There is one point where your critic, imm, may have some relevance as to what a beginner meets in darktable, and that is if they actively select the Module Groups header preset Workflow: beginner.
This module set-up do not fully use the core scene-referred modules like Color balance RGB, but for the Color group displays i.a. the Split toning, RGB levels and the RGB curve modules. These three modules are not anymore among the modules commonly used by dt users, and are not among those referred to in the processing intro part of the manual. (They are likely there because they may perhaps in themselves (their UI) be considered to be more intuitive.) This is confusing when one seek to pursue a modern scene-referred workflow that one is everywhere else persuaded to do. Also its module groupings/tabs use a somewhat deviating terminology. It has one less tab, though, and a less number of Corrective/Effect modules by leaving out some modules that are not essential, but imm mostly easy to anyhow understand the basic function of.
I’ve argued for the removal of this preset – and rather just stay with the default Workflow: Scene-referred.
But where do a beginner see that “overabundance” of modules if one stay within the default UI of Workflow: scene-referred?
As a totally new user of image processing I found this UI set-up to be quite orderly and informative and fairly easy to get an overview over its displayed modules.
(Previously one could unaware easily be flooded with a list of all modules by an extra click on a Module Group tab, but not any longer. It has recently been made a change here so one first need to make a right-click on a tab, which presents a list of relevant modules for removal/inclusion in the specific tab, and at the bottom of this list there is an option for All available modules.)
Except for the Diffuse or sharpen module that do have some function/presets that overlaps with functions in some of the other Effect modules, where specifically do you think a beginner meets an “overabundance”?
As for the “resistance”:
Yes, there are many ideals, (I guess it should also be so intuitive that a beginner can understand how to use all of its functions at first glance …) and many proposals for ideals, but in the real world they are often not possible to achieve.
There are f. ex many clever heads that has put much effort into creating as good a tone mapper as possible. But as varied as images, users and users’ needs are, it turns out that they all have their strengths and weaknesses and one cannot replace all the others. Sigmoid was f. ex. developed because Filimic RGB was by many considered too complex and with unsatisfactorily results for some types of images, but on the other hand Sigmoid then is lacking in features needed for other groups of images
I agree, with emphasis on “talking”, because I’ve learned that in the FOSS-world it is doing that counts in the end – and I think any perception of “resistance” should be considered in this light.
) . In my former day-job such was critically important, as failure investigations depended quite heavily on knowing the pedigree of data reduction back in the early days of design, prototyping, and development. So, keep that in mind when you go mining your repositories’ early years…