Thanks for the explanations Larry, It is what I was guessing.
Disclaimer: I'm totally pedantic at work ;)
On Fri, Nov 10, 2017 at 8:01 PM Larry Gritz wrote:
>
> On Nov 9, 2017, at 8:00 PM, Thomas Mansencal
> wrote:
>
> The fact it is backward incompatible raised to me the question of OIIO
> v
> On Nov 9, 2017, at 8:00 PM, Thomas Mansencal
> wrote:
>
> The fact it is backward incompatible raised to me the question of OIIO
> versioning scheme. What is the current approach since it does not seem to
> follow semantic versioning?
I think for the pedantic, "semantic versioning" has th
It wasn't too bad. A couple weekends of screw-around time. It's similar in
spirit (and syntax) to Boost.Python, but cleaner. I'm very bullish on pybind11
now, I would definitely recommend it for a new project that needs python
bindings. It was perhaps a little less of a clear win for OIIO that h
Wenzel Jakob, the man is everywhere :)
The fact it is backward incompatible raised to me the question of OIIO
versioning scheme. What is the current approach since it does not seem to
follow semantic versioning?
Cheers,
Thomas
On Fri, Nov 10, 2017 at 3:58 PM Jonathan Gibbs wrote:
> Oooh, pybi
Oooh, pybind11 hasn't crossed my path before -- very interesting. How much
work was it to move?
--jono
On Thu, Nov 9, 2017 at 5:48 PM Larry Gritz wrote:
> Those of you who have the OIIO Python bindings on your critical paths,
> please be aware of this outstanding PR:
>
> https://github.
Those of you who have the OIIO Python bindings on your critical paths, please
be aware of this outstanding PR:
https://github.com/OpenImageIO/oiio/pull/1801
in which I refactor the python bindings to use pybind11 instead of
Boost.Python, and in the process make some changes to our Pytho
