I should really have put a suitable emoji on my post.
The HURD really is a dead parrot, I think the talk page on its Wikipedia page sums it up.
I should really have put a suitable emoji on my post.
The HURD really is a dead parrot, I think the talk page on its Wikipedia page sums it up.
Yes, I see now that you were joking about the HURD. Sorry. I’m sometimes surprised how long it takes to get agx done, with new ideas or problems popping up all the time (and sometimes I’m my own worst enemy, let’s just quickly try this one more thing).
On the HURD’s Wikipedia page, I note that it supports the ext2fs file system. I don’t know what you are running, but my Linux system uses BTRFS on the root partition, and XFS for user partitions.
While Richard Stallman was looking for a kernel for the GNU project before Linux was produced, it is arguable that his continuing antipathy to Linux (or GNU/Linux as he calls it) didn’t do the project any favours.
Stallman has form on this, see also the Emacs/XEmacs split, and the Gnome/GTK, KDE/QT conflict.
No, my fault. What I forget is that this is an international forum, and however the excellence of poster’s English, they may not realise that posts are not meant to be taken seriously.
Yes, it takes a long time - and i’m sure it took more time than you expected.
But that is a good thing, i think. Darktable doesn’t need a repetition of the filmic development, with 9 different published versions (3 in filmic, 6 in filmicrgb), and 9 different user interfaces.
Well yes, everything takes longer than you expect
My days of involvement in software development at any level are long behind me. Aurelian obviously knew his stuff, but his focus seemed to be on his own needs, and he didn’t seem to take the use cases of others into consideration. Hence the need for continual changes.
ext4 and zfs here
GNU Hurd was always meant to be perfect, addressing shortcomings of UNIX, while giving freedom to the user, but the users were much more interested in installing and running servers and desktop computers than they were in installing another file system driver (“service”) on an already running big computer system (as opposed to a PC).
Reality outran the imagination of the FSF so fast that the FSF didn’t even notice. The FSF may already have lived in the past at the time the HURD project was started.
It’s a nice discussion, but I think we ought to stop it.
It is a distraction from the primary purpose of the thread.
To give Aurélien credit: he developed that stuff pretty much himself (sure, he talked to people, but those were his ideas, and when he switched from filmic v5 to v6 and then to v7, he did that based on user feedback, mainly). I’m, on the other hand, only porting what others have developed. My only contribution was exposing the controls.
@kofa this is the most positive message of the day! So happy to hear that AgX is reaching that level of maturity. I developed a hundred images with different POC versions and I expect it to bring a new era of darktable - enabling beginners and pros likewise to get good results quicker. I achieved fast results especially with tricky raw files that required a lot of steps before AgX to get acceptable output.
While I agree that this is a very positive message and that many people will get good results quicker than before. I doubt, that this counts for beginners.
To get results quicker, you have to have an imagination how your photo should look at the end. Many beginners don’t have that. So they will probably not profit from the control AgX gives and maybe would be quicker with Sigmoid.
What kofa will bring to darktable is a straight forward Masterpiece of control for all kind of light situations. It is NOT a guarantee for getting the best out of every picture. That is still up to the user, who will get a new very powerful tool. But that’s still just one tool among other very good tools, which you will still need.
I really like Agx, but it is another level of complexity to master on top of darktable’s significant complexity compared to commercial alternatives. I think for many who understand shoulder’s and toes, primaries, and attenuation, AgX is fine. But for others who learned how to expose photos with highlights, shadows, and midtone sliders this new UI language (which doesn’t conform to many other modules) will be a challenge for beginners.
Maybe that’s not the intended audience. Maybe everyone is happy with the freedom and innovation that generates cool new modules like AgX.
I can see the value in the innovation and some increased UI design coherence.
I’ve thought about changing the labels, such as ‘shoulder power’ to ‘contrast in highlights’. Alternatively, there are the tooltips. There will be documentation. There will be videos (Boris, for one, has offered to create one). We’re just waiting for the dust to settle.
This will be a great pleasure!
I will need your expertise again. Like the nice explanation about “Notorious 6”. I have to include the basic ideas behind AgX, because that is very important for understanding the use of the module.
But we’ll talk about that later.
It isn’t thaaaat complex. Yes it offers a lot of possibilities, but you don’t have to use them all. What helps a lot is the curve, which illustrates what you are doing - this helped me a lot.
I wouldn’t change the naming of the sliders- All sliders are referring to the curve and “raising contrast in the highlights” is not always possible to that extend what some users would expect when the slider is named like that (depending on the picture).
Yes it might be challenging for beginners. But that counts for darktable anyway. And by the way for many other advanced RAW-converter.
A good documentation, videos (thanx to Boris) and tool tips should be enough. I just wanted to point out that this module is great but not a “egg laying woolmilksow” like we Germans wood say.
Here is what I gathered about Notorious 6
In image processing, the term “notorious 6” typically refers to six common types of image distortions or degradations that frequently challenge image quality and algorithm performance. This term is sometimes used in the context of evaluating image processing algorithms—especially in image denoising, super-resolution, or compression artifact removal.
While not a formal standard, the “notorious 6” generally includes:
These distortions are “notorious” because they significantly affect both human visual perception and computer vision algorithm performance, and require specialized processing techniques to mitigate.
Are you sure darktable is that much more complex than commercial alternatives? Different, yes, that surely.
Also, it’s not “another level of complexity” on top of what exists, it is an alternative to some of the existing modules, and might in some cases lower the need for extra modules.
Don’t forget that you can safely ignore the AgX module, as long as it’s not the default. I don’t see that happening for at least 2 or 3 versions after introduction.
And as far as I know, there’s no planned date for inclusion yet.
In time, there may be user guides that make it easier to use. Or it may not even be needed for many use cases.
But it is there for when it is needed, and for those who do want to use it (cf. the “Diffuse or Sharpen” module: very complex to use from first principles, very usable with the available presets).
And don’t confuse “beginners” with “users from other programs”.
For users coming from “other programs”, darktable is already challenging in a number of aspects. One issue is that those users are not beginners, but have already a well-developed knowledge of editing with that other program. Real beginners may actually have an advantage, as they don’t have to unlearn anything.
Here is an alternate definition of “Notorious 6”. This is probably closer home
In the context of image editing, the term “notorious 6” is not a formally recognized or widely standardized term in academic or professional literature. However, when mentioned informally, it usually refers to six commonly encountered and often problematic editing mistakes or challenges that degrade image quality or realism.
Here’s a plausible interpretation of the “Notorious 6” in image editing:
These six issues are considered “notorious” because they’re common errors made by beginners or even experienced editors under pressure, and they’re easy for viewers to spot—often breaking the realism or aesthetic of an otherwise good image.
Thanks for the explanation, but I’m referring more to what @kofa has nicely demonstrated in this post regarding color shifts:
I am not a professional, more of an amateur who likes to improve home/holiday photos. I have been experimenting with AgX, from almost the beginning of its creation. Here is my take:
I find AgX to be a great module that is able to handle the functions of several modules, Sigmoid, CB-RGB, TE to name a few.
Earlier I had to move back and forth between the modules to come to a state that I like. Now with AgX, all I need to do is play around with the sliders and the tabs, certainly a more convenient approach.
I, Like you didn’t understand toes and shoulders, but playing with the sliders once or twice, gave me the understanding what the slider does and when to use it So I am not complaining of the terminology
All in all, it is a great module and is capable of producing excellent edits. I am sold on it and cant wait for it to come into the master of nightly builds. Today I have two versions to work (AgX and Non AgX), which is a pain in the wrong place