Hacker News Reader

Comment by geppetto | Hacker News Reader

IMHO it's the best FOSS photo application out there. It's cross platform, fast, stable and manages the whole photo workflow (RAW developing, photo management, non destructive editing).

Thanks darktable folks!

View on HN

Agree. I would add three more qualities to your list:

* Fully programmable, via a nicely document Lua interface: http://www.darktable.org/lua-api/index.html.php -- you can automate all kinds of workflows.

* Truly innovative, with clever new features unavailable in other applications, including Lightroom. For example, five years ago they added a powerful new "equalizer" module: http://www.darktable.org/2011/11/darktable-and-research/ (google "darktable equazlier" for usage examples).

* Always improving: I've been using it for years and the developers keep finding ways to make it better all the time.

Congratulations on another great release!

Definitely agree!

I make a substantial part of my income (just enough to pay rent) selling photography, and I recently dropped ps and lightroom in favor of Darktable. The change didn't happen in a day, but Darktable is so good I don't think I'll ever need to come back to Adobe. Notable plus, I can now spend most of my time on Linux, no more need to use my Macbook for photo work.

The interface is not really good for smaller displays and resolutions (adobe stuff was better on my 11 inch mba), but on anything with more than 1280 x 960 px, it's great.

Congratulations to the Darktable team for this 2.2 release!

Any experience using RawTherapee also? Just an hobbyist myself and moving away from the Mac platform so figuring out the best alternatives.

Never tried RawTherapee, won't be able to help here sorry!

I don't make a living via photography, but I do receive income from taking pictures of escorts, and pets. (Strange how you find your niche!)

I've been using RawTherapee, on Linux, for the past few years. For my needs of marking images, doing post-processing, and mass-operations, it works perfectly.

I'm always reluctant to allow a tool to manage the layout/tagging of my images. So for that I have a strictly sorted hierarchy:

    ~/Images/
    ~/Images/2016
    ~/Images/2016/12/04-Hannah
    ~/Images/2016/12/12-Tiffany

Inside each "event"-folder I store:

    RAW/
    JPG/
    JPG/thumbs/
    Teasers/
    Teasers/thumbs/

Then I write out a meta-data file, ".meta", with detail such as:

    Location=Studio|MyPlace|Their Hotel|Whatever
    Contact-Details=+44...
    Tags=lingerie,nude,shillouette,monochrome
    ..

It works well, and allows me to find files years later via a simple `grep`.

I am also a hobbyist and I use Darktable. I've experimented with RawTherapee. I like RawTherapee's editing controls and love that its sidecar files are somewhat human readable (Darktable's are not).

What Rawtherapee lacks is the 'light table' feature for handling bulk photographs. I'll often shoot several hundred images in a day (even a blind squirrel finds a nut sometimes). Darktable lets me triage those photos and filter them for processing without managing a file system... well now that my workflow uses RapidPhotoDownloader on the front end. It reads off the memory card and renames files and directories to a standard format.

A few months ago, I brought many years worth of sorted directories of images into Darktable and went to bed and when I woke up it had added everything to the database and I could sort by name and remove duplicates and look at thumbnails and remove junk that had been saved alongside stuff I wanted during bulk operations.

Anyway, since Darktable runs on Mac, there's not a drop dead reason not to give it a spin alongside whatever you're currently using.

I'm sure you're sick of getting asked this, since it's asked every time someone who makes some money from photography mentions it, as there are many wannabe photogs out there, but how do you monetize your photos? Stock photo marketplaces like Dreamtime? Event photography (i.e., getting paid to take photos and provide prints of said photos, not getting paid for photos directly)? Online marketplaces for more artistic-oriented photos like 500px?

Personally my photos provide indirect value. People don't pay for them, but they drive traffic to a property that people pay to use. I've dabbled with the concept of selling them directly or taking more stock-style photos, but never really gotten serious about it.

Again, I know this question is annoying, so if you don't wanna answer, no biggie. :)

Many people ask the question indeed!

I sell prints and license photos to a small list of corporate clients I slowly grew with time. I haven't been lucky with marketplaces so far, and I'm doubtful it would be worth the effort with my type of photography (abstract/achitecture).

Any monetization strategy is probably too tied to the corresponding style of photography for any generic advice to make sense. A few of my friends have their income coming mostly (or 100%) from photography, and not three of them monetize the same way. None of them use any marketplace though. Artists, fashion photographers, event photographers, social media content producers, all have a different business model which more often than not looks like classic b2b.

A lot of this comes down to your niche. As you say there are many people who are "photographers", and many many many more who would never pay and take their own with their phones.

I accidentally found a niche when I took some fun pictures of a friend who was an escort. She shared them, for advertising purposes, and that lead to my name being mentioned on an escort discussion forum. Since then people get in touch every now and again and I arrange to shoot them in exchange for cash.

Finding people willing to pay is definitely hard, but sadly my own route is pretty hard to generalise.

> It's cross platform

Are there Windows builds? There was a guy that did but he stopped at 2.0.

Oops! You're right there's no official Windows port. I've run it on Mac and Linux only.

I've been able to run it on the Linux Subsystem for Windows with an X server, but there was still a few quirks I believe.

There's been renewed work on a clean Windows port in recent months: https://github.com/darktable-org/darktable/pull/1327

Interesting! They almost lynched the guy at the beginning but seems that there's progress.

It is very interesting. I don't want to intrude on their thread, but if others here vouch for this technique I'll politely ask them to look into the following:

- https://blog.jessfraz.com/post/docker-containers-on-the-desk...

- http://boot2docker.io/

...has anyone here ever used this approach with nix apps that need to run on Windows "natively"? Writing a powershell script to silently install boot2docker with this special docker image wouldn't be too difficult. The DarkTable folks don't seem interesting in maintaining Win builds so this may be the perfect workaround.

NOTE: I realize that `File > Open Image` may be non-trivial, but surely there's a way to mount a host<->guest folder (e.g.: When using the application, the workspace is located in `C:\Users\$whoami\Documents\DarkTable\.jpg`)

Boot2Docker is not for this purpose. Silently installing a VM on a Windows PC is a very bad idea. If the user is not proficient enough to know about and configure the VM, then what happens if

- B2D installs a Virtualbox instance, but Hyper-V is on. VBox can't run besides Hyper-V.

- How do you configure the resource usage for the VM?Something like Darktable requires a lot of resources

- How do you manage shared folders? VBox shared folder performance is really bad compared to native

- Do you install an X server as well?

Anyway -- it's just a bad idea.

Thanks sz4kerto. I'm glad I got some thoughts here before posting that in their PR (I obviously won't post it now because of the points you brought up).

I guess a user that would be comfortable with b2d or vbox would just have those running anyways!

Note: for you question about X, please check out https://blog.jessfraz.com/post/docker-containers-on-the-desk...

Wow you weren't kidding.

> It may be better to go help out some other free software project that already had the misfortune of working on that platform...

I understand their points about maintenance, but telling someone to go work on something else rather than keeping an experimental branch around, sheesh.

As someone who appreciates open source and works primarily on Windows I really wish they had a better attitude. Take a look at Rust or Elixir, fantastic communities and the software is better for it. I'm sure that Darktable has some interesting things but based on what I saw in that thread I think I'll be sticking to Lightroom.

I've only read that quote, but to me that doesn't seem like a rude thing to say. I interpret it as "This windows branch is awful, it might be better for you to abandon that effort and work on something with more potential". I hear it as criticizing the code, not the person.

Edit: I read a little more, looks like the devs were upset because windows isn't a free platform, and they didn't want darktable's reputation marred by a buggy port.

> This will be very bad for darktable reputation

It was in the context of building trust, basically "go away and work on another project so we know who you are before you can touch our holy code".

That guy came in with a great attitude and approach, he pretty much got told to pack it.

I dont believe in being rude to people, but sometimes you have to make things clear. In their comment chain they asked the dev to rename the application so it didn't give the reputation a bad name. I can understand that, especially since Darktable apparently is a really good app. Open source means you can take the code and build/modify/play with it, it doesn't mean you can take the name and reputation of a product and apply it in a new direction (using the same name).

So the guy forks darktable for windows, calls it something else. Then when the project reaches a level of maturity the original devs approve of, they take his work back into darktable, produce their own windows port under their 'brand' and his project potentially dies because it doesn't carry the same weight?

I got the vibe from some of the comments that "one guy working on a windows port isn't enough, we don't want to help support it so we're going to discourage any 'one man bands' until a team appears from nowhere to do all the work" is the common view, which I sort of understand that they don't want to have to provide end user support to a platform they themselves don't use, but that attitude might ultimately go against the project.

Exactly! Thanks for summarizing much better than I could.

You can set out requirements for merging upstream while also encouraging users to contribute. They're driving off good developers who do use windows primarily and might want to contribute. Like you said, there's no "team" that's going to materialize if that's the approach that they take whenever anyone brings it up.

I dont see anything wrong with what you just described.

    * Yes it's ok for the *owners* of the project to reject buggy code.
    * Yes it's ok for the owners of the project to accept working code.
    * Yes it's ok for the owners to "take" someones work when said person volunteered to it. 
    Thats what "contributing to open source" means.
    * Yes it's ok for the owners to use a working contribution to better their project for each and 
    every one of the users.
    * Yes its ok for someones fork to die.  Thats the fork maintainer's problem. 
    Is there anything preventing him from checking out the latest version of the code (serious question, 
    I dont know), the one that the owners incorporated and improved ?

Also, it's not "his" (the windows dev's) project. It's his contribution, which is to port the original owners' project to windows.

I dont see anything wrong here, other than the owners weren't exactly welcoming, encouraging, or perhaps helpful. That sucks, sounds like they were rude (or perhaps they turned rude when the original "dont call this fork Darktable, please" didn't get heeded ?).

What a shame, I've heard similar stories about people who try to contribute to open source and get treated badly. But again, other than being crass, I dont see any problem. It's great that windows dev contributed so much and improved. He deserves a big Thank You from everyone, especially the owners (and maybe they did ?). But the bottom line is he volunteered. He isn't owed anything.

It's an unfortunate situation, but the way _they_ put it (two sides to every fight) is that they don't know if he'll drop out in a week and all those who downloaded it will come screaming "Were's our updates!!"

It works in Windows 10 using the Linux for Windows feature, and then using a suitable windows X server.

http://i.imgur.com/MxadCpg.png

I regret that I haven't really had the opportunity to explore Darktable very thoroughly. Every time I would try to point it at my library folder it would choke to death. Perhaps I should re-organize around Darktable, but right now I have other components of my workflow that are hard to imagine replacing, primarily DxO -- this was the main driver that got me to switch to a Windows host running a Linux VM and doing all my real work over SSH to the local VM, after years of struggling through painfully slow editing in a Windows VM on a Linux host. I resent that Windows's bad, less flexible behavior has forced me to make it my host OS. :'(

At the moment I don't have a good library app, I just browse the directory structure. I've tried Lightroom but it too struggles with my library folder, although unlike Darktable, it is marginally usable. I've tried Digikam, and while it's the only one that seems to handle the library in a semi-respectable fashion (with the experimental, now-deprecated MySQL backend enabled, which iirc took some manual massaging to even make compile anymore), it's hard to get the components I really wanted to work reliably (it's very important the my camera's GPS tags can be mapped and sorted; this is my main interest in a library manager and as of now, all such programs are too overloaded to do this competently or half-broken like Digikam's integration, and when I need GPS-based information I guess I will have to try to write a script that reads the coords out manually with exiftool instead of trying out Yet Another Library Manager).

Adobe is absolutely that 1k lb gorilla in the room that everyone is afraid to compete with. Add to that that photography programs are very difficult and complex work and it makes the field, particularly the field of volunteer developers who make contributions to a FOSS project, pretty limited.

For the record, the current size of my library:

    [me@my_host photos]$ find . -name .snapshots -prune -o -iname "*.CR2" -print | wc -l | xargs printf "%'d\n"
    97,596
    [me@my_host photos]$ find . -name .snapshots -prune -o -iname "*.JPG" -print | wc -l | xargs printf "%'d\n"
    177,785
    [me@my_host photos]$ du -sh --exclude=./.snapshots
    2.8T    .

.snapshots being the btrfs snapshot tree, which makes these numbers about 16x higher, but is a relatively recent addition (around the time I moved to a Windows host) so isn't the reason everything has choked historically. Hopefully stuff like Lightroom isn't trying to parse these directories as well. I guess I could blacklist them from sharing in Samba to try to get around it.

Open to any tips others may have.

http://www.fastrawviewer.com/ or On1 Browse are good fast raw viewers on windows. I use them to 'manage' my library as I use a mix of raw processors and don't want to be tied to them for my workflow.

Took 5k pictures today so I'll try them out now. Thanks!

Reply on news.ycombinator.com