fwiw, some field line based DLA...

Chris M. Thomasson

unread,

Apr 29, 2016, 6:46:46 PM4/29/16

to

FWIW, some, IMHO, fairly interesting C++ code for Field Based DLA:

https://plus.google.com/101799841244447089430/posts/am5EpMHcmig

https://github.com/ChrisMThomasson/CT_fieldDLA

https://en.wikipedia.org/wiki/Diffusion-limited_aggregation

that can be heavily multithreaded!

the code as is, creates the following rendering:

https://plus.google.com/101799841244447089430/posts/DfjAJbbdLGp

Four C++ files, that use the Cairo graphics lib along with the GLM library:

http://cairographics.org/

http://glm.g-truc.net/0.9.7/index.html

No need to follow random walks, follow field lines for DLA?

C++ seems like a _very_ nice place to implement this.

:^)

Chris M. Thomasson

unread,

Apr 30, 2016, 3:46:14 AM4/30/16

to

> "Chris M. Thomasson" wrote in message
> news:ng0o85$19em$1...@gioia.aioe.org...

> FWIW, some, IMHO, fairly interesting C++ code for Field Based DLA:

> https://plus.google.com/101799841244447089430/posts/am5EpMHcmig

> https://github.com/ChrisMThomasson/CT_fieldDLA

[...]

BTW, if you have any questions about the code,
please ask them. I think we can make this
algorithm much more efficient, humm...

https://github.com/ChrisMThomasson/CT_fieldDLA/blob/master/cairo_test_penrose/ct_field.hpp

The (ct::field::search_hit) function can be
improved, along with (ct::field::calc_point).
These are very important "hot spots" in the
(ct::field::render) function.

If you are interested, and think this is off topic,
we can move it to:

https://plus.google.com/101799841244447089430/posts/DfjAJbbdLGp

However, it is C++ with Cairo and GLM, not that bad
for third part libs...

If I confine posts to this thread only, is it okay to post updates here?

;^o

Chris M. Thomasson

unread,

May 2, 2016, 4:11:03 PM5/2/16

to

> "Chris M. Thomasson" wrote in message
> news:ng0o85$19em$1...@gioia.aioe.org...

> FWIW, some, IMHO, fairly interesting C++ code for Field Based DLA:

> https://github.com/ChrisMThomasson/CT_fieldDLA

> that can be heavily multithreaded!

[...]

FWIW, here is a proxy garbage collector than can be fairly
useful for multi-threaded read-side optimizations.

Think RCU here. The code is here:

http://pastebin.com/f71480694
(uses Relacy Race Detector:
http://www.1024cores.net/home/relacy-race-detector)

How many people have ever used Relacy? FWIW, here is a video
of the man who created it:

https://www.youtube.com/watch?v=a9xrxRsIbSU
(very smart!)

more context on the simple proxy collector I created:

https://groups.google.com/d/topic/lock-free/X3fuuXknQF0/discussion
(fwiw, there are some very smart people in that thread!)

I am thinking about using this for the multi-threaded Field
DLA algorithm.

The vector field can be a shared, dynamic, read-mostly
data-structure protected by this, portable C++ proxy garbage
collector.

Chris M. Thomasson

unread,

May 8, 2016, 3:35:03 PM5/8/16

to

> "Chris M. Thomasson" wrote in message
> news:ng0o85$19em$1...@gioia.aioe.org...

> FWIW, some, IMHO, fairly interesting C++ code for Field Based DLA:
> https://plus.google.com/101799841244447089430/posts/am5EpMHcmig
> https://github.com/ChrisMThomasson/CT_fieldDLA

FWIW, here are some new renderings I created using a
mutated version of the DLA field code:

https://plus.google.com/101799841244447089430/posts/9bz3jgBxAPf

https://plus.google.com/101799841244447089430/posts/7LQhwQVi86V

https://plus.google.com/101799841244447089430/posts/YUBEPZTE5hN

IMVVVHO, this is interesting, and C++ is wonderful medium!

:^)