Am Sonntag, 29. April 2018, 19:11:55 CEST schrieb Timur Irikovich Davletshin: > On Sun, 2018-04-29 at 18:41 +0200, Tobias Ellinghaus wrote: > > Am Donnerstag, 26. April 2018, 13:22:34 CEST schrieb Timur > > > > Davletshin: > > > https://drive.google.com/file/d/14sZLgnpZSV5W3pw1K_8owHW29Heq9EWz/v > > > iew?usp=d > > > rivesdk — there is nothing special about those sidecar files, it is > > > the same > > > I described earlier. They come from my lazy habit of editing mostly > > > only > > > first file in the set with same light conditions. Then I just copy > > > respective module settings and adjust individual like 'spot > > > removal' if > > > needed. > > > > Did you report that to redmine? I thing I remember it being mentioned > > but I > > can't find any ticket. > > > > I am thinking about only importing masks that are actually used in > > the image > > when loading a sidecar (I have the code for that already), but I am > > not sure > > if that is going to break some workflows where you want to keep > > masks. > > > > Tobias > > > > [...] > > No, I didn't. It was just in mailing list. > > I'm not sure if this is right solution. I can imagine that some masks > should be kept even if not used. But masks I talk about are not even in > the list...
I added some change to how masks are handled: - When resetting the spot module we delete all masks used internally in the module. That should prevent dt from piling up masks that are not shown in the mask manager. - When loading XMP files all cloning masks (the ones used in spot removal) that are not actually used by the image are ignored. With those it might be enough to reload your images to get rid of the big sidecar sizes. It would be great if you could try it after making a backup of your files. Just to be on the safe side ... > IMHO, as first step it would be better to have external tool to fix > those XMPs. And then fix code that copies unused masks along with > module settings which do not employ any masks at all. Tobias
signature.asc
Description: This is a digitally signed message part.
