Even Rouault
Even Rouault
> Is there a standard GDAL thread pool facility to use? yes, we already discussed it with Andrew a bit: https://gdal.org/doxygen/classCPLWorkerThreadPool.html . It has probably not all the bells and...
> I can push a branch if you'd like. yes I can give a try
> Here is a branch that removes the threads for each line: https://github.com/abellgithub/gdal/tree/lessthread much better! runs in ~ 15.8 s in the Windows VM . Actually the key change is...
> unless you changed something in the branch above I didn't. Basically in my Windows VM: - https://github.com/abellgithub/gdal/commit/6e1a4ef06a68a2e8a0b87c310355399dd40b66e3 (before async) : runs in ~ 58 s - https://github.com/abellgithub/gdal/commit/05f1967d108c2bb64a3d0331c957b3a065516241 (async) :...
> but I'll defer to you ok, let's drop that complication (special behaviour for one thread) for now and see if someone complains. I believe however we need the std::async...
> I don't understand the CI failure. Perhaps just a system timeout? that one is a known flake. I've restared it > Also, I added the pre-commit hooks but I...
cf CPLGetValueType() : https://gdal.org/api/cpl.html#_CPPv415CPLGetValueTypePKc
FYI, swig/python/README.rst is what is used to put in the "sdist" package and is displayed on https://pypi.org/project/GDAL/ Perhaps part of the information of this PR should be put also (or...
> Overall, I don't think a bit more info in README showing up is a problem. There might be a misunderstanding. I'm not against additions. My above comment was just...
> And I would think if moved, we should move the conda info too? I would keep the existing info in swig/python/README.rst (other people have already worked on those docs...