Tank pressure plots for CCRs

[Dirk Hohndel]

On Mon, 2014-01-06 at 20:01 -0200, Rodrigo Severo wrote:

>         
>         One caveat - Subsurface has always been open to cater to
>         specific
>         subsets of the user base, but we have a hard rule that we
>         cater to the
>         "typical recreational diver" first. So we will not make
>         changes that
>         make Subsurface less useful for recreational OC divers just
>         because the
>         few suicidal people who dive rebreathers think that's what
>         they need.
> 
> 
> I see you don't trust rebreathers very much ;)

We already lost one of our developers to rebreathers. Or to quote one of
my best friends in the dive industry "diving on a rebreather the
question is not if you die, but when".

> But I don't see why these changes would make Subsurface any less
> useful for recreational OC divers. As I mentioned the extra step of
> defining a tank pressure segment as OC, CCR O2 or CCR diluent should
> be optional with the default being OC obviously. AFAICT there would be
> no difference here for recreational OC divers.

I just wanted to make sure this is understood upfront. People get
carried away with making Subsurface do exactly what THEY want, not
realizing what most other people need.

>         > Implementing item 1 could also have the positive side effect
>         of fixing
>         > the gaps that consecutive tank pressure plots for the same
>         tank
>         > present. The last pressure of a tank pressure line is always
>         higher
>         > than the first pressure of the next line for the same tank.
>         They
>         > should be exactly the same. These gaps exist now with the
>         current code
>         > and also with my yet not approved nor refused patch
>         "Discontinuity and
>         > stall on tank pressure interpolated lines". I would never
>         suggest that
>         > a change so big as item 1 implies be done just because of
>         these gaps
>         > but I think it would be great to fix this too.
>         
>         
>         Actually, there is no such thing as "it should be the same".
>         Pressure
>         data can fluctuate for many reasons and it's entirely possible
>         that
>         pressure has gone up because the diver crossed a thermocline.
>         Or
>         pressure could have gone down because of leakage.
>         
>         Since pressure data is NOT continuous but assembled from
>         discrete
>         samples there can and will be sudden changes.
> 
> 
> You are completely right when we have discrete samples for the dive.
> 
> 
> I was talking about interpolated tank pressure plots, like when we
> have only the start and end tank pressures for the whole dive. In this
> situation the last pressure of a tank pressure line should be exactly
> the same as the first pressure of the next line for the same tank.
> There is no reason for us to interpolate otherwise. In fact both
> current code and my patch do the right thing in this aspect. It just
> isn't plotted correctly because current code only plots one tank
> pressure point per sample.

That's something we should be able to address in the new profile code.

/D