[PATCH] Import surface event from Suunto DB

[Linus Torvalds]

On Tue, May 19, 2015 at 11:36 AM, Robert C. Helling <helling at atdotde.de> wrote:
>
> On 19 May 2015, at 18:22, Miika Turkia <miika.turkia at gmail.com> wrote:
>
> Miika,
>
>> This is actually relevant information when someone surfaces
>> accidentally, thus we should not ignore it.
>
> maybe this event should also trigger a zero depth sample. What do you think?

I think Suunto (and probably others) considers "surface" to be 3ft,
not zero. I've seen "lots of surface events" when I do a swim back to
shore, and I think it would actually look bad if it made the profile
jump back-and-forth between 4ft and 0ft for that use case. It may not
be horribly common, but I don't think it's necessarily all that
unusual either.

It might be interesting to see what the samples around it are, and
maybe also take sample rate into account. With a 30s sample rate, I
certainly agree that losing a surface event in between just because
you went up and immediately back down again is a misfeature. But with
a 10s sample rate and the samples around that time being 4ft, I think
it would be wrong to say "0ft" in between, because that really isn't
what the event even means.

The reason we mask out that event is really that it can get very
annoying in that "shallow swim out or back" scenario for shore diving.
Maybe other people really tend to swim at the surface, and I'm just
odd for liking to do the "shallow dive back" thing. But clearly it was
wrong to ignore it *entirely*, and we should do something more subtle.

                   Linus