Pretty new to darktable, so I have lots to learn. I’ve only started to delve into it so far.
I’m also about to replace my desktop with a laptop, so I’ll be tied up for a while getting that tweaked. But, the change from my desktop to a mostly-at-home-but-occasionally-mobile laptop has me curious how best to approach collections / importing, etc.
I have a NAS housing my images (in date-prefixed) directories which I manually create and populate with images from my camera. That process won’t change since it has other advantages. I import (but not copy) images into darktable. I could put all the images on my laptop, but the NAS is more reliable, it’s backed up, centralized and there’s no point in tying up laptop disk space. Even on my older / slower desktop I’ve never run into latency issues between it and the NAS so that’s not an issue.
So my plan is to leave it as-is, since I’m not gone that much. But ideally I’d like to be able to use darktable when not at home if needed. Unlike a non-DAM directory-browser-only tool such as RT, obviously images have to be imported for darktable.
Is there a way to accommodate potential mobility in a static environment as I’ve described? Or are they fundamentally two exclusive approaches: NAS or laptop, not NAS and laptop?
If you can handle a bit more technical complexity, check out git-annex, which fits this type of use case perfectly. There is even a lua script to help integrate it into darktable. Using git-annex would allow you to import files to your laptop, check them into gitamnex, then transfer the files to your NAS when you get home. Then drop the files from your laptop when you’re ready.
That’s how it’s done now. The NAS volume is shared both NTFS and NFS, since I have both Windows and LInux hosts mapping / mounting to it (Linux can do NTFS, but NFS works better). I’m wondering about effectively having multiple sources behind a single collection, more or less.
That sounds interesting and potentially applicable. I’ll check it out.
One of the reasons I want to keep my options open is that I’ve been using (and like) RawTherapee, but I want to learn more about darktable as well. I’m a low-volume amateur shooter, so as useful (essential, actually) as darktable’s DAM capabilities may be to pros and more active shooters, I really don’t generate enough images to benefit from it. But being an somewhat analytical-wannabe type I’ll probably end up tagging what I import anyway…
Since I was just evaluating darktable I had initially set XMP sidecars to edited only, but then entirely disabled them to reduce clutter (particularly in XnView MP). The same info is in the database anyway, right?. But it sounds like I’ll need them at least for every local copy. Is that correct? If so I can filter them out from XnView.
Although git-annex is more capable I think local copied will meet my needs. At least the option is there if I need it.
That’s not the point, think if the global database got corrupted? Are you ready to loose all your edits? And this just to avoid seeing .xmp using XnView !!! At least you are warned now.
Yes, but the database isn’t accessible for other programs, so tags, captions, and such won’t be visible in those other programs (when you work with raw files). And you’ll probably like to start tagging images sooner or later, if only to find back that picture you took of X about 10 years ago (but was it 2010 or 2013? Appears it was 2007 ) And captions and titles are always useful. I can’t remember details from pictures I took last year, let alone 10 years ago (and I have a lot of those, even as an amateur shooter).
Nor can you easily exchange edits with others without sidecar files. And as @Pascal_Obry indicated, sidecars give you another backup as well, for a rather low cost.
I’ll probably end up tagging what I import anyway…
Thanks.