[PATCH] Set in_deco flags for subsurface calculated deco

Dirk Hohndel dirk at hohndel.org
Thu Oct 17 15:06:59 UTC 2013


On Thu, 2013-10-17 at 23:31 +0200, Anton Lundin wrote:
> On 14 October, 2013 - Dirk Hohndel wrote:
> 
> > 
> > On Mon, 2013-10-14 at 22:44 +0200, Anton Lundin wrote:
> > > Now when we look at in_deco for ploting of deco-stops, we need to set
> > > that flag for non-dc calculated deco too.
> > 
> > Careful. We want to be able to tell apart whether your dive computer
> > thought you were in deco or whether Subsurface thinks that you are in
> > deco. These are visualized differently. And IIRC in_deco should only be
> > set if the dive computer told us that was its state at that time...
> > 
> 
> I managed to reply directly to Dirk so we had some discussion off-list.
> After that discussion i thought up some alternatives:
> 
> 1. Add a second set of parameters into the plot_data for the
> subsurface-calculated ndl/tts/deco and plot that besides the
> dive-computer reported data.

I am fine with that

> 2. Only calculate ndl/tts/deco for dc's that is missing those values and
> add a flag for the plot that the ndl/tts/deco is calculated by
> subsurface.

But then you can't do the comparison - if you calculate only one set of
data then you don't need a second set of parameters, you just need a
flag if it's from the DC or from Subsurface

> 1 is interesting to be able to compare our deco code on the same data as
> some divecomputers calculations, but how are we going to fit it all into
> the plot?

This is only in the overlay, anyway. So you make the overlay show 
NDL: 34min(DC) / 22min(calc)
If they are identical do
Ceiling: 12m (DC & calc)

> 2 is more how i thought when i wrote the code, because i have a Suunto
> Vyper that can't do proper deco calculations, and it is interesting to
> see the data.
> 
> 
> I didn't find any assumptions in the code that would break if the
> subsurface-based calculations set the in_deco flag, but i also managed
> to miss the ruler functionality up until some days ago =)

I'm pretty sure that the code that draws the ceiling and colors the
background accordingly would break (but haven't tried so it's possible
it's smarter than I remember it being).

/D





More information about the subsurface mailing list