Mateusz Łoskot

@meshtag Could you tell what is the status of this PR? Is this going to be picked up or drop it and close it?

Sounds good. Yes, there has been lots of activity hence the conflicts. If you will have trouble with resolving them and need help, let me know.

You've made good point about the parameters checking. ----- The this announcement of [Formal Review of IO and Toolbox extensions to Boost.GIL](https://lists.boost.org/boost-announce/2010/11/0273.php) includes points provided by @chhenning - author of...

*** MARKING AS WIP *** On reflection, I'm postponing this until the whole `clang-tidy` refactoring is done. Moving files around while changing them may not be a good idea, esp....

@codejaeger Although manual verification using an example is good, it is not good enough. We need to add proper test for this particular case, then this issue can be closed.

@codejaeger No problem. > what kind of tests would you propose for these ? As @renatoGarcia explains the undesired behaviour is: > writing a gray16 PNG image results in image...

@codejaeger Can you try the same for JPEG and/or TIFF and compare? We need to identify if the issue occurs at I/O level or in the core.

> JPEG - gray16,rgb16 fails to compile, others pass Since support for 16 bpp is dependant on codec library, we may not allow/support it. Simple as that. For libjpeg itself,...

To summarise what we know now about the issue of 16-bit grayscale image I/O: 1. GIL write/read of TIFF works 2. GIL write of PNG and read of PNG fails...

@codejaeger Quick note, I really appreciate your efforts and please be assured I will respond as soon as I have a moment to check the issue in details.