| 30 May 2025 |
 K900 | That expects to be booted by something else | 15:06:08 |
 emily | your flake seems busted(?) btw
error:
… while updating the lock file of flake 'gitlab:K900/nix/51c4148b3c3568b93388e2ebff729ea37f3642c8?narHash=sha256-Q/%2BM5veEoDMyMSm2CT73RVX73zrPImv23%2BdTvz9V6UI%3D'
… while updating the flake input 'firefox'
… while updating the flake input 'firefox/lib-aggregate'
error: cannot find flake 'flake:lib-aggregate' in the flake registries
| 15:06:09 |
| K900 | Ugh | 15:06:42 |
| K900 | It's a cppnix bug | 15:06:44 |
| K900 | It works on cppnix main I think? | 15:07:00 |
| emily | lol okay | 15:07:07 |
| K900 | And Lix something older than cppnix main | 15:07:08 |
| emily | --override-input firefox nixpkgs problem solved | 15:07:24 |
| K900 | I can 100% tell you it works on Lix main | 15:07:27 |
| K900 | Oh yeah also I think it might borf because some of the gitlab repos it actually pulls might be private? | 15:07:46 |
| K900 | I don't remember how much of that shit I set to publiv | 15:07:54 |
| K900 | * I don't remember how much of that shit I set to public | 15:07:55 |
| emily | this is truly an installation procedure of all time | 15:08:53 |
| emily | can't wait until I have this stuff in my own repo that nobody else can consume instead :P | 15:09:02 |
| K900 | I mean why wouldn't you want a 100% organic home grown Rust based policy routing implementation that relies entirely on assumptions about the ambient system | 15:09:51 |
| emily | I can think of no good reason. | 15:12:09 |
| emily | I'm a bit concerned that I'm about to hit a brick wall learning curve of making any of this work at all. | 15:12:35 |
 ElvishJerricco | well, you can't use dd. You have to do the whole "erase the flash before you write it" dance | 16:54:10 |
| K900 | Uhh do you? | 16:54:23 |
| K900 | I think dd worked for me | 16:54:27 |
| K900 | But I don't remember at this point | 16:54:50 |
| ElvishJerricco | I have flash_erase /dev/mtd0 0 0 followed by nandwrite --pad /dev/mtd0 ./uboot.img in my notes because it was booting if I just overwrote it | 16:54:53 |
| ElvishJerricco | * I have flash_erase /dev/mtd0 0 0 followed by nandwrite --pad /dev/mtd0 ./uboot.img in my notes because it wasn't booting if I just overwrote it | 16:55:02 |
| K900 | Did you write to /dev/mtd0 or /dev/mtdblock0 | 16:55:17 |
| ElvishJerricco | mtd0 | 16:55:24 |
| K900 | I think /dev/mtdblockN should automatically do the erase dance | 16:55:28 |
| ElvishJerricco | ahh ok | 16:55:34 |
| ElvishJerricco | anyway emily I did figure out how to format an SD card to get it booting off that instead of the NAND if you want | 16:56:36 |
| emily | yeah I'm interested | 16:57:16 |
| emily | I'll probably want to do NAND + M.2 long-term but right now I kinda just want to get something booting | 16:57:26 |
