Nature documentation narrator voice: The main admin shook his head. “We are not going to do that.”
1 Like
These are two independent conversations. I wanted to explore an alternative path to having some more theming options available for the average user, as devs are not willing to have new themes added to the repo.
If there is an interest in making it part of the official DT offering, then we can spend some time refining it.
1 Like
Pixls is pretty robust and will definitely still be here in a year. Maybe post stuff here as well, just in case.
1 Like
True, but it’s also more sparse and automating things as @Pascal_Obry was suggesting would be more complicated.
I’d imagine most people who can get Lua stuff working are pretty comfortable with copying-and-pasting some small bits of CSS.
For me what I think could be useful is having something like your website and having options posted there, but in addition you have a dropdown that exposes the css. People could tweak and experiment and your theme thumbs would update to reflect the edits…so someone could find a theme that was close and then just experiment to add or change something a little. Then you just have a copy button like we see with code or commandline snippets and you copy it when you are happy and just paste it into DT. So no fancy connection… but still very useful…
1 Like
Would it?
You can get https://discuss.pixls.us/t/darktable-css-tweaks/53649.json and do some basic parsing and come out with the css and the screenshots of it too.
I think the main point is that we do not have a lot of stuff on our infra because we do not have the person-power to run it. PIXLS (which serves more than just the darktable community) has approximately 1.7 admins, and I count @darix as 1.5 persons worth of admin because he is amazing, while @patdavid and myself are approximately 0.1 persons worth of admin.
I appreciate that you’re trying to push things forward, but we grow in a very controlled manner because what we have is already quite fragile.
The attitude of “written disclaimers” and “its just downloading random CSS from the internet” is not great. You’ve made something and you’ve put it out there and you’re asking people to use it, but you don’t want any of the responsibility of it.
3 Likes
Writing a parser that is robust to different people writing posts in different ways and without using a rigid schema is going to be a fun undertaking 
1 Like
I looked at this a long time ago, specifically populating the user.css and then applying it. I think I have to add a hook to the API to do the apply.
EDIT: Figured out what to hook
That is a bit of misrepresentation of what I did and of my intentions.
I said that it’s a proof of concept and that I would like to people to help test it. I also said that if there is interest in establishing something like this I am happy to work on improving it (I used “we” because I see it as a community project, even if no one else is going to actually work on it).
So let me be a bit more clear about the whole thing:
- The idea has been floating around for a while. In the past I have been criticized on this forum for proposing ideas and not following up. Now I picked someone else’s idea and I went ahead and did it (which, apparently, is still not good enough).
- This is a proof of concept, which I made to gauge the community’s interest around something more concrete than a vague proposal. If the interest is there, I am happy to work on it, improve it and maintain it.
- If the DT team wants something like this to become more officially part of DT ecosystem I am happy to transfer hosting and ownership.
- I am happy to be in charge of its maintenance, regardless of where it is hosted.
As for the security concerns:
- There is no security risk that I can think of for end users, in the worst case they may download an invalid or spammy CSS. No big deal, they can delete it and move on with their lives.
- That said, I can easily implement a quarantine zone and only publish a CSS after it’s been manually inspected. We are not talking about instagram here. We are not expecting 1M posts per hour. On average I would be happy to get one submission per week. And even if it were 10 submissions per day I would be happy to take care of vetting them individually.
- On the server side, the input is already sanitized, and of course it can be sanitized even more (see point (1) above).
4 Likes
That would be nice to have, however it would be a different project: a theme editor vs. a theme gallery, which is what this thread is about.
BTW, now you can see the highlighted CSS in the theme/snippet view, and you can copy the CSS to the clipboard as an alternative to downloading the CSS:
6 Likes
I just wanna thank you @Masterpiga
I’ve really wanted contrasting module headers and color-indication of off/on status for a long time. Now i have it!
I tried once myself, but getting the class-names was too cumbersome.
Thanks man!
2 Likes
So, the website has been online for 10 days. It has received quite some visits and the assets have been downloaded, but there have not been new contributions, and the only hosted themes/snippets are the ones that I uploaded.
I am a bit surprised, as the idea to have a repository of themes seemed to be well received, more than one person had suggested it as a good alternative to adding new themes to the release, and users seemed to be happy to have a central repository of themes.
I don’t know why people didn’t contribute. Maybe users perceived the effort as a way to showcase my contributions, even though I think that it was clear that it was not the intention. Maybe it’s just the usual user inertia, and users just prefer to keep using the forum to share themes and snippets as they did before. Maybe they didn’t want to share their email address during the registration, or they didn’t trust the usage that I would make of their data. Maybe it’s about ego, and theme authors just prefer to host their contributions on their own websites instead of using a shared resource. Maybe it’s a combination of everything above, or maybe it’s something else altogether.
Anyways, the lack of contributions suggest that it does not make sense to spend more effort on this project. Since I don’t like clutter, abandonware and unfinished things, I am going to turn down darktable-css in the next couple of days.
Of course, happy to change my mind if conditions change.
Thanks, cheers
1 Like
I do from time to time try out other people’s css tweaks, but I literally have nothing of my own to share. Perhaps most people are much the same as me, just using what is there. People may also have been reluctant to upload anything in case your site proved to be short-lived, but there is really no way to tell.
1 Like
I would gladly contribute, but my free time for FOSS tinkering is limited. If you keep it up for a while, I’m sure I’d contribute eventually.
It’s not just laziness, either. But publishing something requires a bit more polish than the little hacks I use for my own purposes. We’ve only had ten days to polish things up. I’d give it a bit more time.
I looked at the site and the first attempt showed a mess. I suspect that this might be caused by the use of JavaScript on the site (I haven’t checked this further). JavaScript is disabled by default in my browser; scripts are only enabled on trusted sites one by one.
More importantly, logging in to a site hosted by a private individual I don’t know with personal data is unacceptable for me.
I think this is a really good idea - a central repository. If I see something interesting when scrolling through pixls late at night on my phone then I will copy the link to Google Keep notes for exploring at a later date on my pc. Obviously I will miss things so having a central place to contain snippets of code that as far as I can tell will not break your darktable is to be recommended.
This along with planet darktable and libre arts are really good developments and to be applauded.
Some time ago on this forum, and sorry I forget who did it for me, gave me a short code to darken the active module and it was a real improvement.
1 Like
Currently, there’s a fundamental problem. The various topics related to darktable are widely scattered and sometimes difficult to find.
As long as there’s no effort to unite “everything” under one roof, it will remain difficult.
A few examples:
-
Darktable.org
Homepage photo is version 5.1 and the tooltip is 5.0
Contact page → Facebook, last entry 2021 - Summer release DT 3.6 -
Darktable Github
Photo version 3.1
Wiki last modified Mar 13, 2022 -
darktable user manual not (entirely) up-to-date
-
darktable gitlab
Last updated 5 years ago.
“darktable 3.4” in the darktable usermanual - version 3.4
These are just a few examples to illustrate the issue.
What’s missing is a central hub for all the information available about darktable, creating a single home. Or at least as few different locations as possible. These locations must still be accessible from a central site.
Old information simply needs to be deleted.
Darktable is fantastic software, but it’s (partially) difficult to use, and I’m not talking about the UI 
To be clear, this is not criticism but a description of the current situation.
3 Likes
This is all true, but probably it’s not very related. Hundreds of people have read the OT and seen the website, including those who over the last weeks have been sharing their CSS on this forum, yet they have decided not to contribute to darktable-css.
The connection is that support would be significantly higher if it weren’t another “isolated” location.
I think many people are aware of the problem. This doesn’t just affect you, but many others as well.
Currently, there’s no solution.
It’s not easy when you have a good idea and want to contribute something to the community.
But it’s also not easy for the developers, who, as it currently stands, are at the limit of their capabilities. It doesn’t matter where this situation stems from.
1 Like