Incorrect out of gamut indicators in non-sRGB color spaces
When the GIMP-2.99 babl space invasion is complete, hopefully this won't be a problem. But as of commit 04bbe941 the situation is as follows:
-
Open a linear gamma Rec.20202 image and make a color patch that's out of gamut with respect to GIMP's built-in sRGB color space. Rec.2020's greenest green, reddest red, and bluest blue will do nicely, but there are many less saturated colors that also will do, the only requirement is that the color is out of gamut with respect to sRGB and in-gamut with respect to Rec.2020.
-
Color pick the color patch using the FG/BG tool. Also add Sample Points and use the Pointer dialog to hover over the color patch:
Actual result is that the Pixel, and LCH (and presumably LAB, I forgot to check) readouts are correct for all the dialogs that read colors from the image, which is good. But all the other ways to read colors ("RGB %", "RGB 0-255", HSV - I'm ignoring CMYK readout), still give results that would only be correct if the image were converted to sRGB. Also, any color that's out of gamut (out of display range) with respect to sRGB is still marked with the upper right corner triangle as "out of gamut", even though the image isn't really in the sRGB color space.
Expected results are that all the ways to read and display color information from an image should give results relative to the image's actual color space rather than "as if" the image were in the sRGB color space. Also, colors shouldn't be marked with the little triangular "out of gamut" indicators unless the color really is out of gamut WRT the image's actual current color space.
Speaking for myself I'd cheerfully ignore incorrect "converted to sRGB" readouts for "RGB %", "RGB 0-255", and HSV (well, I already ignore these readouts). But that out of gamut triangular indicator is very attention-grabbing and is the sort of thing that really ought to mean "ought of gamut for the actual image's actual color space", not "would be out of gamut if the image were converted to sRGB".
Regarding HSV, maybe digital artists would prefer "in terms of sRGB" so that they know how to judge "what sRGB HSV hue", "how sRGB HSV Saturated" and such. I'm not qualified to even suggest a course of action because I don't use HSV and so have no expectations as to how or whether to somehow link HSV readouts in "sRGB only GIMP" with HSV readouts in "anyRGB GIMP".
Here's the linear gamma XCF file with the sample green color: