[PATCH] Planner - change timestep to 2 seconds

[Rick Walsh]

Steve,

On 22 June 2015 at 08:48, Steve Butler <kg7je at comcast.net> wrote:

>  On 06/21/2015 03:23 PM, Rick Walsh wrote:
>
> On 22 Jun 2015 8:16 am, "Steve Butler" <kg7je at comcast.net> wrote:
> >
> > On 06/20/2015 06:25 PM, Rick Walsh wrote:On 21 Jun 2015 11:19 am, "Dirk
> Hohndel" <dirk at hohndel.org> wrote:
> >> >On Sat, Jun 20, 2015 at 11:17:28AM +1000, Rick Walsh wrote:
> >> > > But what about me?  I like SI units and whole decimals.  Don't
> worry, 2 s
> >> > > timesteps fit nicely when using 10 m/s ascent rate (18 s between
> stops).
>
> snip
>
> > Also my DC records every 10s.  One of the DCs i'm looking at does 5s
> intervals.
> >
> > How would this work comparing the pre-dive plan with the post-dive
> profile?
> >
>
> The change has nothing to do with setting or limiting an ascent rate.
> Currently the ascent to the next stop is done in 3 second increments. If
> you ascend at 30ft/min (Subsurface default, which matches most DCs) it
> should take 20s to ascend 10ft. But in 3s increments it is bumped out to
> 21s. No huge issue but it makes the calculated plan have some odd runtimes.
>
>  So that would be 10 calculations (one every 2s) between stops.  If you
> slid the other way and went every 4s then its 5 calculations up to the next
> stop.  Any concerns with snappiness (performance) on slower machines?
>

I have an 8 year old laptop, and calculations appear immediate, so I don't
think it's too intensive.  Changing from 3s to 2s means we do 10
calculations rather than 7, which isn't that dramatic.  Changing to 4s is
ok for a 30ft (9m) /min ascent rate, but for 20ft (6m) /min ascent (also
common), a stop should take 30s but that becomes 32s. 5s is ok for 20ft/min
or 30ft/min but isn't good for 5 m/min or 10 m/min.

You are right, it isn't all that efficient.  I tried setting it to the time
it takes to ascend to the next stop (do the ascent in one jump), which I
thought should work.  And to quote a great of modern philosophy, "60
percent of the time it worked every time".  Occasionally, it would just
break a ceiling.

R
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20150622/4cb4de6d/attachment.html>