"Select Events..." versus "Heading: X°"

[Lutz Vieweg]

On 12/30/2012 10:37 PM, Jan Schubert wrote:
> Sorry, was thinking more about tons of events overlapping the graph in
> subsurface (as we have seen with beta version of ldc before).

I did not see such a symptom - the many "Heading: ..." events are
amassed only in the "Select Events..." dialog, originating from many
different dives with only a few of those events.

> OK, seems screenshot 2 and 3 gets mixed up

Yes. It seems the order that one selects attachments in is
happily ignored by my mail client. :-/

> Anyway, they use some overlay at mouse position!?

Yes, and this is made less useful than it sounds by SDM showing the
popup content for only a second or so, as hard to capture on a screen
shot as it is to read ;-)

> Then this could be
> visualized similar to the partial pressure data in subsurface?

Instead of an under-the-mouse popup I would rather recommend to
designate some text field in the GUI to display whatever details
are available regarding the current position in the graph.

> I doubt that the Suunto protocol
> is giving us this floating deco in the samples

Yes, AFAIK, it doesn't.

> so it would need some magic to show the ceiling

Not only magic, but also science (to compute the decompression model)
and a violation of some countries copyright laws (as the Suunto DC
decompression model is most likely proprietary and would need to be
either reverse-engineered or blatantly copied).

But I guess if subsurface would include any one (or more) reasonable
decompression model implementation, that would be good enough to
visualize some ceiling, it could even be educational to the subsurface
users if they saw that decompression models are what they are - coarse
approximations based on many assumptions and a little empirical data - 
and not precise computations of one's DCS risk.

And btw.: All mainstream DCs that I know of voluntarily avoid to use
the more computational intensive decompression models (like VPM-B/E)
to not drain too much battery power for the model computation.
(Well, the DR5 dive computer doesn't, but that is certainly not
mainstream.)
Subsurface, running usually on some fat CPU, would not need to make
compromises. It could utilize the most accurate models available so far.
That would IMHO be better than trying to emulate whatever manufacturer
specific models are out there.

One could start by porting this open source VPM implementation from Java
http://webspace.webring.com/people/fv/vpmopen/

> fairly easy to implement some progression between deco start and end

Ehm... no. It would be more misleading than beneficial to interpolate
without actually computing a decent decompression model over the full
dive time (including the dives before).

Regards,

Lutz Vieweg