| 29 Oct 2024 |
 atemu12 | Huh, so I am not alone in this | 00:38:18 |
| atemu12 | Yeah this is definitely a software issue; some state getting corrupted | 00:38:31 |
| atemu12 | You can get it to work by clearing userdata and I think I had to wipe system too | 00:39:03 |
 oak 🏳️🌈♥️ | The Robotnix just probably ends up using the same stuff from vendor repository | 00:39:20 |
| atemu12 | Yup, TheMuppets | 00:39:40 |
| oak 🏳️🌈♥️ | By inspecting logcat logs, fingerprint related issues could be probably found | 00:40:18 |
| atemu12 | Though we have to infer which vendor dirs we need from device names alone, so we might accidentally be missing some | 00:40:35 |
| atemu12 | I only have vendor/fairphone/FP4 | 00:41:32 |
| oak 🏳️🌈♥️ | Below the Android Java stack there's low level Unix daemon fingerprintd | 00:41:36 |
| oak 🏳️🌈♥️ | On some phones it needs some calibration files for the camera from some of the partitions | 00:42:22 |
| atemu12 | Thank you so much, that'll already give me something to poke at tomorrow. Do you know whether any of these hold significant state in userdata? | 00:43:02 |
| oak 🏳️🌈♥️ | I would check if the fingerprintd fails to start or something | 00:43:56 |
| oak 🏳️🌈♥️ | It's also somewhat related to unlocking the Android Keystore | 00:44:43 |
| oak 🏳️🌈♥️ | Probably read from the source how it decides whether fingerprint options are shown in the settings | 00:45:48 |
| atemu12 | In reply to @oak:universumi.fi
It's also somewhat related to unlocking the Android Keystore Interesting, yeah right that can be locked behind bio auth | 00:46:05 |
| atemu12 | Though keystore in general appears to work for Signal and Element | 00:46:17 |
| oak 🏳️🌈♥️ | Did you try to go to that fingerprint debug menu with *#*#2886#*#* | 00:47:14 |
| oak 🏳️🌈♥️ | My OnePlus 8 Pro does not seem to have that with LineageOS Robotnix build | 00:47:42 |
 samueldr | I'm low-key wondering if it's something "simple" like it's using some identifier that changes as a "trust anchor" type thing to prevent system software swap | 00:48:00 |
| atemu12 | It didn't work on my FP4 either | 00:48:04 |
| atemu12 | Just says "Call ended" | 00:48:14 |
| samueldr | * I'm low-key wondering if it's something "simple" like it's (accidentally?) using some identifier that changes as a "trust anchor" type thing to prevent system software swap | 00:48:16 |
| atemu12 | And when I typed it into the dialer it just wiped my input upon entering the last * O.o | 00:48:35 |
| samueldr | In reply to @atemu12:matrix.org
And when I typed it into the dialer it just wiped my input upon entering the last * O.o that would mean it got handled I think | 00:48:58 |
| atemu12 | Huh | 00:49:06 |
| atemu12 | No menu or anything popped up though | 00:49:17 |
| samueldr | "handled" | 00:49:24 |
| atemu12 | Right | 00:49:30 |
| atemu12 | Perhaps LOS is missing whatever blob is responsible for this | 00:49:50 |
| samueldr | the ##...## numbers are to launch some arbitrary stuff, and very build-dependent (and kind of a misfeature imo) | 00:50:04 |
