[GRASS5] Re: r.flow: strange compile error M_PI

GRASS GIS GRASS GIS developer
1

Hi,

I have tried to clean up the [M_]PI mess a bit,
hope I got it right. It is now defined in include/gis.h.

Untouched (not sure how to solved this - precision):
./lib/vector/rtree/gammavol.c
./lib/vector/rtree/rect.c
./lib/vector/rtree/sphvol.c

./lib/cdhc/as241.c
./lib/cdhc/cdh.f
./lib/cdhc/dagstndn.c

Unclear:
./vector/v.label/main.c
which came with 3.1415926535897932384626433832795029L definition (changed to M_PI now).

Markus

2

Markus Neteler wrote:

I have tried to clean up the [M_]PI mess a bit,
hope I got it right. It is now defined in include/gis.h.

Untouched (not sure how to solved this - precision):
./lib/vector/rtree/gammavol.c
./lib/vector/rtree/rect.c
./lib/vector/rtree/sphvol.c

./lib/cdhc/as241.c
./lib/cdhc/cdh.f
./lib/cdhc/dagstndn.c

Unclear:
./vector/v.label/main.c
which came with 3.1415926535897932384626433832795029L definition (changed to M_PI now).

Change all of them (except cdh.f) to use the definition from gis.h.

I'm quite certain that none of the ones which are using a
lower-precision definition really need it to have that exact value.

And v.label definitely doesn't need a "long double" definition.

--
Glynn Clements <glynn@gclements.plus.com>

3

On Thu, Jul 14, 2005 at 01:46:35AM +0100, Glynn Clements wrote:

Markus Neteler wrote:

> I have tried to clean up the [M_]PI mess a bit,
> hope I got it right. It is now defined in include/gis.h.
>
> Untouched (not sure how to solved this - precision):
> ./lib/vector/rtree/gammavol.c
> ./lib/vector/rtree/rect.c
> ./lib/vector/rtree/sphvol.c
>
> ./lib/cdhc/as241.c
> ./lib/cdhc/cdh.f
> ./lib/cdhc/dagstndn.c
>
> Unclear:
> ./vector/v.label/main.c
> which came with 3.1415926535897932384626433832795029L definition (changed to M_PI now).

Change all of them (except cdh.f) to use the definition from gis.h.

I'm quite certain that none of the ones which are using a
lower-precision definition really need it to have that exact value.

And v.label definitely doesn't need a "long double" definition.

OK, all done.

Markus