Real support for multiple concurrent sensors..

[Dirk Hohndel]

> On Jul 27, 2017, at 10:22 AM, Linus Torvalds <torvalds at linux-foundation.org> wrote:
> 
> On Thu, Jul 27, 2017 at 1:59 AM, Robert Helling <helling at atdotde.de> wrote:
>> 
>> do I understand you correctly that with multiple sensors, we would give up
>> to try to determine which gas the diver is breathing (at a given time)?
> 
> No, we can definitely do it, it's just that the C++ model that does it
> right now is not doing it right.
> 
> It abstracts things away the wrong way.
> 
> The information about the primary active cylinder is still there, it's
> just that it isn't in the pressure samples any more, because we can
> have many pressure samples. It's still there in the gas change
> information, though.
> 
> The problem really is that the TankItem model has saved all the wrong
> things. For example, instead of just saving the displayed dive (like
> everything else does), it copies the current set of cylinders. Which
> is pointless - if it had just saved the current dive and dive
> computer, it would have had the cylinders through that.
> 
> So the TankItem class actually does extra work in order to throw
> useful data away. Which is why I refuse to touch that code.

I'll fix that - the class was largely written by me, anyway :-)

So far I haven't seen any substantial feedback on your code changes,
I'm hoping to get a few more people to actually read through the patch...

/D