Current subsurface issues..

[Linus Torvalds]

So there's two new issues I found so far (not counting the renumbering bug).

Our handling of "air time remaining" events is pretty nasty. This is
not new, but I noticed just because I added support for it to the EON
steel backend now that the v1.1.15 release has airtime support.

We treat it as an event, which means that we get an event for every
sample (since generally you'd get airtime remaining for every sample).
Which makes the save files unnecessarily big, but also makes the
profile look horrid, because you just have this endless stream of
event triangles for the airtime handling.

I guess I'll take a look after the dive trip. Most of our airtime
handling isn't about GUI issues.

The other issue is with the new dive site handling. It interacts very
badly indeed with the GPS download. It's basically unusable.

So what happens is:

 - you download the GPS information

 - all the dives are now "Autogenerated dive", but they have the right
location, so the map looks good.

 - you want to *name* them, and maybe use the existign names, so you
use the drop-down menu to do so.

 - that now overwrites the GPS information. And the *good* case is
that it overwrites it with valid GPS information that is at least
close. But you have no way to tell, and it could be completely bogus
GPS information, or it could be just a name with no GPS information at
all.

Very annoying. Not really useful.

I'm willing to take a look at the airtime handling, since I have a
test-case and it's not mainly about UI. But the divesite thing I'm not
touching.

                      Linus