Thanks to dcraw source code, Adobe Photoshop and dozens of other image tools now have built-in support for raw photos, and the popularity of raw photography has grown enormously.
Does this mean that PS used dcraw on Camera Raw? Does someone know more details about it?
To end up with a usable photo, a program needs to:
Decode the raw file from mostly proprietary formats (NEF, CR2, ARW, PEF, etc.) or open formats (DNG) to get the raw image, typically using dcraw, libraw or rawspeed.
Get the metadata parameters required to work with the image (white levels, black levels, the coordinates of the usable image without the surrounding black frame if any, bit depth, color matrix, etc.), either from the decoded raw file or from external sources (exiv2, camconst.json, etc.).
Demosaic the raw image if it needs demosaicing (needed when the camera uses a Bayer or X-Trans color filter array, not needed for Foveon sensors and Pixel Shift images without motion correction).
White-balance the image.
Pump the image through a series of manipulations in appropriate colorspaces to make it look good.
dcraw was last updated in 2016, no reply from Dave Coffin.
RT uses dcraw and will adopt something else hopefully soon, darktable uses mostly rawspeed.
What Adobe uses is a mystery to me, but I doubt a company with ~15000 employees and probably over 100 full-time programmers, a world leader in the field with a lot of money and access to all camera manufacturers, would rely on dcraw which has a single point of failure (one developer) and which was created through reverse-engineering, when it could just buy the needed information and write the code itself.