I like this. Bitbake is a steep learning curve. Nowhere near as simple as buildroot. But I maintain that if you can get over the first few slopes, the payoff is worth it.
However, I don't like new files as patches. I really prefer to have my device tree be a dts file that I bring in instead of bundled into a patch. Maybe I'm not following the guidelines, but I think it's nicer to be able to search for dts things in .dts files and I get nice syntax highlighting and whatnot.
I also like their stance that you only need one layer. I've had people push for a layer per machine. Not needed as shown here and most other places.
> However, I don't like new files as patches. I really prefer to have my device tree be a dts file that I bring in instead of bundled into a patch. Maybe I'm not following the guidelines, but I think it's nicer to be able to search for dts things in .dts files and I get nice syntax highlighting and whatnot.
This is what I do on custom boards. It's better to "look" at files and link to others when they're files and not patches.
Would there be any advantage in using Yocto if you only ever have one target (x86 in my case)? Been happily using Buildroot but wondering just how greener the grass is on the other side.
The advantages aren't strictly on how many architectures you have. There's more facility to put things in the layer as steps instead of hacky surrounding scripts, and I've never had it mess up needing to rebuild something.
However, I don't like new files as patches. I really prefer to have my device tree be a dts file that I bring in instead of bundled into a patch. Maybe I'm not following the guidelines, but I think it's nicer to be able to search for dts things in .dts files and I get nice syntax highlighting and whatnot.
I also like their stance that you only need one layer. I've had people push for a layer per machine. Not needed as shown here and most other places.
This is what I do on custom boards. It's better to "look" at files and link to others when they're files and not patches.