Quirks of the Subsurface dive planner

[Willem Ferguson]

On 30/01/2017 10:28, Robert Helling wrote:
>
> I will have to look into this with more time. Please keep nudging me 
> until this is resolved.
>
I did quite a bit of checking of the planner using Version 4.6.1. I have 
three points that I think need attention (marked by ****) and some comments.

*Merging dives with dive plans:*

As far as I can see there are no bugs except for:

*** *In the dive notes the concatenation of notes for the real dive and 
for the dive plan are done as (Dive notes) or (Dive plan text)
The text in red, above, is actually printed as part of the dive notes. 
/A clear bug/.

*Editing a dive plan using the planner:*

I save the plan, select it again on the dive list and open it for 
editing within the dive planner.
1) **** Points with zero time duration have been added in the Dive 
Planner Points table. Actually, when re-editing the plan and the 
duration of the descent/ascent segment is less than 1 min, the segment 
is given as having zero duration. In the dive plan details before 
storing the plan, the ascent/descent durations have correct durations, 
usually 1 minute. This creates ambiguity in interpreting the plan being 
edited. /See the attached Screenshot.jpg. I think this needs to be 
changed to be unambiguous. At present, the printed dive plan and its 
representation in the Dive Planner Points table actually do not 
correspond if one compared the runtime and depths for each row in the 
two representations. (This is the 2nd last 26m dive in the dive log I 
previously sent you)./

2) On the profile, a section on back gas has been inserted. If I delete 
the zero-duration-points referred to above, this extra gas segment 
disappears: /This issue is solved as far as I can see./

3) In the planner, the cylinder pressure is not shown in the Information 
Box in the Profile panel. This is a pity because it is extremely useful 
to be able to see the expected cylinder pressure during the different 
phases of the dive: /This issue is solved as far as I can see./

4) Re-opening a dive for editing in the planner, the part of the dive 
where the planner initially calculated the deco profile, (upon 
re-editing) comes up as hard depth/time milestones (i.e. fixed and 
indicated by small white circles as if they were forced (explicitly 
specified in the Dive Planner Points table) all the way to the surface 
during the initial planning.

> There is (at least currently) no way in our file format to record the 
> distinction between manually entered points and computed points. 
> Whenever we go through a file on disk, the whole profile is considered 
> manually entered. That is a problem of the file format.

/This is a limitation we have to live with for now, but we need more 
elegant treatment of the dive plan. I have a few ideas about dealing 
with this within the context of the existing dive structure, but I think 
they will be frowned upon. For instance, the depths of auto-planned 
points could be represented as negative numbers. Basically, outside of 
the planner, negative /*planner*/points on the profile are drawn by 
ignoring the sign. In contrast, within the planner, the negative numbers 
are deleted, forcing the planner to recalculate the plan by using the 
existing "hard-coded" time/depth milestones. But I suspect most 
experienced programmers will consider such an approach as very ad hoc. 
The first prize is the ability of the dive plan to look exactly like it 
was when it was saved with no hard time/depth milestones in the part of 
the dive that was planned by the software.

/*Unwanted gas change events:*/
**** I thought Linus resolved this matter when looking at Miika's case 
of two dive computers with clashing orders of cylinder definitions. At 
present, if I first define a cylinder with EAN50, then afterwards a 
cylinder with EAN32: if the first segment of the dive plan uses EAN32 
(i.e. the second cylinder and gas definition in the Available Gases 
table), there is an unwanted gas change from EAN50 to EAN32 at the very 
start of the dive. Is there any way avoid creating this gas change event?

/*Multi-row edits of dive points table:*/
/

> This is (at least semi-) intentional. You get the behavior you want by 
> right clicking on the waypoint in the profile and inserting a gas 
> change. I always thought this is the expected behavior: In the table 
> you change a single leg of the dive, as this is what a row corresponds 
> to, in the profile you insert a gas change that stays valid until the 
> next explicit gas change. One might think about allowing several rows 
> of the table to be selected simultaneously and then editing one 
> affecting all selected. Let’s see if we can do that.
>
/I do not think this has been implemented yet?/

*Weird gas pressure profiles:*
There are two anomalies that probably have nothing to do with your two 
pull requests, but maybe you understand this better than I do. There are 
unexpected horizontal and/or vertical sections in the cylinder pressures 
in the plan. These anomalies are associated with gas changes. /I cannot 
repeat these anomalies reliably. If I come across this again, I will 
make contact.
/
Finally, the planner operates in a much more predictable way at the 
moment (V4.6.1) and I have confidence in using it for complicated dive 
plans. There has really been some significant progress and I can see 
some time has lately gone into this code. Thank you very much for this time.

Kind regards,
willem



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20170213/929be8df/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Screenshot.jpg
Type: image/jpeg
Size: 33017 bytes
Desc: not available
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20170213/929be8df/attachment-0001.jpg>