Add proper smooth scrolling by default.

[Pale as the Moon]

Hi there,

currently I have both Waterfox and Pale Moon running on the same Windows machine(s). One thing I have noticed is that Waterfox does scroll far smoother by default. It is more pleasant to look at for the eye and no micro-stuttering can be seen. Pale Moon appears to micro-stutter in comparison. I have noticed this on both new and old profiles and on two different machines, using the same websites - so it doesn't seem to be a coincidence. I ask you to compare the scrolling behavior of the two browsers and have a look yourself. Would be glad if Waterfox' scrolling behavior could be introduced in Pale Moon as well, as it is really a good experience.

Thanks for your attention, hope this is given some consideration.

[adisib]

Waterfox uses Asynchronous Panning and Zooming. This requires e10s, so is unlikely to make it into Pale Moon. Pale Moon would have to address this issue differently.

[Moonchild]

Waterfox uses Asynchronous Panning and Zooming.

Waterfox uses whatever Firefox comes up with, since it's a re-build. Anything like this that is findamentally deep in the rendering engine will not have been touched by the Waterfox developer who hasn't touched its C++ for a long time now.

This requires e10s, so is unlikely to make it into Pale Moon.

APZ has nothing to do with e10s. I think you have your technologies mixed up XD. I'm also not sure if this perceived smoothness actually has anything to do with APZ, since that would not necessarily address "micro-stutters".
APZ is, however, not desired for our current implementation of Goanna. It simply won't work well.

Pale Moon would have to address this issue differently.

And I have played with some stuff to improve this already, however I'm not yet sure if it's ready to be enabled by default. Probably is, but will need testing. I'll flip it on in the unstable channel and people can have a go at it.

[Pale as the Moon]

Good to hear something is done about that. Yeah, I had no better word than "micro-stutters" at hand for this behavior, since it does not really "hang". It's more like multiple small interruptions when scrolling which do add up to something visible, whereas Waterfox' scrolling behavior appears to be fluid.
It feels like Pale Moon was the beta and Waterfox was the final, so to speak.

When the improvements @Moonchild mentioned roll out (I assume with Pale Moon 27.4) I will intensively test this against Waterfox and will give you further feedback.

[back2themoon]

When the improvements @Moonchild mentioned roll out (I assume with Pale Moon 27.4) I will intensively test this against Waterfox and will give you further feedback.

The unstable channel was mentioned, so you can probably already check it out there. Make sure to read instructions and backup your profile.

[adisib]

Waterfox uses whatever Firefox comes up with, since it's a re-build. Anything like this that is findamentally deep in the rendering engine will not have been touched by the Waterfox developer who hasn't touched its C++ for a long time now.

I never said it was Waterfox's own solution.

APZ has nothing to do with e10s. I think you have your technologies mixed up XD. I'm also not sure if this perceived smoothness actually has anything to do with APZ, since that would not necessarily address "micro-stutters".
APZ is, however, not desired for our current implementation of Goanna. It simply won't work well.

I know they aren't the same thing. The fact is, however, that Firefox will not use APZ unless e10s is enabled. Flip off e10s, you are no longer are using APZ. Presumably, this would mean that APZ is tied to e10s in some way, unless you are suggesting they are arbitrarily not allowing it just so that Firefox will feel smoother with e10s to make people more accepting of it or some such idea?

And I have played with some stuff to improve this already, however I'm not yet sure if it's ready to be enabled by default. Probably is, but will need testing. I'll flip it on in the unstable channel and people can have a go at it.

Glad to hear!

[Moonchild]

I know they aren't the same thing. The fact is, however, that Firefox will not use APZ unless e10s is enabled.

That's really not a necessary requirement. APZ is NOT dependent on e10s. Why Mozilla links those two, only Mozilla can answer.

[LimboSlam]

Has this improved feature been in implemented, or enabled? If disabled, then what is the preference name?

[adisib]

Has this improved feature been in implemented, or enabled? If disabled, then what is the preference name?

If you are referring to "And I have played with some stuff to improve this already, however I'm not yet sure if it's ready to be enabled by default. Probably is, but will need testing. I'll flip it on in the unstable channel and people can have a go at it.", I believe these are the prefs related to the mentioned unstable channel change that was added to 27.4:
gfx.vsync.compositor;true
gfx.vsync.refreshdriver;true

[LimboSlam]

If you are referring to "And I have played with some stuff to improve this already, however I'm not yet sure if it's ready to be enabled by default. Probably is, but will need testing. I'll flip it on in the unstable channel and people can have a go at it."

Yes, I am. I have found the answer in the release notes. See below:

Enabled v-sync aligned refresh for a smoother scrolling experience.