WIP: preliminary Haiku port

[mmuman]

Still a lot to do but enough to get glances running.

[giampaolo]

Mmm... I've never heard of this OS before. How much used is it? Honestly I'm not sure if it's worth it.

[mmuman]

Well, there are so many OSes out there, Haiku is one of them. You'd probably have said this of Linux 15y ago btw.
We're far from the user base of Linux and I don't have download figures, but I don't know how many people use Solaris (or IllumOS) nowadays either.

[mmuman]

Just rebased, no new feature yet.

[giampaolo]

Looking back at this, I'm sorry but I'm gonna reject it. There's been no demand for Haiku, and overall it looks like a very niche OS. To give an idea, these are the current psutil's monthly download stats:

| System name                    |       Downloads |
|:------------------------------ | ---------------:|
| Linux                          |      11,084,549 |
| Windows                        |         790,397 |
| Darwin                         |         232,315 |
| FreeBSD                        |           3,499 |
| OS/390                         |              88 |
| AIX                            |              84 |
| SunOS                          |              74 |
| OpenBSD                        |              39 |

Considering OpenBSD as the most underused platform with 39 monthly downloads, I expect Haiku numbers would be close to zero.

More in general: since all major and most known minor platforms are already supported, the bar for entry for adding a new exotic/niche platform at this point is very high. The technical debt [1] simply outweighs the benefits. A separate fork (link to Haiku fork) + a specific PYPI package would serve this task more effectively.

[1] new code in main core and unit tests, doc adjustments, occasional bug reports which will likely be stagnating, etc.