You have no obligation to us…I will look again and see if anything looks strange…I am not a programmer so as noted I’m not much help but I will poke around in case I stumble in to anything
1 Like
Interesting concept. Trying to get a handle on the above, Rec2020 is simply another RGB colorspace. What’s its purpose here?
1 Like
THX Christian … now I have a better understanding of the new module !!!
Your effort is much appreciated .
Compared to your " base curve " module , the spectral version is for sure easier to use !!
More gentle approach … IMHO.
Keep up your great work
Cheers Andreas
1 Like
Very interesting approach
however, If you want more people to test this, it might be worthwhile to create a pull request on the master branch on the darktable repository it self.
If all developers would follow your approach - create their own experimental of darktable - we end with 10+ version to test, that is imho very undesirable.
Personally, I will not download or install your ‘experimental’ version on my system. I hope that someday you will create a pull request to contribute to darktable.
3 Likes
Yes, I share the feeling and I wrote to Christian privately a few weeks ago about it.
There are some very nice ideas here, but I have no interest in using something that I won’t be able to incorporate in my daily workflow.
4 Likes
populating the darktable repository with unmature stuff as pull request isn’t an option - a pr should be something worth the maintainer and core developer will check and review.
First it needs a good answer to the question “why is this beneficial to darktable” - if not such prs are following the principle “flood the zone with shit” …
3 Likes
I am pretty sure that @martinus was not proposing that. Rather, expressing the desire that some of these ideas receive enough love and polish to make the leap from concept to feature.
3 Likes
that leap must be made before initiating a pull request.
A pull request is just initiating the last mile with the reviewers.
1 Like
Well sometimes it may also be good to do a first draft pr so we devs can get an idea where it might lead to. But for sure it must stick into current codebase.
2 Likes
Personally I don’t have a problem with a second install to test out things. And the way Christian B has set this up makes it very easy to handle both installs without conflict, at least on windows.
I also hope that some of his work will make it into a stable release (particularly Contrast & Texture). But after all, with AgX, we also had quite a long test version period which was, I think, beneficial.
5 Likes
I have installed one of Christians previous installs with no trouble and it allowed me to test out his new idea. This seems a reasonable first approach for developing a new module that may or may not prove worthy of inclusion in the master DT. It seems Christian gets enough testing support this way and others can wait for the PR request to see it included in DT nightly builds if the ideas prove worthy enough. I certainly welcome the work and effort Christian is putting in.
3 Likes
Indeed I was not proposing that. And if GitHub is configured correctly it also should not possible to just push to any branch in the official darktable repository.
I was more expressing that I hope that at some point an PR is pushed. And that @Christian-B is not working on an experimental version that will exist besides darktable.
I tried to express also some worry that if every developer will create his own version and ask us to download and to test, it get’s very messy, very quickly.
1 Like
Can echo that … for MacOS .
Running two DT versions completely separate from each other , no conflict at all … here on my system .
I do actually use the " experimental build " more than the stable one .
I will support his idea … as long as it is running as stable as nowadays .
Just my 2cents
2 Likes
Let me elaborate a bit. I appreciate the effort that Christian is putting into this and I am supportive. However, when there was only one feature on the radar, the attempt to make it into a mergeable PR eventually derailed, so in the end nothing was merged.
Now the features are piling up, I guess there is at least 3 modules and 1 UX change, so I wonder if and when these will move from the proof-of-concept stage to become something more. In the absence of a somewhat clear path to merging I don’t feel motivated to try these things out, as I am not very optimistic that they will eventually land.
But I would be very happy to be proven wrong 
1 Like
I also hope that you will be proven wrong 
, only Christian B will be able to tell us. Everybody has his/her own style of working and we have can only encourage him to go forward but I think that’s what we do.
2 Likes
If you don’t want to download and install the software, that is your own choice and you should do that.
It is starting to feel borderline coercive with “I won’t install this, work with upstream.” And that is a feeling I think we should avoid. Free Software is working as intended here, that should be celebrated. Christian obviously feels strongly enough that he went to the trouble of changing the about screen.
There are multitude of reasons why this might never get upstreamed.
5 Likes
Oh well, far from my intentions. Also, I have no coercion power whatsoever. Just expressing an opinion as a darktable user in a thread in the darktable category.
4 Likes
Yes sometimes our words get interpreted incorrectly. I am sorry this has happened to you on this occasion.
I too look forward to the inclusion of Christian’s modules when they are mature enough to make the PR standard required for DT. I remember how long AgX was in the making before it became a part of the official DT. It was a wait worth waiting for.
I am sure some of Christian’s ideas will prove valuable as well. His current pathway of testing seems suitable to me so as not to burden the DT team responsible for PR requests too early. I am particularly interest in in his contrast module (or whatever it is now called).
1 Like
Hello everyone,
While I was fixing the bug I was having with Windows on this project, I was following the discussion about my fork. Honestly, what I’m reading about my work style makes me chuckle a bit. We’re a long way from the spirit of “free (libre)” software advocated by Richard Stallman and “open source.”
I want to remind everyone that my goal is to learn and test freely in an “independent lab,” and that the code is available to everyone!
Furthermore, I have no problem if someone improves the code of a project and submits a PR, everyone is free to do so.
Regarding the introduction of certain projects, the first PR I’ll submit will be the addition to the color equalizer module; the code is ready and has no known bugs, and I’m planning to release it by the end of June (after the launch of 5.6).
For the basecurve module project, a discussion is underway on GitHub, and for the “contrast & texture” module, see this discussion where opinions differ on what should be included. In my fork, I’ve finalized it and also added the OpenCL code for the module and EIGF.
In any case, it’s the darktable team that will decide whether or not to incorporate a project.
I think I made myself quite clear in my communication. Now, there are several ways to respond to my fork, and everyone is free to react and/or use it as they see fit.
Personally, I’m disappointed by some of the reactions, as most of the posts don’t address the proposed code. But I’m pleased with other reactions that helped me debug by pointing out various issues in a constructive way. Regarding this module project, I’d especially like to thank Todd @priort , who took the time to compile it on Windows and managed to provide me with a log that helped me debug it.
Have a great day, everyone,
Christian
8 Likes
A couple things here:
-
I love what you are doing! Let me clear on that. I love how open source is working here. And really I have no issue with you creating your own experimental version. I actually admire that you have the time, and are experimenting. And there is nothing wrong with your style of working. It is very nice that you are sharing and asking us to give feedback.
-
But, and this has nothing to do with ‘borderline coercive’, I have the freedom to decline because it is not working for me. And of course, I understand that leads to disappointment. But you when you ask for feedback, you cannot expect that people just adopt to your style of working.
What is holding me back for of the darktable nightly or your test version, is that I don’t a true option of running it in a virtual machine (I don’t have proper gpu support). I want to absolute sure that my production darktable catalog is not accidentally touched by testing software. I the past I have run the nightly under a separate user account, but that is incredible annoying to me.
I don’t think that protecting the integrity of my personal data is against the ‘spirit of open source software’?
Richard Stallman, wrote an interesting article about communication in open source software development: https://www.gnu.org/philosophy/kind-communication.html. I hope that you recognise that by remarks are not a personal attack!
1 Like