I agree: stacks and stacks of try / except statements can lead to indent hell.
One way to cut that down is to use functions. Each function would try / except a set of related operating systems, and if none of them work, raise NotImplementedError. The main function would then just have to catch NotImplementedError. It will cut down the bulk. -tk On Mon, Apr 15, 2019 at 2:56 PM Jon Marnock <jon.klo...@gmail.com> wrote: > That is a good point, and it doesn’t look from that code as though windows > is currently supported. I might take a look at writing something that does > switch on OS with differing strategies. I kind of don’t like that approach > in general (because it feels like the ‘ol “version check” in windows which > everyone got so wrong for so png that MS actively fights you if you try and > get that info on recent versions). Mostly though my issue is that OS > versions aren’t necessarily correlated directly with methods - eg, until > about 5.7 iirc, OpenBSD had a procfs you could use, but it was dropped. > Arguably we should be testing fpr the existence of methods, not versions > (the current exception driven design is actually closer to this, though > just trying t and dealing with the fallout when it fails is perhaps less > elegant!)
