It's a separate project of mine, actually

https://asvd.github.io/intence/

it works... somehow :-)

http://asvd.github.io/microlight/haskel.png

well, since the lib is general, it's built upon compromises. But I am open for suggestions concerning updating the logic for some particular cases

Does it also work for languages where you can define // to be an operator?

That's a different thing. The menu on the Intence page is implemented using viewport.js: http://asvd.github.io/viewport/

The Intence itself is the one which highlights top and bottom edges of the area in case if there is something to scroll.

Have you faced any complications in understanding how navigation is performed on the page? (if we imagine that the content is not actually related to the subject and does not explain how the indication works)

Complications? My belief is that good UI is there, but not noticed. You shouldn't have to ask how to interface with a website.

A week ago this project has been published on the news.layervault.com. The UX experts have already submitted tons of negative feedback on the concept:

https://news.layervault.com/stories/46773-intence---the-scro...

Now what do hackers think? Should scrolling indicator necessarily provide the way to control the scrolling position, or can these issues be solved by different widgets? Is this way of indication far too distracting, or does it probably depend on the designer's choice of a texture for the indicator? Is this really a small UX revolution, or just another useless experiment which will soon fall into oblivion?

This is defined by the developer actually. And besides there is no problem for a user if he does not know that he can scroll with dragging.

then you will have to manually reset the mouse on the table :-D

Sorry, I think I don't clearly get what you mean. Tablets normally scroll smoothly pixel-wise as long as you move the finger. In fact, dragscroll implements the similar behaviour for desktops, because I consider such behaviour as more reasonable in many cases (comparing to selecting the content of the area, I mean).

You are right about how tablets are supposed to work.

The problem is that in your examples this behavior no longer functions as expected on tablets.

The example should work well for both desktop and tablet. Your current implementation changes the normal behavior on tablets.

There probaby an bug in dragscroll then. It is not supposed to break the behaviour on tablets, will test it..