| 29 Sep 2025 |
 outfoxxed | were supposed to be the same thing | 05:39:37 |
| outfoxxed | I guess I'm building qt again today | 05:40:02 |
 K900 | Also I think I now have the context to explain what I was saying before about nativeBuildInputs | 05:40:28 |
| K900 | My idea was to instead of doing what you did and propagating host tools everywhere | 05:40:41 |
| K900 | To build a special "cross Qt" for (build -> target) | 05:41:09 |
| K900 | So we have tools for build and libraries for target in the same package | 05:41:19 |
| K900 | And then we just use those everywhere | 05:41:24 |
| K900 | Which works without cmake patching but does lead to funni in Canadian use cases but no one does that anyway | 05:42:02 |
| outfoxxed | I'm not sure what the advantage of that would be but I think that is an officially supported mode of operation | 05:42:10 |
| outfoxxed | honestly I'm not sure why I had to patch the cmake for this, it should work with a host build in one folder | 05:42:38 |
| K900 | That is the officially supported mode of operation | 05:42:39 |
| K900 | That's mostly the advantage yes | 05:42:39 |
| K900 | But seeing how small the cmake change is I think I like yours more | 05:42:42 |
