Haven Dive

[Linus Torvalds]

On Wed, Aug 1, 2012 at 2:07 PM, Andrea Spano <andrea.spano at quantide.com> wrote:
>
> Today I was diving and the xml file is attached.
> The dive started with a Al80 filled with 15/55 for the first 15 minutes
> It continued with a Double 18Lt for the next 35 minutes
> At min 50 the average bottom dept was about 57 mt
> We used EAN50 and O2 for the Deco

Ok. You can actually add gas change events to let subsurface know when
this happened, but sadly, we dont' have a GUI interface for it yet.

Plus our gaschange event interface is horrible, because it's based on
the dive computer interfaces that (at least for Suunto) is about what
O2 *percentage* you switch to, not which tank you switch to. So
because you have two tanks with 15/55, the gas change thing works
really really badly.

I'll have to think about this, but right now you'd have to add something like

  <event time='0:00 min' name='gaschange' value='15' />
  <event time='10:00 min' name='gaschange' value='50' />
  <event time='42:00 min' name='gaschange' value='100' />
  <event time='54:00 min' name='gaschange' value='15' />
  <event time='60:00 min' name='gaschange' value='100' />
  <event time='72:00 min' name='gaschange' value='15' />
  <event time='78:00 min' name='gaschange' value='100' />

to the xml file manually (I didn't check that I got the minute offsets
right, it's roughly what your plan seems to be). And as noted, even if
you add it by hand the above gets things wrong because it doesn't
understand the difference between your tank 1 and tank 4 (both 15/55).

So we have both an interface bug (lack of any sane way to set these
events) *and* a data format bug (bad gaschange event format due to
just copying the event format from libdivecomputer).

I'll get around to this, but because of timing (merge window for the
kernel followed by vacation in Finland), it will be quite some time
before I can fix these kinds of things. So if somebody else wants to
work on it, I can only encourage that person,

                 Linus