| 25 Feb 2025 |
 @hexa:lossy.network | https://github.com/christiaangoossens/hass-oidc-auth | 01:23:43 |
 macaroniskengz | Not really, seems overkill to be honest as I don't have any OID infrastructure yet | 01:23:57 |
| macaroniskengz | Maybe I just have to accept that Home Assistant will come with some state and I cannot do anything about it | 01:25:51 |
| macaroniskengz | A more HASS related question: can I obtain a "short-lived" access token by performing the usual login flow as part of the frontend to use the WebSockets API? I mean: instead of using a Long Lived Access Token to use the API, I perform the flow and after passing my credentials to Home Assistant I receive a access token that expires in some time. | 01:29:43 |
| @hexa:lossy.network | home-assistant stores the session data in local storage | 01:31:14 |
| @hexa:lossy.network | * home-assistant stores the session data in local storage in the browser | 01:32:00 |
| @hexa:lossy.network | not sure how it looks on the wire | 01:32:06 |
| @hexa:lossy.network | zigbee2mqtt 2.1.1 on master https://github.com/NixOS/nixpkgs/pull/371053 | 04:09:29 |
| @hexa:lossy.network | it is opt in | 04:09:32 |
| @hexa:lossy.network | use the 25.05 cycle to migrate to zigbee2mqtt_2 | 04:09:51 |
| @hexa:lossy.network | will remove zigbee2mqtt_1 after the 25.05 branch-off, so it will not be in 25.11 | 04:10:06 |
| 27 Feb 2025 |
|  mike changed their profile picture. | 13:53:50 |
 Sandro 🐧 | is that better than the ldap login script? | 18:54:53 |
| @hexa:lossy.network | did you intend to quote the message about auth-oidc? | 18:55:36 |
| @hexa:lossy.network | I have not yet tried it, and I'm not sure I will | 18:55:49 |
| @hexa:lossy.network | I'll prefer an OIDC flow any day over LDAP | 18:55:58 |
| 28 Feb 2025 |
| @hexa:lossy.network | ok, so the frigate update is kinda blocked on onnxruntime support the openvino execution provider | 05:21:39 |
| @hexa:lossy.network | what I have is here: https://github.com/NixOS/nixpkgs/pull/380543 | 05:22:32 |
| @hexa:lossy.network |
/nix/store/0d0l820v8vnw7kf959xr6bvpvyg21r2b-onnxruntime-1.20.2/lib/libonnxruntime_providers_openvino.so: undefined symbol: _ZN9OrtDevice3NPUE
| 05:29:54 |
| @hexa:lossy.network | * /nix/store/0d0l820v8vnw7kf959xr6bvpvyg21r2b-onnxruntime-1.20.2/lib/libonnxruntime_providers_openvino.so: undefined symbol: _ZN9OrtDevice3NPUE
| 05:30:01 |
| @hexa:lossy.network | dunno, maybe it is just something stupid, like openvino built w/o npu support | 05:30:20 |
 K900 | npue | 05:30:26 |
| @hexa:lossy.network | 
Download image.png | 05:30:45 |
| @hexa:lossy.network | 🫠 | 05:30:48 |
| @hexa:lossy.network | openvino> /build/source/src/plugins/intel_npu/tools/protopipe/src/scenario/accuracy_metrics.hpp:9:10: fatal error: opencv2/core/core.hpp: No such file or directory
openvino> 9 | #include <opencv2/core/core.hpp>
openvino> | ^~~~~~~~~~~~~~~~~~~~~~~
openvino> compilation terminated.
| 05:46:07 |
| @hexa:lossy.network | ok, that looks like a fairly simple error | 05:46:16 |
| @hexa:lossy.network | but I'm already passing opencv.cxxdev in buildInputs | 05:46:25 |
| @hexa:lossy.network | - # src/graph/src/plugins/intel_gpu/src/graph/include/reorder_inst.h:24:8: error: type 'struct typed_program_node' violates the C++ One Definition Rule [-Werror=odr]
- env.NIX_CFLAGS_COMPILE = "-Wno-odr";
+ env = {
+ CXXFLAGS = "-I${lib.getDev opencv}/include/opencv4";
+ # src/graph/src/plugins/intel_gpu/src/graph/include/reorder_inst.h:24:8: error: type 'struct typed_program_node' violates the C++ One Definition Rule [-Werror=odr]
+ NIX_CFLAGS_COMPILE = "-Wno-odr";
+ };
| 07:09:12 |
| @hexa:lossy.network | did this | 07:09:13 |
| @hexa:lossy.network | and now I'm here | 07:09:22 |
