Please see the attached tarballs, previous announcements in this forum 
and INSTALL.txt for suggestions on compiling and running via wrapper script.

Note. The yuqk_DocAndAid_3317dfc0.tar.gz and 
yuqk_Examples_3317dfc0.tar.gz files will be attached to a response to 
this posting.

A reminder the yuqk fork is not POV-Ray. As time passes, it's less 
likely existing POV-Ray scenes will work without modification.

Note. Those using g++ compiler version 13 or later must add the compiler 
flag -fno-finite-math-only, if using the flag -ffast-math.

Note. Those compiling to the c++20 (or later) standard must currently 
use the -fno-char8_t compiler flag.

Bill P.

---

---------------- povray-3.8.0-x.yuqk_e6e987ab.tar.gz  Novemeber 04, 2024
yuqk R16 v0.6.10.0

Moved dens_object pattern 'exponent' keyword to 'ip_exponent'. Moved 
dens_object and hard_object pattern keywords: 'radius', 
'recursion_limit' and 'samples' to 'ip_radius', 'ip_recursion_limit' and 
'ip_samples', respectively.

Added the inbuilt function f_popnrm_rnoise(). Fast, normal-ish fit 
random noise values based upon x,y,z coordinates and a user seed value.

The dependency on boost/functional/hash/hash.hpp (boost::hash_combine() 
and boost::hash()) has been removed.

The crackle scalar value pattern largely re-written to address thread 
cache collisions and a small collection of issues and bugs. Though the 
re-write is both faster and slower, for the traditional scenes at small 
scale downs (large crackle cells) it tends to be slower. Of note the 
dependency the ip_solid option had on the noise generators was removed. 
The new caching is currently not optimal as it is set up to experiment 
with larger cell counts around the center cell.

Added ip_cache, ip_exponent and ip_seed features to the crackle pattern.

Updates to support OpenEXR 3.0 changes. Code adapted from that 
implemented by Ton van den Broek (TonVanDenBroek) and Christoph Lipka 
(c-lipka) as the merged github pull #423 against the official v3.8 beta 
2 release. Note. Input and output remain at 16 bits / half floats rather 
than 32 bits.

Changes to INSTALL.txt suggested performance build flags which include 
-ffast-math to also include -fno-finite-math-only given the g++ version 
13 compiler turns on finite-math-only and many g++ implementations of 
finite-math-only causes various errors and fails while rendering. Some 
of these fails might not be immediately apparent.

Post a reply to this message

Attachments:
Download 'povray-3.8.0-x.yuqk_3317dfc0.tar.gz' (1649 KB)

On 11/4/24 01:27, William F Pokorny wrote:
> Note. The yuqk_DocAndAid_3317dfc0.tar.gz and 
> yuqk_Examples_3317dfc0.tar.gz files will be attached to a response to 
> this posting.

Files attached to this post.

Bill P.

Post a reply to this message

Attachments:
Download 'yuqk_docandaid_3317dfc0.tar.gz' (635 KB) Download 'yuqk_examples_3317dfc0.tar.gz' (462 KB)

hi,

William F Pokorny <ano### [at] anonymousorg> wrote:
> Please see the attached tarballs, previous announcements in this forum
> and INSTALL.txt for suggestions on compiling and running via wrapper script.
>
> Note. The yuqk_DocAndAid_3317dfc0.tar.gz and
> yuqk_Examples_3317dfc0.tar.gz files will be attached to a response to
> this posting.

thanks.  builds and 'make check' ok, on Slackware 15.


regards, jr.

Post a reply to this message

On 11/5/24 12:15, jr wrote:
> thanks.  builds and 'make check' ok, on Slackware 15.

Thank you for testing it! :-)

Bill P.

Post a reply to this message

On 11/4/24 01:27, William F Pokorny wrote:
> Please see the attached tarballs

For the record.

There are problem(s) in this release (and my development space) with 
'some' normal{} use. I now have one case where the normals just don't 
work and a second which core dumps.

I've not got a clue what change(s) might have tripped this behavior. I 
don't believe I touched anything related to normals for this last 
release. :-(

Bill P.

Post a reply to this message

On 11/19/24 19:26, William F Pokorny wrote:
> On 11/4/24 01:27, William F Pokorny wrote:
>> Please see the attached tarballs
> 
> For the record.
> 
> There are problem(s) in this release (and my development space) with 
> 'some' normal{} use. I now have one case where the normals just don't 
> work and a second which core dumps.
> 
> I've not got a clue what change(s) might have tripped this behavior. I 
> don't believe I touched anything related to normals for this last 
> release. :-(
> 

Attached is a patched source code tarball. After compiling,  if one runs 
the version.pov file shipped with yuqk it will show v0.6.10.1 rather 
than v0.6.10.0.

I was away from POV-Ray / yuqk coding and play for more than two months 
after the early July yuqk release. I forgotten I'd slapped at some code 
faster than using a dynamic_pointer_cast for possible normal blend maps 
on every Perturb_Normal() call. The code wasn't supposed to be released 
yet, but I let it slip by...

In my development space now trying another method. This time using a 
'hard assert' in BOTH the performance compile and debug compile. The 
debug compile of v0.6.10.0 had an assert which immediately pointed to 
the problem for scenes showing issues, but the assert was only in the 
debug compile. FWIW.

Bill P.

Post a reply to this message

Attachments:
Download 'povray-3.8.0-x.yuqk_3317dfc0_patch1.tar.gz' (1668 KB)