Problems - apd ccr dives import

[Miika Turkia]

On Mon, Apr 30, 2018 at 6:18 PM, Willem Ferguson <

Hallo Miika,
I was wondering whether you would be prepared to look at these three CCR
1) Three separate text files (.apd) that need import using the Import ->
Import Log Files dialog. They are labelled dive1 - dive3.
2) A xml dive log file containing these three dives.
You will notice that the setpoint data for dive 1 is recorded correctly,
but the setpoint data for the other two dives are erroneous: the po2
attributes in the xml is frequently updated to reflect some aspect of the
sensor pO2 and not the setpoint. The reason for this is most probably that
there has been a a new software release by apd. Look at the headings for
the columns in the .apd files and you will see that there are differences
in the rightmost three columns of the text files: Dive 1 differs from the
other two that come from the newer version of the apd log viewer. My
feeling is that, if this is indeed due to the upgrade, it would be
fantastic to cover the new software and forget about the older version.

The difference you see is probably due to different Subsurface versions
when you imported the dives. The format difference should only affect deco
ceiling and GF values as there was one field added before those two.

What is currently imported as po2 field is the one labeled "PPO2 - Setpoint
(Bar)" in the log. If that field is not set, then we use "PPO2 - C1
Measured (Bar)". I think you would prefer this the other way around?

(I am totally ignorant when it comes to CCR diving, so when you say
setpoint, I try to look for a field named that)

miika

[Anton Lundin]

Post by Miika Turkia
On Mon, Apr 30, 2018 at 6:18 PM, Willem Ferguson <

Hallo Miika,
I was wondering whether you would be prepared to look at these three CCR
1) Three separate text files (.apd) that need import using the Import ->
Import Log Files dialog. They are labelled dive1 - dive3.
2) A xml dive log file containing these three dives.
You will notice that the setpoint data for dive 1 is recorded correctly,
but the setpoint data for the other two dives are erroneous: the po2
attributes in the xml is frequently updated to reflect some aspect of the
sensor pO2 and not the setpoint. The reason for this is most probably that
there has been a a new software release by apd. Look at the headings for
the columns in the .apd files and you will see that there are differences
in the rightmost three columns of the text files: Dive 1 differs from the
other two that come from the newer version of the apd log viewer. My
feeling is that, if this is indeed due to the upgrade, it would be
fantastic to cover the new software and forget about the older version.

The difference you see is probably due to different Subsurface versions
when you imported the dives. The format difference should only affect deco
ceiling and GF values as there was one field added before those two.
What is currently imported as po2 field is the one labeled "PPO2 - Setpoint
(Bar)" in the log. If that field is not set, then we use "PPO2 - C1
Measured (Bar)". I think you would prefer this the other way around?
(I am totally ignorant when it comes to CCR diving, so when you say
setpoint, I try to look for a field named that)

Somewhere back in the dark ages we mixed up the measured po2 and the
expected po2 (setpoint) into the same variable, and used only the
measured po2 if we had it, and if we didn't have it, we filled that
value with the setpoint.

As usual, the reality is more complex than that simple model, and now we
have setpoint, messured po2, messured po2 by cell, calculated po2 and so
on.

In this world, having a attribute in the xml named "po2" is just really
confusing, and we should probably rename it to setpoint, because that's
what we're using it for nowadays. With this, the problem of migrating
old data comes to.


//Anton

[Willem Ferguson]

Post by Miika Turkia
The difference you see is probably due to different Subsurface
versions when you imported the dives. The format difference should
only affect deco ceiling and GF values as there was one field added
before those two.
What is currently imported as po2 field is the one labeled "PPO2 -
Setpoint (Bar)" in the log. If that field is not set, then we use
"PPO2 - C1 Measured (Bar)". I think you would prefer this the other
way around?
(I am totally ignorant when it comes to CCR diving, so when you say
setpoint, I try to look for a field named that)
miika

It is my stupid fault, as usual. Yes, the dives were imported using
different versions of Subsurface. Importing them all with the latest
build resolves all problems. Both the sample structure and the plot_data
structure have setpoint members. I think the present parsing of apd
files is reasonably clear. But the handling of the po2 variable could do
with some simplification.

Kind regards,

willem