waypoint above ceiling??

[Sven Lanckmans]

Hi,

I'm running 4.8.6 on a Mac.

When doing a dive plan, I get the yellow triangle when using VPM-B +3, but not when using +2 or +4. This doesn't make sense to me. Isn't the whole idea of an algorithm to propose me a plan which gives me no danger of going through a ceiling? Note that I have plenty of air to stay longer, why would the program then force me to break a ceiling?

Here's a zoom of the area of the triangle, clearly showing that the program is telling me to ascend when there's still deco left at the current ceiling. 

For reference I do not get this error when using conservatism +2 or +4

[Robert C. Helling]

Am Mittwoch, 5. Juni 2019 10:26:10 UTC+2 schrieb Sven Lanckmans:

Hi,

I'm running 4.8.6 on a Mac.

When doing a dive plan, I get the yellow triangle when using VPM-B +3, but not when using +2 or +4. This doesn't make sense to me. Isn't the whole idea of an algorithm to propose me a plan which gives me no danger of going through a ceiling? Note that I have plenty of air to stay longer, why would the program then force me to break a ceiling?

I have seen this behaviour before but have been unable to reproduce it. It often goes away when make some changes back and forth that cause a replot.

I tried to reproduce your situation, but can't: As indicated in the planner notes, you are planning a repetitive dive with a surface interval of one hour. Since I don't have your previous dive, I get slightly different deco times.

Could you please try to reproduce this from an empty log? If that does not work, would you mind sending me the relevant previous dives from your log (in the export menu you can select annonymize dive to remove all the text data), my address is hel...@atdotde.de and I will have a look at it.

Thanks 

Robert

[Sven Lanckmans]

Hi,

I hadn't thought about the surface time. I'm still learning the application and thought that the "save" button was to save my dive plan not save it as a logged dive. :-)

It's indeed related to this, because if I change the date to tomorrow, the warning dissapears.

I've mailed you the two dives directly.

Thanks,

Sven.

[Robert C. Helling]

Unfortunately, VPM-B is somewhat hard to debug as the ceiling depends on so many things and here the violation of the ceiling is rather small so that anything could cause it without being obvious. But this should not be an excuse.

I still have to dive a lot deeper into this but this is not even guaranteed to have a solution: The ceiling plotted in the profile is actually computed using our extension for VPM-B to real dives whereas the planned schedule is computed by our attempt to reproduce the original implementation of VPM-B (which only works for planned dives and thus does not directly apply to the profile). I have blogged about this in the past here: https://thetheoreticaldiver.org/wordpress/index.php/2017/12/22/vpm-b-for-real-dives-or-not/ but in your case the "deco time" seems not to be the immediate problem (as both the planner an the profile agree that it is 1580 seconds).

So yes, this should not happen, but figuring out the detailed reason might take a but longer (and as I said, the violation seems to be very small making probably making it irrelevant for all practical purposes). But please stay tuned.

Best

Robert

[Robert C. Helling]

Am Mittwoch, 5. Juni 2019 12:25:55 UTC+2 schrieb Robert C. Helling:

I still have to dive a lot deeper into this but this is not even guaranteed to have a solution: The ceiling plotted in the profile is actually computed using our extension for VPM-B to real dives whereas the planned schedule is computed by our attempt to reproduce the original implementation of VPM-B (which only works for planned dives and thus does not directly apply to the profile). I have blogged about this in the past here: https://thetheoreticaldiver.org/wordpress/index.php/2017/12/22/vpm-b-for-real-dives-or-not/ but in your case the "deco time" seems not to be the immediate problem (as both the planner an the profile agree that it is 1580 seconds).

I finally got around to investigate this a bit further and I am still not 100% sure but have the strong suspicion that it is exactly one of the problems described in that blog post due to the fact that for real dives (or the profile for that matter) there is no guarantee that what we determine to be the "end of bottom time" is the same as the end of the manual entered waypoints (we rather take it to be the moment of the deepest ceiling).

But still looking into this.

Robert