Quirks of the Subsurface dive planner

Rick Walsh rickmwalsh at gmail.com
Sun Jan 29 23:32:29 PST 2017


On 30 January 2017 at 17:44, Willem Ferguson <
willemferguson at zoology.up.ac.za> wrote:

> Robert,
>
> Thank you very much for your feedback on this issue and for the associated
> patches you submitted. Apologies for a long reply. During the last week I
> have been using the planner quite intensively. The reason is mainly
> twofold: the graphical interface that allows a visual confirmation that the
> dive plan is set up as intended, including partial gas pressures; 2) the
> facility to check expected gas pressures at critical points along the
> planned profile: this is of enormous value for gas management during the
> dive. Below are a few comments on the behaviour of the planner that I think
> need attention.
>
> The types of dives that we planned involved the use of a travel gas in
> trimix dives of a cave system. therefore a large part of the profile after
> the deepest point of the dive are specified in the dive points table
> (therefore differing from an open-water dive where the planner deals with
> the total ascent from the end of the bottom section of the dive to the
> surface). The strategy was to initiate the dive with nitrox, switch to
> trimix as a main back gas for the deeper part of the dive, then switch back
> to nitrox upon getting back to shallower levels and with final deco using
> oxygen. In short, there are two problems: 1) The order in which the gases
> are entered in the cylinder table has an effect on the rendering of the
> data; 2) when saving a plan and re-opening and editing it in the dive
> planner, the gas changes are forgotten. Below are some images of a
> stereotyped dive plan using the above approach.
>

I'm just a bit jealous about the dives you have planned.

I'll try to have a look at this if I get a chance and I'm not beaten to
it.  Maybe this weekend.

>
> Case 1: Here the order in which the cylinders were defined are: nitrox,
> trimix, oxygen. Image Case1a indicates a plan after saving it and viewing
> it from the dive list. Looks ok, no problems. Image Case1b shows the same
> dive after opening it again for editing in the dive planner. The second gas
> change has been forgotten, requiring extensive editing of the Dive Points
> table.
>
> Case 2: Here the order in which the cylinders were defined are: trimix,
> nitrox, oxygen. Many divers would reason to use this order because the
> trimix is the main back gas, the other cylinders being ancillary. Here
> there are problems arising from the fact that the first cylinder defined
> was trimix, but the dive started with nitrox. The points to be seen in
> image Case2a is that
>
> (i) the cylinder pressure graphs for nitrox are not shown on the planning
> screen. The gas pressure calculations are, however done, evident from the
> final gas pressure for nitrox actually being shown.
>
> (ii) Also, notice on the left margin of the profile (near the 20m depth
> tick mark) that there are two cylinder labels: trimix as well as nitrox. I
> assume what happens is that, since the trimix cylinder was the first one
> defined in the cylinder table, the planner assumes that the dive begins
> with trimix. Then, when in the dive points table, the first segment is in
> nitrox, the planner inserts a gas change event from trimix to nitrox right
> at the start of the dive. It would be wonderful if the profile could
> automatically start off with the first gas defined in the dive points table
> and not with the first cylinder defined in the cylinder definition table. I
> realise that this issue is probably difficult to deal with because of the
> integration between the planner profile display and the "normal" dive log
> display where the cylinder definition table is filled in the order in which
> the dive computer provides the cylinder gas data after the dive. But the
> usability of the planner interface would really be much more user-friendly
> if the profile display is not dependent on the order of the cylinder
> definitions.
>
> Image Case2b shows the profile after being saved, and viewed from the dive
> list. Pretty much as it looked on the dive planning screen. Still no gas
> pressure graphs for nitrox.
>
> Image Case2c shows the same dive plan after re-opening it for editing in
> the planner. In this case, both gas changes involving trimix has been
> forgotten, again requiring extensive re-editing of the dive points table.
>
> There are some smaller issues that relate to the ease of use of the
> planner:
>
> i) If I change the gas in the dive points table, it would be very useful
> if all segments below the one being edited were also changed to the same
> gas. For instance, if I change the gas in the 3rd row of the table to
> EAN40, it would be helpful if all the rows underneath were also changed to
> EAN40. This way I do not have to change by hand each of the subsequent
> segments to EAN40. I realise that, if final deco is done with oxygen, these
> will also be overwritten with EAN40, but that would require only one
> re-edit of a row lower down in the dive points table, changing the gas to
> oxygen: then all the rows underneath would also be changed to oxygen. I
> think the fundamental reason for such a request is that the dive points
> table for these sorts of dive plans includes quite a long list of
> explicitly-defines dive points.
>

I can see this being helpful in some cases, and really annoying to some
users in others.  What about rather than changing the gas of all following
segments, the change applies to the segment being edited, plus any
following segments up until the next defined gas change.  That way you
wouldn't have to re-edit the oxygen segment in your example above.

>
> ii) It would be invaluable if it were possible to insert a row in the dive
> points table. Let's say I planned a dive, and in one of the dive segments I
> want to add another depth level during the descent (or ascent). Currently I
> can do that by double-clicking the profile and creating a new dive point.
> But a) this operation does not always have a predictable effect and b) now
> I am dependent on interaction with the dive profile panel for doing changes
> to the dive points table. Our approach in the past (I think) has always
> been that editing the dive points table is the better way to put together
> the segments for display in the profile panel.
>

Currently, the table is resorted by runtime when it is refreshed/loses
focus.  So if you insert a new row in the table (displays as last row), and
set the runtime to an earlier point in the dive, the row/waypoint will be
where you intend.

It should be feasible to use a bit of Qt magic so that when a new row is
created with the plus button, it is inserted between the row currently with
focus and the following.  If the table didn't already have focus, the row
would be inserted at the end (current behaviour).

>
> iii) When re-opening a dive plan for editing in the planner, the final
> part of the dive plan (determined by the planner) is hard-coded into
> explicit dive points represented in the dive points table. This has some
> very nasty consequences. For instance, if I reposition a gas change along
> the profile using the profile panel, gas changes later on in the plan are
> often screwed. In many cases, the change to oxygen is moved to the very
> last segment of the dive plan: clearly erroneous. This means I have to
> delete all the dive points from the end back towards the point where I exit
> the cave into open water and let the planner recalculate the final deco
> again.
>

> iv) When re-opening a dive plan for re-editing the plan in the planner, a
> zero-time dive point is still inserted as the very first dive point.
>
> Please do not take these remarks as criticism of any sort. One could argue
> that the way that I use the planner here is a corner case. My motivation
> for adding improvements to the planner is twofold: 1) I place very large
> value on the Subsurface planner because it has unique advantages; 2)
> robustness of the planner is an important part of making this software a
> prime choice for other divers.
>

Thank you for your real use, testing, and reporting.  It is very valuable,
as your real-use cave diving and planning differs substantially from mine.
We have some very nice caves not too far from home, but the vast majority
of them, at least the ones I'm qualified to dive, are relatively shallow
and/or short.  Our sinkholes can be deep and require multiple gasses, but
these can be planned similar to open water (with appropriate contingency
planning).  With a few exceptions, most true cave dives here can be done no
deco or short deco on back gas.  Some divers will use a single deco bottle,
but travel gas is rare.

One feature I was thinking could be useful, especially for deep cave dives,
would be to check each manually entered ascent leg to see if it breaks the
deco ceiling, and automatically calculating and inserting required stops
and gas changes in the plan.  Do you think this would be helpful?  The way
I envisage it working would be that the points are not inserted in the
manually entered points table, but do exist in the profile display and
output text table.

Cheers,

Rick
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20170130/32f084c9/attachment-0001.html>


More information about the subsurface mailing list