Re: [Clanlib-devel] Patch for warnings

[Seth A. R. <se...@rt...>]

Hi Ben,

Sorry for the slow reply and thanks for the patch.

Anyway, I applied it, but figured I'd better not commit right away =
because I
noticed I was still getting compiler warnings for many things that were
changed, for instance:

The old float offx1 =3D cos(M_PI * 2.0 * (e / numsegments))*radius;

And the new

float offx1(cos(M_PI * 2.0 * (e / numsegments))*radius);

give the same warning under VS8:

.\Display\Collision\collision_outline.cpp(333) : warning C4244:
'initializing' : conversion from 'double' to 'float', possible loss of =
data

I can fix it by doing this:

float offx1 =3D (float)cos(M_PI * 2.0 * (e / numsegments))*radius;

so I wanted to ask, what compiler(s) are you using on your side? GCC =
4.X? Is
it possible for you to also test under MS VS8?  Probably not but hey,
thought I'd ask :)
=20
-Seth
=20

-----Original Message-----
From: Ben Chambers [mailto:bdc...@ya...]=20
Sent: Thursday, December 21, 2006 5:31 AM
To: ClanLib Devel
Subject: [Clanlib-devel] Patch for warnings

Since warnings bug me, I've patched a number of files to get rid of =
them.

Mostly, I've included explicit type castes where the casting was done=20
implicitly.

Also, in collision_outline.cpp, on lines 300, 339, 366, I changed=20
truncate-by-caste (requires two castes) to floor().

Note that, in collision_outline.cpp, on lines 456, 457, 519, 520, 588,=20
589: Changing what's included in the parenthesis could affect rounding.

...Chambers



__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around=20
http://mail.yahoo.com=20