Log for Tusa Zen DC

[Dirk Hohndel]

On Sat, Sep 13, 2014 at 07:23:45AM -0700, Dirk Hohndel wrote:
> On Fri, Sep 12, 2014 at 08:55:12AM +0200, Willem Ferguson wrote:
> > 
> > Thirdly the depth scaling that Subsurface does for the dives is crazy for
> > all except the first and last dives.
> 
> Can you tell me what exactly you consider crazy? Just so we know what we
> are talking about? :-)

Ah, sorry, missed the attached file at first.

No, the scaling is totally sensible. We trust libdivecomputer. So we plot
what libdivecomputer gives us. In this case look at the XML for dive 2

<trip date='2014-08-16' time='11:40:00'>
<dive number='2' date='2014-08-16' time='11:40:00' duration='8:52 min'>
  <cylinder o2='87.0%' />
  <cylinder />
  <cylinder o2='58.0%' />
  <divecomputer model='Tusa Zen (IQ-900)' deviceid='1b4bbd37' diveid='4807f68c'>
  <depth max='629.222 m' mean='4.13 m' />
  <temperature water='-17.778 C' />
  <sample time='0:02 min' depth='4.858 m' temp='-17.778 C' pressure='17.582 bar' />
  <sample time='0:04 min' depth='4.953 m' />
  <sample time='0:06 min' depth='4.858 m' />

So he was diving with 87% O₂ in your first cylinder, air in the second,
58% O₂ in the third. That sounds odd - can you check if that's right?
He went down to 629.222m but had a mean depth of 4.13m. Long deco stop at 3m?
And the water was -17.778°C - I hope he brought a saw.

I'd agree with Jef, there are a couple of parser issues here :-)

Subsurface scales for the maximum depth in the header. The samples never
reach that deep. Maybe we could scan for the maximum depth in the samples
and scale for that instead. Since that's what we plot...?

This is clearly a situation of garbage in, garbage out - but this
particular one we could have avoided.

Thoughts? Arguments why that's a bad idea before I make the change?

/D