Consider only making fields optional if they were added after Linux 2.6

[cjbassi]

The docs say that this library intends to support Linux versions greater than 2.6. Would it make sense to only make fields optional if they were added after 2.6? Anything before 2.6 is pretty old, and from my experience with rust-psutil, there's a lack of documentation and a lot of missing features from kernels older than 2.6.

edit: This would be a breaking change, and would require some changes to struct parsing too, but would also make things a lot simpler.

[eminence]

In general, yes the idea is that fields added in newer kernels become a Option in the procfs structs. The stat struct is good example.

What I think you're asking about is some really old fields (like exit_signal or processor) that were added in Linux 2.1 or 2.2 (which is way before 2.6). Is that right?

I suppose I don't have any major objection to make a breaking change like this. I'm not sure if I'd want to do it right now. But if we had to make another breaking change then turning some of these fields into non-Option would make sense.

[cjbassi]

Yeah, any fields that existed before 2.6 should not be options.