Greasy Fork

Regarding Zoom Factor

I understand that width altering makes impossible to return to marked position because the dependency from width is not linear. But I think, that zoom altering can be recalculated if keeping zoom factor together with position. Or even more tricky: to keep the position recalculated to zoom 100% and recalculate it for other zooms. The dependency from zoom should be close to linear.

As far as I noticed, zooming does not change webpage pixel Y-position, bookmarks remain valid after pressing Ctrl-plus/minus, so I don't see any reason to correct anything yet.

Zooming does not change webpage pixel Y-position, but it changes the actual page width. (Real geometrical page width it might or might not change, but the pixel size becomes bigger/smaller and the width in pixels changes). But I agree that it is nothing to correct in your script - if zooming causes width altering it is nothing to do - if the geometrical size will be changed - the zoom altering will not affect the page position, if will not change - will affect. But it depends on site behavior and page layout and it is impossible to take this into account.

But it depends on site behavior and page layout and it is impossible to take this into account.

in fact, everything is possible - but is it worth the effort?

I agree. The user can take this into account, but it doesn't worth.