[PATCH] VPMB gradient of imperm: don't confuse atm and bar

[Rick Walsh]

On 17 September 2015 at 04:21, Dirk Hohndel <dirk at hohndel.org> wrote:

> On Wed, Sep 16, 2015 at 07:57:24PM +1000, Rick Walsh wrote:
> > Other VPM-B implementations (V-Planner, MultiDeco, Fortran code) take the
> > default gradient of onset of impermeability as 8.2 atm.  We use bar as
> the
> > pressure unit, and 8.2 atm = 8.30865 bar.
> >
> > This has negligible effect on all of the calculated plans I've tested,
> resulting
> > in a maximum decrease of 1 minute in total dive time for ~150 m deep
> dives, but
> > we should get unit conversions correct anyway.
>
> I'm not arguing against that. But with this patch TestPlan fails.
> So either submit a patch that also adapts the runtime we expect in the
> test (and argue why that is the right thing to do), or...
>
> NAK as it is
>

Oops - I should have checked that - it's a bit embarrassing.

While sitting in the naughty corner, I realized we were calculating the CVA
next gradient based on the tissue loading at the end of the previous
iteration, rather than the end of the user-entered portion of the dive.
The attached patch fixes that, and it also makes all the tests pass when
applied on top of the previous patch.

Rick
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20150917/ddb69a21/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-VPM-B-restore-deco-state-before-calculating-next-gra.patch
Type: text/x-patch
Size: 1144 bytes
Desc: not available
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20150917/ddb69a21/attachment.bin>