[PATCH] Visualisation of individual oxygen sensor data for CCR dives

Anton Lundin glance at acc.umu.se
Sat Jan 10 03:00:33 PST 2015


On 09 January, 2015 - Willem Ferguson wrote:

> On 09/01/2015 16:32, Anton Lundin wrote:
> >
> >I've bin thinking if this is a good idea to plot in our already massive
> >plot.
> >
> >I'm more leaning towards just plotting a computed po2 value here, and
> >the setpoint of course, and then rather add a some sort of separate tool
> >to analyze the o2 cells, and how they correlate to each other, outside
> >the general dive plot.
> >
> >I think thats going to be a way more useful tool.
> >
> >
> >//Anton
> >
> >
> Anton,
> 
> I hope I can offer a systematic argument here. So, for this reason, I have
> numbered the points below.
> 
> 1. In order to perform evaluation of oxygen sensors it is, as a start,
> necessary to visualise the individual sensor readings. Other CCR software
> also do this. Attached, as Image 1, is a screen dump from the APD software.
> This is part of the main graphical display of that software. It shows the
> measurements for each of the three sensors, the setpoint and the depth
> profile. Even though their graphical environment is somewhat rudimentary (at
> least compared to Subsurface) this is the sort of information that APD
> divers in this part of the world look at as an aid to evaluate the condition
> of their equipment's oxygen management system and also the diver's
> interaction with that system. Unfortunately my Poseidon software is not
> running at the moment (license issues), so I use APD software as an example.
> This type of visualisation is not exceptional. It is close to the norm.
> 

Yea, its what everybody else is doing, but what does those values
actually tell you? Can you see with which margin the red sensor is
undershooting and with which margin the orange sensor is overshooting?
I'd say no.

I'd say its way better to plot the sensor lines only if they differ more
than X %/ppo2/mV from the expected readout. That will provide only the
relevant information and not the noise from just normal values visually
clogging the plot.

> 4. I can sympathise with with your point of view of creating a separate
> tool. Now I wish to ask a question. Would you be prepared to generate the
> separate tool that you intend? Until now, I have been working mostly on my
> own, but with significant help from Miika and Robert. Do we need to wait
> until someone, some time in the future, will be prepared to push the
> Subsurface CCR initiative further? Or are you prepared to commit to a plan
> of action to create a new Subsurface tool?
> 

Did i step on any virtual toes that i didn't notice?

> 5. Would you like this to be an independent non-Subsurface tool? Would
> someone need to separately rewrite the whole infrastructure for and
> visualisation of dive logs that Sursurface already has?
> 

Now you're just misinterpreting things for the sake of finding things to
argue about. Of course not, it will not be a standalone tool outside
Subsurface, it would be a tool inside Subsurface.

> 5. My suggestion is this: Please, let's for the meantime work with the
> present code which is not perfect. There are one or two rough edges that
> need to be smoothed over. But let us use it until someone comes with a
> better idea of how to deal with CCR dive logging and interpretation in the
> open source domain. I perceive this to be an opportunity for Subsurface, not
> a threat.
> 

I've never suggested to remove all the code and start over. I'm
suggesting better goals to strive towards.

We have discussed o2 sensor plotting on this mailinglist for 2-3 years,
and everytime before we have come to the conclusion that plotting
individual sensors is just noice. Yes, its what everybody else does but
everybody else are still butt ugly and stupid =)


Yes, if this is your puppy and your holy grail build it so then, but
make sain defaults like not plotting all the sensors individually, but
rather have the default as plotting the consensus/avenge/computed/whatnot
value and if the user like to, can turn on plotting the individual
values.


I hack on the parts of Subsurface which i think are fun. With this
attitude CCR-parts are not fun.


//Anton


-- 
Anton Lundin	+46702-161604


More information about the subsurface mailing list