| 9 Mar 2025 |
 K900 | Unfortunately, Thread doorbells or buttons don't seem to, like exist | 07:22:52 |
| K900 | * Unfortunately, Thread doorbells or buttons don't seem to, like, exist | 07:22:53 |
 @hexa:lossy.network | * https://www.bleepingcomputer.com/news/security/undocumented-backdoor-found-in-bluetooth-chip-used-by-a-billion-devices/ | 07:59:51 |
 mattleon | In reply to @lynatic:catgirl.cloud
Doesn’t the E dongle also technically support Thread in promiscuous mode while the P dongle only supports Zigbee? I have the E-dongle and it supports both zigbee and thread with the multiprotocol firmware | 12:53:03 |
 dotlambda | In reply to @mattleon:matrix.org
I have the E-dongle and it supports both zigbee and thread with the multiprotocol firmware Don't do multiprotocol. Just get two sticks | 15:01:29 |
| @hexa:lossy.network | the home-assistant people gave up on coexistence for their stick fwiw | 15:04:13 |
| dotlambda | In reply to @k900:0upti.me
Unfortunately, Thread doorbells or buttons don't seem to, like, exist Can't you use any existing button with a Thread "switch" like https://m.intl.taobao.com/detail/detail.html?ut_sk=1.ZePUU%2BurHrQDABCFeGaKc/IL_21380790_1730683154411.Copy.1&id=759224027511 ? | 15:11:16 |
| 10 Mar 2025 |
| @hexa:lossy.network | https://github.com/AlexxIT/go2rtc/releases/tag/v1.9.9 | 08:24:32 |
| @hexa:lossy.network | supports v4l2 as a source now | 08:24:43 |
| @hexa:lossy.network | so like a one-stop-shop for v4l2 to webrtc ;D | 08:24:57 |
| @hexa:lossy.network | * so like a one-stop-shop for v4l2 to webrtc 😄 | 08:24:58 |
|  Gaël joined the room. | 22:26:43 |
| 11 Mar 2025 |
| K900 | OK, update on my zigbee adventure | 06:08:05 |
| K900 | It actually worked surprisingly well | 06:08:17 |
| K900 | I now have a coin battery powered button double sided taped to the outside of my door | 06:08:34 |
 @lynatic:catgirl.cloud | I'm having this problem with my nixpkgs-unstable override right now, is there any fix? | 17:08:05 |
| K900 | You can't really just stack a package from unstable on top of stable | 17:08:35 |
| K900 | And expect it to work in the general case | 17:08:40 |
| @lynatic:catgirl.cloud | huh I thought that was the entire point of nix | 17:09:25 |
| K900 | There are intentional impurities | 17:09:48 |
| K900 | There are also unavoidable impurities | 17:10:05 |
| K900 | Like the kernel | 17:10:08 |
| @lynatic:catgirl.cloud | so this just doesn't work as of now/we have to either use stable or make the full system from nixos unstable?
https://wiki.nixos.org/wiki/Home_Assistant#Tracking_the_latest_release | 17:11:17 |
| @hexa:lossy.network | this was the issue | 17:12:22 |
| @hexa:lossy.network | I LD_PRELOAD'ed and incompatible libjemalloc | 17:12:33 |
| @hexa:lossy.network | commit 8c9755e515014f97a1985d19030ec90cb60785b4
Author: Martin Weinelt <hexa@darmstadt.ccc.de>
Date: Wed Jan 1 23:59:50 2025 +0100
home-assistant: use unstable jemalloc to match native dependencies
Otherwise home-assistant runs into import errors, e.g. on microvad
ImportError: /nix/store/ybjcla5bhj8g1y84998pn4a2drfxybkv-gcc-13.3.0-lib/lib/libstdc++.so.6: version `CXXABI_1.3.15' not found
diff --git a/modules/home-assistant/default.nix b/modules/home-assistant/default.nix
index 538c627..26a958e 100644
--- a/modules/home-assistant/default.nix
+++ b/modules/home-assistant/default.nix
@@ -20,7 +20,7 @@ in
./zigbee2mqtt.nix
];
- systemd.services.home-assistant.environment.LD_PRELOAD = "${lib.getLib pkgs.jemalloc}/lib/libjemalloc.so";
+ systemd.services.home-assistant.environment.LD_PRELOAD = "${lib.getLib unstable.jemalloc}/lib/libjemalloc.so";
services.home-assistant = {
enable = true;
| 17:13:45 |
| *  philipp is avoiding impurities by having home-assistant in a microvm.nix | 17:15:57 |
| @lynatic:catgirl.cloud | I somehow managed to have this problem while having this in a microvm already aaaa | 17:19:49 |
| @lynatic:catgirl.cloud | thanks, I'll try fixing my install to use that module declaration then | 17:20:25 |
| @hexa:lossy.network | are you also using jemalloc? | 17:20:41 |
