proposed changes to libdc

[Dirk Hohndel]

> On Jun 23, 2017, at 11:22 AM, Anton Lundin <glance at acc.umu.se> wrote:
> It's just not VID / PID, but more importantly the Product and
> Manufacturer fields. VID / PID will just tell me this is a ftdi device,
> but if we can match to things like:
> Product: HeinrichsWeikamp OSTC3
> Manufacturer: HW
> 
> We can just do the right thing.

Yes. I didn't realize those were separate fields. How do we get access to them?

> Best would be if we can use that information on desktop os'es to, to map
> whats actually connected to which virtual serial port. It's a bit work,
> but it would be cool.

Agreed

> The question is if we should pack this information into our
> libdivecomputer branch, or we just should keep this in subsurface. If
> we can upstream this, it should go into libdivecomputer, but if Jef
> doesn't want it we should just keep it in subsurface and steer clear
> from the potential merge hell.

I'll admit that I don't spend much time anymore thinking about what he 
may or may not take - I don't think he cares about our improvements 
and the way we want to use the library.

Having this in Subsurface seems incredibly painful and duplicative;
I looked into that option, first, TBH. But it seems much more useful
to have this in the logical place to collect the information...

But you have a valid point about merge hell.

/D