How hard would setting up a remote monitored test: the user and the monitor?
It can be effective with very few participants - the Nielsen Norman Group does mostly web site testing, but some of their techniques could work on software. If you have 5 users and they all fail on the same spot, you don’t need another 45 users and a fancy graph … you have a problem.
A website is still an area of little content and some menus, with a form of use that is the same for the whole population, are not comparable to a program. To believe that with 5 people you will obtain reliable data because, for example, 3 of the 5 complain about something is to ignore the law of small numbers.
That’s not really pointing to any proper design standards for UI/UX. Let’s be clear, everything is called design. UX is about focusing on the user experience. No matter how good your UI designers are, if they don’t focus on the user experience, it’s just opinion. And when it’s opinion it’s all down to how good that team and person is. In my experience having a UX focus can give a broader look at the UI, when you can bring actual data to a conversation it stops being about “I feel this button is easier to use”.
I also think if they have written this down in details, it’s odd that it’s not public.
Ton is right, because we too often assume other software or defacto standards are what they are because they are thoroughly researched and well made. But the fact is that a lot of products from big companies are lucky coincidents. And you can get a lot of bad code and sloppy work from them too.
Taking a step back and rethinking things is important. That being said, often you’ll find out the way other people do it is because that is intuitive. For different reasons. So while it’s important to do, it’s also dangerous to refuse to do something in a similar way just because. (Can’t speak for what he meant though 
)
I don’t think the meaning of Ton’s words is to refuse doing something similar, just to reach that solution in an independent manner, if the solution achieved is the same one… it probably is the most optimal one 
So … reininventing the wheel basically?
Blender, houdini, maya, zbrush, marvelous, modo,… all of them created their own wheel without paying attention to anyone.
Not exactly, a wheel is something simple, like a node, however the nodes in Houdini, Bifrost, ICE, Unreal, they all work differently, you may get inspiration based on your knowledge, but it’s better to avoid specific and detailed examples to try to avoid direct contamination, and better to propose a solution to a problem under the umbrella of the software designed to solve it, neither Houdini, Maya, Softimage amor Unreal function in the same way of Blender or have the same functionality and target, so their solutions, while may have some interesting points, may not be the most optimal solutions for Blender, so reaching an “own” solution is far better than using an existing one, I think that’s the idea behind “no use other softwares” to “solve problems”
Sadly, one sometimes has to reinvent the wheel because otherwise some other company with a patent or something similar will force you to remove your wheels that seem copied from their wheel design.
Also true… sadly…
So, I think the conclusion we can draw by this thread, and other developer responses I’ve seen on this forum, is that while Blender has grown and received a lot of money lately, they haven’t allocated any of that to documenting user feedback. In fact, while they provide a number of places for users to leave feedback, they may or may not read it. What they still want most is community contributions, that they may or may not include.
Did we get any concrete response on the process of implementing Ux in Blender in this thread, apart from that there currently doesn’t seem to be one?
(Also, if the people with more than 15+ posts already in this thread could possibly restrain themselves from replying to this, perhaps this thread would be taken more seriously…)
If you don’t come up with assumptions like this, people with more than 15 posts here could have restrained themeselves, but if the information is wrong or the assumption is incorrect anyone would chime in to say it, no matter if that person has 1 or 15 posts.
They read this forum, and there is also a lot of user feedback in developer.blender.org, in the different design tasks and even when a bug is reported.
I’m not sure how can you extract that conclusion from this thread.
Was there a concrete question?
Even you in this answer haven’t made a question, but once again an assumption / conclusion.
What I see in this thread are continuous assumptions about how UI/UX is being developed in Blender, or about how it is not being developed 
, but not an actual question about it.
Going with the assumption that there is no work being done in UI/UX or that there is no work made by the UI team regarding UX is not a good idea I think, would you like to enter in a diffuse discussion with someone that denies your daily efforts?
Ask a concrete question, mention the proper developer, like the UI module owner, and then you may receive an answer.
I think a customizable UI might solve a lot of UX problems because everyone has unique preferences.
This is the year 2030.
Blender now has an interface with Open AI with GPT - 7
it is an AI that perfectly understands the users, they speak and the ai executes orders, it creates whole scenarios just “describing them to it”. The classic UI still exists, but it is now almost completely hidden, also because blender is now used almost exclusively in VR and AR.
Unfortunately I cannot give you proof from 2030, but here is a primordial example of what will be (GPT-3).
Well, here is a problem. A Production.
And massive production deadlines.
User in not a “constant” - it is, actually, a “variable”.
User that start learning software and the same user several years ago, surviving through neat production deadlines are two different users with different goals and demands.
There is a huge difference between learning to extrude and doing billions of extrusions.
Regular car is all about comfort - it is full of consistency solutions, which are easy to learn, but are not effective enough to perform professional race.
Race car is all about efficiency - it is based on relevancy solutions, which are hard to learn, but allow you to do massive extreme work faster.
A sustainable business is a race.
In UI/UX (workflow) design this dilemma is called a “consistency vs relevancy and their balance” issue.
There will be no wide-range effective professional tools and methods that will be easy to learn, just because of their nature. Sorry.
Blender IS a regular car, although with a non industry standard dashboard that is hard to learn…
However, because it is a regular car, people can mod it and apply their own customization and kits to it. You’d be amazed to see what performance people can squeeze out of regular cars. A lot more people use regular cars than racing cars.
Those racing cars you mention, they’re built by companies like Autodesk who specialize in racing. Most people just want to drive, and nobody starts out by racing.
… also: A 3D Software is not a car. 
Nobody says that consistency and intuitive usability should come at the expense of speed. I am sure the Blender developers don’t want that, either. But getting to the ideal ballance of pleasant and easy use without the loss of speed takes time as well as trial and error. Changes will have to be dialed in and maybe partially rolled back as well.
Ultimately ease of use in the sense that you don’t have to constantly think about what to do (like for example the viewport setup for retopo which takes steps in several different places across viewport and object settings) also can mean a lot of speed improvement in the end.
The point of this topic all along was that these things should be taken more into an overarching and program wide consistent consideration manifest that we can see and talk about.
Not just as a modular thing that is attached to each individual module.
I think the real question is, what if Blender was an insect ? would it be more like a cicada, or more like a termite ?
As I mentioned, there are valid arguments regarding the fact that “easy to learn” (as a core focus and as a core philosophy) can mean significantly reduced power for those who want to go beyond putting together basic scenes, but I don’t think the car analogy fits very well.
The reason why that is so is due to the fact that while a race car has a lot more horsepower, it is built primarily for speed sports on a track. It would not be your first choice for things like errands (like shopping ect…) or even just driving in the real world where traffic laws need to be followed (because of the regulations regarding safety and because every part needs to last a long time over thousands of miles).
The Race Car excels at what it does because it is specialized for a specific task, Blender covers a lot of areas and a lot of use cases so it can’t specialize too much without alienating thousands of users. In the big picture of things I wouldn’t exactly call Maya and Max a race car either, though maybe Zbrush counts as one since its core use case is sculpting.
If it were a regular car, it would not have won Autodesk’s competitions several years in a row.
We use Blender for uniquely complex modeling projects where all other software fails.
May be you just using it wrong, but no worries, a lot of people use software just for fun.
Most people want to make money. If you want to make money in computer graphics, you have to be better than others, and… welcome to the business race. That’s how it works.
This is horizontal shift - between specializations (more ARTwork - less CADwork). There is also vertical shift - between professionals and new users (more consistency for new users - less relevancy for professionals).
Also, I’ve never compared Blender to a car. I compared the two cars to clarify one of the deepest UI/UX design problems. If something is hard to learn sometimes that means that you have to be smart enough to learn this, because further it will be even more hard.
Learning Blender is not that hard compared to making money in computer graphics.