sasl/jookia) | 03:16 | |
| + esad (~textual@84-114-111-23.cable.dynamic.surfer.at) | 09:08 | |
| - esad (QUIT: Quit: My MacBook has gone to sleep. ZZZzzz…) (~textual@84-114-111-23.cable.dynamic.surfer.at) | 12:16 | |
| + esad (~textual@84-114-111-23.cable.dynamic.surfer.at) | 12:17 | |
| - esad (QUIT: Quit: My MacBook has gone to sleep. ZZZzzz…) (~textual@84-114-111-23.cable.dynamic.surfer.at) | 12:59 | |
| + esad (~textual@84-114-111-23.cable.dynamic.surfer.at) | 13:09 | |
| - esad (QUIT: Quit: My MacBook has gone to sleep. ZZZzzz…) (~textual@84-114-111-23.cable.dynamic.surfer.at) | 14:05 | |
| + esad (~textual@84-114-111-23.cable.dynamic.surfer.at) | 14:06 | |
| - esad (QUIT: Read error: Connection reset by peer) (~textual@84-114-111-23.cable.dynamic.surfer.at) | 14:09 | |
| + esad (~textual@84-114-111-23.cable.dynamic.surfer.at) | 14:09 | |
| + Perkol (~Perkol@46.53.242.169) | 14:44 | |
| - esad (QUIT: Ping timeout: 265 seconds) (~textual@84-114-111-23.cable.dynamic.surfer.at) | 15:28 | |
| - B[] (QUIT: Ping timeout: 240 seconds) (~Thunderbi@122-61-176-31-fibre.sparkbb.co.nz) | 15:50 | |
| - Perkol (QUIT: Remote host closed the connection) (~Perkol@46.53.242.169) | 16:10 | |
| + esad (~textual@84-115-64-174.wifi.dynamic.surfer.at) | 16:32 | |
| - esad (QUIT: Ping timeout: 240 seconds) (~textual@84-115-64-174.wifi.dynamic.surfer.at) | 17:28 | |
| + esad (~textual@84-115-64-174.wifi.dynamic.surfer.at) | 17:29 | |
| - esad (QUIT: Ping timeout: 265 seconds) (~textual@84-115-64-174.wifi.dynamic.surfer.at) | 18:29 | |
| + esad (~textual@84-115-65-215.wifi.dynamic.surfer.at) | 18:32 | |
| - esad (QUIT: Ping timeout: 276 seconds) (~textual@84-115-65-215.wifi.dynamic.surfer.at) | 19:28 | |
| + esad (~textual@84-115-64-14.wifi.dynamic.surfer.at) | 19:28 | |
| - esad (QUIT: Ping timeout: 265 seconds) (~textual@84-115-64-14.wifi.dynamic.surfer.at) | 20:28 | |
| + esad (~textual@84-115-64-213.wifi.dynamic.surfer.at) | 20:31 | |
| - wiedi_ (QUIT: Ping timeout: 245 seconds) (~wiedi@ip5b4096a6.dynamic.kabel-deutschland.de) | 21:07 | |
| + wiedi (~wiedi@ip5b4096a6.dynamic.kabel-deutschland.de) | 21:07 | |
| - wiedi (QUIT: Ping timeout: 240 seconds) (~wiedi@ip5b4096a6.dynamic.kabel-deutschland.de) | 21:25 | |
| + wiedi (~wiedi@ip5b4096a6.dynamic.kabel-deutschland.de) | 21:25 | |
| - esad (QUIT: Ping timeout: 276 seconds) (~textual@84-115-64-213.wifi.dynamic.surfer.at) | 21:27 | |
| + esad (~textual@84-115-64-213.wifi.dynamic.surfer.at) | 21:30 | |
| + wiedi_ (~wiedi@ip5b4096a6.dynamic.kabel-deutschland.de) | 21:44 | |
| - wiedi (QUIT: Ping timeout: 265 seconds) (~wiedi@ip5b4096a6.dynamic.kabel-deutschland.de) | 21:45 | |
| - esad (QUIT: Quit: My MacBook has gone to sleep. ZZZzzz…) (~textual@84-115-64-213.wifi.dynamic.surfer.at) | 21:45 | |
| + marcodiego (~pi@177.37.130.253) | 22:32 | |
| marcodiego | Considering https://stikonas.eu/wordpress/2019/09/15/blobless-boot-with-rockpro64/ , https://www.collabora.com/news-and-blog/blog/2019/08/06/rockpi-panfrost-wayland/ , https://wiki.radxa.com/Rockpi4/hardware/rockpi4 and https://media.ccc.de/v/2019-209-challenge-blob-free-mainline-open-source-software-on-a-single-board-computer-sbc-arm64-foss-efi- wouldn't it be a good idea to use rk3399 on the mnt? | 22:35 |
|---|---|---|
| Jookia | marcodiego: it already uses the imx8 | 22:40 |
| marcodiego | Jookia, I mean... imx8 needs proprietary blobs right? I think it is possible to run the rk3399 with no blobs | 22:40 |
| Jookia | marcodiego: there's always blobs | 22:43 |
| Jookia | the page you link even says 'there seems to be a small binary blob responsible for DisplayPort DRM' | 22:43 |
| marcodiego | Jookia, and it also says: "I have removed it in my fork of ATF (actually replaced it with a file of 0 size)" | 22:44 |
| Jookia | hardware is a big blob too | 22:44 |
| Jookia | the imx8 which the reform uses requires a blob for hdmi and training ddr4 at boot | 22:45 |
| mntmn | yeah and HDMI is optional... the main display path is blob free, already validated | 22:46 |
| mntmn | i think imx8m is the better choice also because it will have a long life, well documented, made by a company that has a long track record, at least has some CSR standards etc | 22:47 |
| marcodiego | and I think a rk3399 based one wouldn't need any | 22:47 |
| Jookia | I guess if all you care about is blobs rk3399 would be a good choice at the moment | 22:48 |
| marcodiego | mntmn, hmm... didn't thought about the track record of the manufacturer. Makes sense | 22:48 |
| marcodiego | how does the imx8 compares to rk3399 performance-wise? | 22:49 |
| Jookia | marcodiego: i think hardware developers end up being less concerned about blobs than software people, given that hardware at this point is just big blob chips with no source code | 22:49 |
| mntmn | also, (at least from my perspective), very little is known about rockchip as a company vs NXP... even their about page is pretty empty | 22:51 |
| Jookia | mntmn: maybe that's just in the english internet | 22:52 |
| mntmn | Jookia: possible | 22:52 |
| marcodiego | the mnt reform will feature a user upgradeable SOM? | 22:53 |
| Jookia | it is worrying that the imx8 can't have HDMI without blobs though :( | 22:53 |
| mntmn | marcodiego: i'm not sure about the perf. i guess the A72s should be faster than the A53s. i do have a rk3399 board here but i never even turned it on so far | 22:53 |
| mntmn | Jookia, usb3 hdmi could be possible for example | 22:53 |
| mntmn | depending on your use case | 22:54 |
| Jookia | neat | 22:54 |
| mntmn | also, the blob is (IMHO) not that different from a ROM. it's not ARM code | 22:54 |
| Jookia | marcodiego: the reform 2 has a user upgradable SOM i think | 22:54 |
| mntmn | it's code that would traditionally be embedded in the silicon | 22:54 |
| mntmn | Jookia, marcodiego yes the som is upgradable | 22:55 |
| mntmn | so someone could definitely make a rk3399 module for it | 22:55 |
| mntmn | or a risc-v module if a fitting chip comes out ;) | 22:55 |
| marcodiego | cool! Happy to know that | 22:55 |
| Jookia | i like the modular aspect since it means new core != new laptop board layout | 22:55 |
| mntmn | yeah, there are some restrictions but not many | 22:56 |
| marcodiego | is the SOM on a standard format (if such thing exists) | 22:56 |
| marcodiego | ? | 22:56 |
| mntmn | like, the chip would need to do USB3 to be able to use the USB ports, and either eDP or DSI for internal display | 22:56 |
| mntmn | marcodiego, not really apart from it being SO-DIMM | 22:56 |
| mntmn | but it's not standard like MXM or COM or such (these are normally quite big though) | 22:57 |
| Jookia | mntmn: how much would i break the board if i powered it on with a RAM stick inserted | 22:57 |
| mntmn | Jookia: good question, i don't know the pinout of those by heart ;) | 22:57 |
| marcodiego | the green board in the middle https://files.mastodon.social/media_attachments/files/018/754/231/original/891ad8bb4e73e76d.jpeg is the SOM, is that? | 22:57 |
| mntmn | marcodiego: yep | 22:58 |
| marcodiego | cool! | 22:58 |
| mntmn | this is Nitrogen 8M SOM | 22:58 |
| mntmn | by boundary devices | 22:58 |
| Jookia | I think that might be the same SOM used in the librem 5 devkit, but in a different form factor | 22:58 |
| mntmn | no, they used an emcraft som | 22:58 |
| mntmn | but same chip | 22:58 |
| Jookia | https://boundarydevices.com/product/nitrogen8m/ | 22:59 |
| mntmn | i chose nitrogen also because you can download the schematics on their website | 22:59 |
| mntmn | Jookia: that is a different version (single board computer) | 22:59 |
| Jookia | that's very helpful | 22:59 |
| mntmn | this is the module https://boundarydevices.com/product/nitrogen8m-som/ | 22:59 |
| Jookia | i just pasted the link before it could load because it froze on my computer | 22:59 |
| mntmn | :D | 23:00 |
| Jookia | spectre affects humans too | 23:00 |
| mntmn | haha | 23:00 |
| mntmn | also i'm looking forward for future i.MXes | 23:00 |
| mntmn | for -> to | 23:00 |
| Jookia | Why's that? | 23:01 |
| mntmn | i've seen the roadmap and it looked quite good :) | 23:01 |
| mntmn | so there will be some upgrade potential | 23:01 |
| Jookia | I think the imx6 stops fabrication in 2020 or 2024 | 23:01 |
| mntmn | i could imagine they'll prolong it though | 23:02 |
| mntmn | imx6 is quite popular still in industrial | 23:02 |
| swivel | mntmn: what's the ribbon cable for? | 23:02 |
| mntmn | swivel: it feeds DSI from the som to the board | 23:02 |
| mntmn | (it will be much shorter, this is just one that i had to hack up) | 23:03 |
| marcodiego | is there a complete prototype yet? | 23:03 |
| Jookia | i think mntmn is currently making the prototype for the reform 2? | 23:03 |
| mntmn | almost. i think rev2 will be the first one considered complete | 23:03 |
| Jookia | he's posting lots of updates on mastodon | 23:03 |
| swivel | ah | 23:04 |
| Jookia | uh, mntmn is posting lots of updates on mastodon | 23:04 |
| mntmn | marcodiego, there are 13 complete reform 1s but the reform2 is a redesign, i just validated the first motherboard rev | 23:04 |
| mntmn | so now i have a list of changes for rev2 | 23:04 |
| Jookia | mntmn: so what's happened to the reform 1s? | 23:04 |
| mntmn | Jookia, well, 1 is at crowd supply HQ, 10 are with beta testers who bought them, 2 are in our office :) | 23:05 |
| Jookia | nice | 23:05 |
| mntmn | reform 2 exists as a bunch of case prototypes, one mostly working motherboard and one keyboard | 23:05 |
| Jookia | you can see them at https://mastodon.social/@mntmn/102786341049801261 | 23:06 |
| mntmn | oh yeah! | 23:06 |
| Jookia | it's a bold design going with non-standard keyboard layout, but reminds me of retrocomputers | 23:06 |
| mntmn | yeah | 23:07 |
| mntmn | i mean i also work on retro computers, so there... might be an influence | 23:07 |
| Jookia | a good influence i think | 23:07 |
| Jookia | the US ANSI keyboard layout isn't exactly the best | 23:07 |
| Jookia | i'm a little surprised you haven't put arrows on the hjkl though ;) | 23:08 |
| mntmn | haha | 23:08 |
| mntmn | well, i'll have to out myself as an emacs user | 23:09 |
| mntmn | (actually i use both vim and emacs) | 23:09 |
| Jookia | that explains all the extra keys | 23:09 |
| mntmn | (but as a heretic i use the arrow keys in vim) | 23:09 |
| Jookia | I am glad to see better placement of modifiers, especially replacing the capslock with ctrl | 23:10 |
| mntmn | yeah. the amiga keyboard on my desk here has the ctrl also in that correct place | 23:11 |
| Jookia | will the reform have a free pcie slot or something for expansion? | 23:13 |
| Jookia | aka FPGA slot | 23:13 |
| Jookia | though maybe USB3 is fast enough for that kind of stuff now | 23:15 |
| mntmn | yeah i mean it depends what else you need in the slots. it has a mPCIe slot and a M.2 PCIe slot. | 23:15 |
| mntmn | you could have a very small usb3 wifi stick in a usb port, then the mPCIe would be free for such things | 23:15 |
| Jookia | yeah that's a fair tradeoff | 23:15 |
| Jookia | oh, what's the 'hack the planet' pinout | 23:16 |
| mntmn | you could also put this in the m.2 slot https://www.crowdsupply.com/rhs-research/nitefury | 23:16 |
| Jookia | is it a floppy driver connector | 23:17 |
| mntmn | Jookia: haha. no it's just a bunch of I/O pins from the onboard LPC11U24 mcu | 23:17 |
| Jookia | I love IO pins | 23:18 |
| mntmn | including one A/D pin | 23:18 |
| mntmn | ah no two actually | 23:19 |
| Jookia | nice | 23:19 |
| mntmn | they are 10-bit, 0 to 3.6v, conversion time 2.44us | 23:20 |
| marcodiego | the reform will have a user programmable mcu? Like a built-in arduino? | 23:20 |
| mntmn | marcodiego: yes, pretty much like that | 23:20 |
| marcodiego | cool! | 23:20 |
| marcodiego | on the motherboard or on the SOM? | 23:20 |
| mntmn | on the motherboard | 23:21 |
| mntmn | here is the manual https://seeeddoc.github.io/Seeeduino_Arch/res/LPC11UxxUserManual.pdf | 23:21 |
| Jookia | mntmn: oh, it's free to use? | 23:21 |
| mntmn | (it's LPC11U24) | 23:21 |
| mntmn | Jookia, well you should leave some of the code intact, as it will control the battery charging | 23:21 |
| mntmn | but it has a lot more space for user applications | 23:22 |
| marcodiego | great! Seems like a very advanced project. Only downside I can point at are the required proprietary binary blobs | 23:22 |
| mntmn | marcodiego: correct | 23:22 |
| mntmn | an upside is that the video decoder is now blob free | 23:22 |
| marcodiego | hmm... hantro, right? | 23:23 |
| mntmn | yep | 23:23 |
| Jookia | i think the only thing blobby that's required is the lpddr4 training? | 23:23 |
| marcodiego | since there will be one more revision... is it possible to make it compatible with eoma68? | 23:24 |
| + erlehmann_ (~erlehmann@46.114.36.150) | 23:24 | |
| - erlehmann (QUIT: Disconnected by services) (~erlehmann@46.114.36.171) | 23:24 | |
| Jookia | marcodiego: i think you mean having the eoma68 as a module? i dont think that would work well since it uses direct LCD connectors | 23:26 |
| Jookia | there's also not enough pins for the imx8 i dont think | 23:26 |
| marcodiego | no, I mean... suppose I have a eoma68 card, would it be possible to attach it to the reform (instead of the nitrogen SOM)? | 23:28 |
| mntmn | marcodiego: no, that won't be something i'd do | 23:28 |
| mntmn | the form factor of eoma68 is a pcmcia module with very limited space, this wouldn't even work heat-wise | 23:29 |
| mntmn | the opening paragraph here doesn't inspire confidence https://elinux.org/Embedded_Open_Modular_Architecture/EOMA68 | 23:30 |
| * erlehmann_ -> erlehmann | 23:30 | |
| Jookia | I bought the EOMA68 when it was on crowdsupply and it still seems to be in development hell | 23:31 |
| mntmn | yeah | 23:32 |
| marcodiego | a SOM based laptop looks more reasonable | 23:33 |
| marcodiego | not as easily upgradeable, not portable but more reasonable | 23:33 |
| mntmn | from my POV it makes more sense to carry around your data between CPUs than your CPU between shells | 23:34 |
| Jookia | I guess the implication is that e-waste will happen if you upgrade and junk the old machine | 23:35 |
| mntmn | yeah, i get that | 23:36 |
| mntmn | i work also with 30 year old computers ;) | 23:37 |
| Jookia | i'm under the impression those don't have BGAs to make user servicing nigh impossible | 23:37 |
| mntmn | yes, that didn't exist | 23:37 |
| mntmn | i think a better form factor for something like eoma could be HDMI out + USB in, then you could plug it into any old tv and plug in any old keyboard + mouse | 23:38 |
| mntmn | well, that's a raspberry pi then | 23:38 |
| Jookia | well, the next usb should be able to handle screens | 23:39 |
| Jookia | so maybe we'll just end up having usb as everything | 23:39 |
| mntmn | true, usb4 will solve this | 23:39 |
| Jookia | plug your PC tower in to an array of 20 USB ports | 23:39 |
| Jookia | maybe it's time to have slower computers that can be user serviced again | 23:40 |
| mntmn | yeah, or maybe we can hit an equilibrium like with audio quality | 23:41 |
| mntmn | you mostly don't need something better than cd audio quality, that's what i mean | 23:41 |
| mntmn | and i rarely need more than full HD or maybe 4k if i'm fancy | 23:41 |
| mntmn | so there will be a point where compute + bandwidth will be enough for most tasks | 23:42 |
| Jookia | then we can have more SOM designs with hand solderable everything else | 23:42 |
| mntmn | yeah | 23:43 |
| mntmn | and that can help with e-waste etc | 23:43 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
