Lars Grüter

Okay, to summarize the action plan: - Deprecate `is_rgb_like` and `adapt_rgb`. - Improve docstring of `skimage.transform.rescale`.

Well, apparently [clang 15.0.7 is currently available in the CI environment](https://dev.azure.com/scikit-image/scikit-image/_build/results?buildId=10222&view=logs&j=feb0b66f-d63c-5533-bf34-a72e08b12053&t=0a19090f-74e9-594e-2229-1bd445296bf2&l=39). I wonder what is causing the new requirement for 16.0.0. [Clang 16.0.5 would be available via chocolately](https://community.chocolatey.org/packages/llvm), I wonder...

Hmm, apparently a faulty setup in the image, see https://github.com/actions/runner-images/issues/8125. The fix is already merged and should arrive soon, so the issue should resolve itself "TM" at some point. :crossed_fingers:...

I think this was fixed in #7200. And anyway, it's no longer a concern with the CI currently.

Hmm, I certainly not grasping everything fully. But as far as I understand the example, `ORB` is used to determine the points in each image and not `FundamentalMatrixTransform` itself. Instead,...

Re https://github.com/scipy/scipy/pull/19814#issuecomment-1879762067, thanks for pinging me @ilayn and sorry for the somewhat late response. I don't think I will have much to add here. I'll keep an eye on this...

> peak method is the only method that can yield asymmetric envelopes. [...] Unfortunately, scipy does not have similar data in its dataset, so I haven't added an example for...

Thanks for pointing this out @ns-rse. It seems to me that both parameters allow for the same end result; is that assumption correct? So I'd even go so far that...

Unless I misunderstand your point @ns-rse, I don't think the math will be a problem? Simply multiplying with `-1` seems to align the directions. E.g. ```python import skimage as ski...

I'm totally with you on this. And if you want, please take a look and see if you can align both algorithms. :) Though, there's the requirement, that we don't...