Subsurface deco calculations

[Linus Torvalds]

On Wed, Feb 8, 2017 at 9:50 AM, Linus Torvalds
<torvalds at linux-foundation.org> wrote:
>
> That code was written to basically create the dives with the primary
> dive computer first, and then merge the random secondary ones we would
> carry. So the code expected that the base dive is the primary and more
> capable dive computer.

The comment we have there even talks about it, and makes sense in that
situation:

 * Merging cylinder information is non-trivial, because the two dive computers
 * may have different ideas of what the different cylinder indexing is.
 *
 * Logic: take all the cylinder information from the preferred dive ('a'), and
 * then try to match each of the cylinders in the other dive by the gasmix that
 * is the best match and hasn't been used yet.

and the issue is that the "preferred dive" that we take the cylinders
from is the dive with the worse data, so what we do doesn't end up
working.

If you download the ostc first, and then download the vyper second, it
should all "just work".

I'll fix it, never fear. I'll just have to think more about it.

                Linus