 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Wasn't it Tyler Eaves who wrote:
>I'm looking for some way to specify a box at an angle. The way I see it, a
>box could be specified by giving 2 points, which would be the center of
>two oppisite faces, and would define one dimesion, call it length, and
>then giving the length of the width and height. This would be roughly
>comprable to the way a cylinder is specified. I can't seem to think of any
>obvious way to do this (3d math is not a strength of mine...)
The first thing that springs to mind is to write a macro like this:
#include "transforms.inc"
#macro aligned_box(P1, P2, H, W)
#local L = abs(vlength(P2-P1));
box {<0,-H/2,-W/2><L,H/2,W/2>
transform {Reorient_Trans(x,P2-P1)}
translate P1
}
#end
--
Mike Williams
Gentleman of Leisure
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Wed, 07 Sep 2005 14:58:11 -0400, Mike Williams
<nos### [at] econym demoncouk> wrote:
> Wasn't it Tyler Eaves who wrote:
>> I'm looking for some way to specify a box at an angle. The way I see
>> it, a
>> box could be specified by giving 2 points, which would be the center of
>> two oppisite faces, and would define one dimesion, call it length, and
>> then giving the length of the width and height. This would be roughly
>> comprable to the way a cylinder is specified. I can't seem to think of
>> any
>> obvious way to do this (3d math is not a strength of mine...)
>
> The first thing that springs to mind is to write a macro like this:
>
> #include "transforms.inc"
> #macro aligned_box(P1, P2, H, W)
> #local L = abs(vlength(P2-P1));
> box {<0,-H/2,-W/2><L,H/2,W/2>
> transform {Reorient_Trans(x,P2-P1)}
> translate P1
> }
> #end
>
That comes VERY close. Actually, it works as I requested, although missing
an implied criteria. I didn't specify it, but I need the box to maintain a
vertical orientation, whereas your macro induces a rotation around the
P1-P2 axis. I'm guessing this is an artifact of the Reorient_Trans
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Tyler Eaves wrote:
> On Wed, 07 Sep 2005 14:58:11 -0400, Mike Williams
> <nos### [at] econymdemoncouk> wrote:
>> #include "transforms.inc"
>> #macro aligned_box(P1, P2, H, W)
>> #local L = abs(vlength(P2-P1));
>> box {<0,-H/2,-W/2><L,H/2,W/2>
>> transform {Reorient_Trans(x,P2-P1)}
>> translate P1
>> }
>> #end
>>
>
> That comes VERY close. Actually, it works as I requested, although
> missing an implied criteria. I didn't specify it, but I need the box to
> maintain a vertical orientation, whereas your macro induces a rotation
> around the P1-P2 axis. I'm guessing this is an artifact of the
> Reorient_Trans
I'm assuming you're generating P1 & P2 somewhere else? I believe you
either have to ensure both y points are the same before passing them to
the macro or else remove them in the macro. Does this work?
#include "transforms.inc"
#macro aligned_box(P1, P2, H, W)
#local L = abs(vlength(P2-P1));
#local FP1 = <P1.x, 0, P1.z>;
#local FP2 = <P2.x, 0, P2.z>;
box {<0,-H/2,-W/2><L,H/2,W/2>
transform {Reorient_Trans(x,FP2-FP1)}
translate P1
}
RG
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Mike Williams wrote:
> Wasn't it Tyler Eaves who wrote:
>
>>I'm looking for some way to specify a box at an angle. The way I see it, a
>>box could be specified by giving 2 points, which would be the center of
>>two oppisite faces, and would define one dimesion, call it length, and
>>then giving the length of the width and height. This would be roughly
>>comprable to the way a cylinder is specified. I can't seem to think of any
>>obvious way to do this (3d math is not a strength of mine...)
>
>
> The first thing that springs to mind is to write a macro like this:
>
> #include "transforms.inc"
> #macro aligned_box(P1, P2, H, W)
> #local L = abs(vlength(P2-P1));
> box {<0,-H/2,-W/2><L,H/2,W/2>
> transform {Reorient_Trans(x,P2-P1)}
> translate P1
> }
> #end
>
Hey cool! Thanks Mike, I was just trying to figure out how to make a
box that acted like a cylinder!
RG - I can't count the times I've had a question to ask here that I
never asked because someone had already posted the answer
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Wasn't it Tyler Eaves who wrote:
>On Wed, 07 Sep 2005 14:58:11 -0400, Mike Williams
><nos### [at] econymdemoncouk> wrote:
>
>> Wasn't it Tyler Eaves who wrote:
>>> I'm looking for some way to specify a box at an angle. The way I see
>>> it, a
>>> box could be specified by giving 2 points, which would be the center of
>>> two oppisite faces, and would define one dimesion, call it length, and
>>> then giving the length of the width and height. This would be roughly
>>> comprable to the way a cylinder is specified. I can't seem to think of
>>> any
>>> obvious way to do this (3d math is not a strength of mine...)
>>
>> The first thing that springs to mind is to write a macro like this:
>>
>> #include "transforms.inc"
>> #macro aligned_box(P1, P2, H, W)
>> #local L = abs(vlength(P2-P1));
>> box {<0,-H/2,-W/2><L,H/2,W/2>
>> transform {Reorient_Trans(x,P2-P1)}
>> translate P1
>> }
>> #end
>>
>
>That comes VERY close. Actually, it works as I requested, although missing
>an implied criteria. I didn't specify it, but I need the box to maintain a
>vertical orientation, whereas your macro induces a rotation around the
>P1-P2 axis. I'm guessing this is an artifact of the Reorient_Trans
Point_At_Trans seems to produce boxes that are vertically aligned more
of the time, but I can't predict which of the directions is going to be
the height and which will be the width
#include "transforms.inc"
#macro aligned_box(P1, P2, H, W)
#local L = abs(vlength(P2-P1));
box {<-W/2,0,-H/2><W/2,L,H/2>
transform {Point_At_Trans(P2-P1)}
translate P1
}
#end
--
Mike Williams
Gentleman of Leisure
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> That comes VERY close. Actually, it works as I requested, although missing
> an implied criteria. I didn't specify it, but I need the box to maintain a
> vertical orientation, whereas your macro induces a rotation around the
> P1-P2 axis. I'm guessing this is an artifact of the Reorient_Trans
Replace Reorient_Trans with this macro and it should work (untested code,
let me know if it doesn't):
// maps the x axis to vec while keeping "sky" as up,
// doesn't work if vec is parallel to sky
#macro Reorient_Trans_Sky(vec, sky)
#local mapxto = vec;
#local mapzto = vnormalize(vcross(sky, vec));
#local mapyto = vcross(mapxto, mapzto);
matrix <
mapxto.x, mapxto.y, mapxto.z,
mapyto.x, mapyto.y, mapyto.z,
mapzto.x, mapzto.y, mapzto.z,
0,0,0
>
#end
Use it like Reorient_Trans_Sky(P2 - P1, y)
- Slime
[ http://www.slimeland.com/ ]
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Wasn't it Slime who wrote:
>> That comes VERY close. Actually, it works as I requested, although missing
>> an implied criteria. I didn't specify it, but I need the box to maintain a
>> vertical orientation, whereas your macro induces a rotation around the
>> P1-P2 axis. I'm guessing this is an artifact of the Reorient_Trans
>
>Replace Reorient_Trans with this macro and it should work (untested code,
>let me know if it doesn't):
>
>// maps the x axis to vec while keeping "sky" as up,
>// doesn't work if vec is parallel to sky
>#macro Reorient_Trans_Sky(vec, sky)
> #local mapxto = vec;
> #local mapzto = vnormalize(vcross(sky, vec));
> #local mapyto = vcross(mapxto, mapzto);
> matrix <
> mapxto.x, mapxto.y, mapxto.z,
> mapyto.x, mapyto.y, mapyto.z,
> mapzto.x, mapzto.y, mapzto.z,
> 0,0,0
> >
>#end
>
>Use it like Reorient_Trans_Sky(P2 - P1, y)
A quick test reveals that "sky" is a reserved word, and the mapxto and
mapyto need to be vnormalized. So:
#macro Reorient_Trans_Sky(vec, Sky)
#local mapxto = vnormalize(vec);
#local mapzto = vnormalize(vcross(Sky, vec));
#local mapyto = vnormalize(vcross(mapxto, mapzto));
matrix <
mapxto.x, mapxto.y, mapxto.z,
mapyto.x, mapyto.y, mapyto.z,
mapzto.x, mapzto.y, mapzto.z,
0,0,0
>
#end
#macro aligned_box(P1, P2, H, W)
#local L = abs(vlength(P2-P1));
box {<0,-H/2,-W/2><L,H/2,W/2>
transform {Reorient_Trans_Sky(P2-P1,y)}
translate P1
}
#end
--
Mike Williams
Gentleman of Leisure
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Thu, 08 Sep 2005 04:07:01 -0400, Mike Williams
<nos### [at] econymdemoncouk> wrote:
> #macro Reorient_Trans_Sky(vec, Sky)
> #local mapxto = vnormalize(vec);
> #local mapzto = vnormalize(vcross(Sky, vec));
> #local mapyto = vnormalize(vcross(mapxto, mapzto));
> matrix <
> mapxto.x, mapxto.y, mapxto.z,
> mapyto.x, mapyto.y, mapyto.z,
> mapzto.x, mapzto.y, mapzto.z,
> 0,0,0
> >
> #end
> #macro aligned_box(P1, P2, H, W)
> #local L = abs(vlength(P2-P1));
> box {<0,-H/2,-W/2><L,H/2,W/2>
> transform {Reorient_Trans_Sky(P2-P1,y)}
> translate P1
> }
> #end
Works great! This is insanely useful.
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Thu, 08 Sep 2005 00:40:14 -0400, gonzo <rgo### [at] lansetcom> wrote:
> Hey cool! Thanks Mike, I was just trying to figure out how to make a
> box that acted like a cylinder!
>
> RG - I can't count the times I've had a question to ask here that I
> never asked because someone had already posted the answer
>
It's such a (to me) obvious thing that I'm surprised there isn't built in
support for it. It's worth pointing out that the code as posted by Mike
isn't quite perfect, but the code he posted later in reply to Slime *is*.
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> A quick test reveals that "sky" is a reserved word, and the mapxto and
> mapyto need to be vnormalized. So:
Thanks for testing. I don't think mapyto needs to be normalized, since it's
the cross product of two normalized, perpendicular vectors. Doesn't hurt
though.
- Slime
[ http://www.slimeland.com/ ]
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|
