Plotting OC-pO2 on SCR dives and accommodating bailout

[Willem Ferguson]

Hallo Anton, Robert,

Would you be prepared, at all, to review PR #1166? I have a follow-up PR 
that creates a preferences option of selecting OC-pO2 display for SCR 
dives, similar to the way setpoint-display is selectable for CCR dives.

The PR #1170 for accommodating bailouts is a proof-of-concept, not a 
finalised PR for potential merging. I would like some comment on the 
approach taken, on whether it appears viable or not. May I try and 
explain how I see the important issues?

1) No of variables used.

a) The dive sample structure: Jef reported that those dive computers 
that register bailout use two systems: a) discrete bailout events and 
(b) a bool with every dive sample that indicates whether the sample was 
obtained while on closed loop or during OC (=bailout). With respect to 
the dive sample structure, this may mean that one would need to 
accommodate both mechanisms, unless libdivecomputer transforms one of 
these mechanisms into a standardised mechanism. At the least we need a 
mechanism where a rebreather diver can create bailout events by hand in 
order to reflect gas pressures in the log more accurately whn using 
technical dive computers that do *not* register bailout. This means that 
the sample structure may need a bool/int that takes into account 
loop/bailout.

b) The plot_info structure: If one wants to internalise the divemode to 
a sample-level variable, and not a characteristic of the dive computer, 
it means that the plot_info structure would also require a bool/int to 
indicate divemode for that point on the profile. So the variable 
reference would become plot_info -> divemode instead of dc->divemode. Is 
this a realistic view of the problem? Of course, if this was done, it 
would be invasive with widespread effects throughout much of dive.c, 
planner.c and profile.c and also some c++ files. Because of this, #1170 
attempts to address bailout but sticking to the existing dive and 
profile structures and methods.

c) Populating a potential plot_info->divemode variable with the correct 
data. This depends on how libdivecomputer presents the data. But, at a 
minimum one would need a facility to translate bailout events (i.e. in 
the events linked list) into sample->divemode or plot_info->divemode 
values. In this case, the code in #1170 would be usable with very little 
change to interpret the bailout events at a sample level or at a 
plot_info level. Only after allowing for bailout is it possible to have 
realistic partial pressures that could be used to calculate Bühllmann 
tissue compartment pressures and deco parameters for dives that involve 
bailout. In the case of SCR dives, bailout is a standard part of the 
dive procedure. So dive planning for SCR cannot realistically be 
performed without provision for bailout.

Robert, what is your perspective with respect to what all this might 
mean for the dive planner?

Kind regards,

willem



-- 
This message and attachments are subject to a disclaimer.
Please refer to 
http://upnet.up.ac.za/services/it/documentation/docs/004167.pdf for full 
details.