If you've seen the sample illustration for SphereSweep in the Object 
Collection, you have noticed that the cockpit is empty.  I plan to 
rectify this, so long as The Google has not (yet) unveiled its 
autonomous airplane.  The white rim at the top of the hat presents 
another opportunity to show off how the module works with trace().

Of course, a photorealistic human being would be its own Object 
Collection module.

All rights reserved for this illustration, but the SDL for the completed 
figurine will eventually find its way into the GNU-LGPL.  When that will 
be depends on how soon I grok the math behind B-splines.

Post a reply to this message

Attachments:
Download 'design_pilot.jpg' (48 KB)

Preview of image 'design_pilot.jpg'

On 8/10/2015 9:52 PM, Cousin Ricky wrote:
> If you've seen the sample illustration for SphereSweep in the Object
> Collection, you have noticed that the cockpit is empty.  I plan to
> rectify this, so long as The Google has not (yet) unveiled its
> autonomous airplane.  The white rim at the top of the hat presents
> another opportunity to show off how the module works with trace().
>
> Of course, a photorealistic human being would be its own Object
> Collection module.
>
> All rights reserved for this illustration, but the SDL for the completed
> figurine will eventually find its way into the GNU-LGPL.  When that will
> be depends on how soon I grok the math behind B-splines.


:-)

-- 

Regards
     Stephen

Post a reply to this message

Cousin Ricky <ric### [at] yahoocom> wrote:
> If you've seen the sample illustration for SphereSweep in the Object
> Collection, you have noticed that the cockpit is empty.  I plan to
> rectify this, so long as The Google has not (yet) unveiled its
> autonomous airplane.  The white rim at the top of the hat presents
> another opportunity to show off how the module works with trace().
>
> Of course, a photorealistic human being would be its own Object
> Collection module.
>
> All rights reserved for this illustration, but the SDL for the completed
> figurine will eventually find its way into the GNU-LGPL.  When that will
> be depends on how soon I grok the math behind B-splines.

I'm studying your patch.

Post a reply to this message

Cousin Ricky <ric### [at] yahoocom> wrote:
> If you've seen the sample illustration for SphereSweep in the Object
> Collection, you have noticed that the cockpit is empty.  I plan to
> rectify this, so long as The Google has not (yet) unveiled its
> autonomous airplane.  The white rim at the top of the hat presents
> another opportunity to show off how the module works with trace().
>
> Of course, a photorealistic human being would be its own Object
> Collection module.
>
> All rights reserved for this illustration, but the SDL for the completed
> figurine will eventually find its way into the GNU-LGPL.  When that will
> be depends on how soon I grok the math behind B-splines.

thank you.
I'm glad someone finally got around to this.
[ : ) ]

Post a reply to this message

On 08/10/2015 04:52 PM, Cousin Ricky wrote:
> [...]  the SDL for the completed
> figurine will eventually find its way into the GNU-LGPL.  When that will
> be depends on how soon I grok the math behind B-splines.

Progress on the B-spline front.  The gray curves are POV-Ray native 
sphere_sweeps.  The colored circles were positioned using a rather 
profligate algorithm, but I have to understand the math a bit more than 
I do now before I can optimize it.

Post a reply to this message

Attachments:
Download 'b-spline_tryout.png' (19 KB)

Preview of image 'b-spline_tryout.png'

Le 14/08/2015 00:52, Cousin Ricky a écrit :
> On 08/10/2015 04:52 PM, Cousin Ricky wrote:
>> [...]  the SDL for the completed
>> figurine will eventually find its way into the GNU-LGPL.  When that will
>> be depends on how soon I grok the math behind B-splines.
>
> Progress on the B-spline front.  The gray curves are POV-Ray native
> sphere_sweeps.  The colored circles were positioned using a rather
> profligate algorithm, but I have to understand the math a bit more than
> I do now before I can optimize it.
>

If it can help you, for all spline types, despite the visual effect, 
there is no influence of one coordinate axis on any other axis.

Also, stick to one kind of spline when learning, you will be able to 
generalize later, but mastery of one type (instead of a small bit of 
many types) is the path that is probably the best.

Post a reply to this message

On 2015-08-14 03:33 AM (-4), Le_Forgeron wrote:
> If it can help you, for all spline types, despite the visual effect,
> there is no influence of one coordinate axis on any other axis.

So I noticed, and it has already helped.  In SphereSweep 1.0, the code 
was simplified by treating the radius as merely a 4th dimension tacked 
onto the spline position.

Post a reply to this message

On 2015-08-14 03:33 AM (-4), Le_Forgeron wrote:
> Also, stick to one kind of spline when learning, you will be able to
> generalize later, but mastery of one type (instead of a small bit of
> many types) is the path that is probably the best.

I've already had a decent grasp of Bézier splines for several years now, 
but have not included them in SphereSweep because they are not supported 
by the sphere_sweep primitive.  As my original rationale for writing 
this module was to circumvent the flaws in the sphere_sweep primitive, I 
decided to prioritize B-splines.

However, with interest in Bézier splines awakened in the past couple of 
weeks, they are definitely on the table for a future version.  I already 
have a sketchy idea of how to make them 1st-derivative continuous with 
little pain to the user.

Natural spline and quadratic spline sphere_sweeps aren't natively 
supported either, but since they /are/ supported by the experimental 
spline feature, they were trivial for me to include from day one.

As the experimental spline feature already supports linear and cubic 
splines, and as my brain is not as agile as it was when I was 11, I feel 
no pull to try to understand these.  (Except than the linear spline is 
intuitively obvious.)

Post a reply to this message