On Sat, Oct 28, 2006 at 06:18:09PM +0100, Paul Makepeace wrote:
I currently have t/53 failing due to WGS84 coords not
being
calculated. I traced this back to Geo::HelmertTransform failing due to
it not being installed.
I was going to patch this but wanted to check first. My initial
thought was if the force_wgs84 param is set to 0 (no) during perl
Build.PL then the Geo::HelmertTransform requirement should be added to
the build requires section. Something ditto for
Geography::NationalGrid::(IE|GB). What say ye?
Hrm, given this is guide-specific rather than installation specific I
would rather fix the test so that it doesn't fail, and then make sure
that a nice run-time error happens when Geo::HelmertTransform is needed
but not installed.
Also, I'd like to see $@ tested after the
transformation; I have that
as a die in my local source right now. It's at the least irritating
having silent failures requiring source spelunking.
Nick, any, chance you could look at this bit of it?
As an aside it made me wonder whether force_wgs84 = 1
is a more
appropriate default given that it doesn't seem to matter much, and
it's an optimisation away from performing the transform. I really
havent looked at the code all that much though so this might not be a
sensible comment.
Hrm. Not so sure about this really. It does make a difference of about
20 or 30 metres and in towns this is very noticable.
Dominic.
--
Dominic Hargreaves |
http://www.larted.org.uk/~dom/
PGP key 5178E2A5 from the.earth.li (keyserver,web,email)