Improvements

[kusti420]

is it possible to somehow detect when youre hovering over a hyperlink or anything that you can click? cause i at least feel that sth like that should be added so that autoscroll wouldnt get enabled when youre middle clicking on links to open em in a new tab.
right now you have to press once to open sth in a new tab, then press again to disable autoscroll. when it shouldnt get enabled in that scenario in the first place.
thats at least how it works on windows and i hope its possible to make your linux-autoscroll work like that as well

[kusti420]

also is it possible to add a feature to be able to autoscroll in any direction? not only up/down but also left/right?

[TWolczanski]

is it possible to somehow detect when youre hovering over a hyperlink or anything that you can click? cause i at least feel that sth like that should be added so that autoscroll wouldnt get enabled when youre middle clicking on links to open em in a new tab. right now you have to press once to open sth in a new tab, then press again to disable autoscroll. when it shouldnt get enabled in that scenario in the first place. thats at least how it works on windows and i hope its possible to make your linux-autoscroll work like that as well

I've been thinking about adding some kind of delay between clicking the middle mouse button and entering the scroll mode: in order to enter the scroll mode you would have to hold the button for e.g. 300 ms. This would allow you to avoid unintentional autoscroll activations, in particular when hovering over a link - if you wanted to open it in a new tab, you would just click the middle mouse button, and if you wanted to enter the scroll mode, you would hold the button until the autoscroll icon appears on the screen. However, it turns out that this feature is not that trivial to implement - I haven't yet figured out how to make it work in every scenario.
By the way, you can modify the script to use a different mouse button for entering the scroll mode by changing the BUTTON_START and BUTTON_STOP constants. As for me, I have these constants mapped to one of my side buttons. This way I can enter the scroll mode everywhere, without any conflicts.

also is it possible to add a feature to be able to autoscroll in any direction? not only up/down but also left/right?

Yes, it's possible, but I can't imagine a situation in which you would find it practical to autoscroll horizontally.

[kusti420]

ill look into it as well. i was thinking that since every os has sth built in to know when to change the cursor then there should be a way to read that info as well. then its just a matter of adding an if condition

[kusti420]

but so far i havent found anything about how linux knows when to show which cursor

[TWolczanski]

i was thinking that since every os has sth built in to know when to change the cursor then there should be a way to read that info as well.

Even if there was a way to read the cursor info, it wouldn't be enough. Imagine a situation in which the cursor doesn't change, but the thing you're hovering over is clickable.

[kusti420]

but why should it not change? cant think of a situation where sth would be clickable and it wouldnt

[kusti420]

even if it wasnt enough (idk how it wouldnt), it would still be significantly better

[kusti420]

2022-06-10.17-28-10.mp4

was able to get something working (somewhat) regarding detecting when cursor is hand pointer. need to figure out how to fix that "maximum number of clients reached, segmentation fault" problem, then implement that logic into your code and it should be done

[kusti420]

idk.zip

im still kinda stuck with solving that "maximum number of clients reached, segmentation fault", maybe it could be solved if someone else takes a look at it as well. look at test.py file.

[kusti420]

legit clueless at this point
linux-autoscroll.zip
idk. is anyone able to fix this?

this issue can be closed as soon as someone is able to fix that "Maximum number of clients reached\nSegmentation fault (core dumped)" problem

[TWolczanski]

I'll take a look at your code in a few days when I'll be working on the feature I mentioned in my first comment.

[TWolczanski]

I've managed to add a few new functionalities to the script, maybe some of them will resolve your issue.

Unfortunately, I don't have time to look thoroughly at your code and at the PyXCursor code, but my initial suspicion is that creating a new Xcursor instance every time the sense function gets called is the cause of this segfault.

[guigirl42]

FYI:
#3 (comment)