News for hackers. Stuff that matters.
Re:Libinput's bus factor is 1 (2019) by wang_li on Friday May 22, 2020 @02:41AM (#23269286)
Go read Microsoft's mouse ballistics from way back, if you can find it, then use a mouse on any Windows desktop since XP for a week. Then come back and try a Linux or Mac with an actual mouse instead of a touchpad. You can fix that.

It's here: https://web.archive.org/web/20101222190833/http://www.micros...

Re:Libinput's bus factor is 1 (2019) by teekert on Friday May 22, 2020 @03:32AM (#23269551)
I don't know from this comment on what side of the fence you are but plugging a mouse into a mac has always felt horrible to me, like I was mousing through mud, I would always feel like the pointer slowed down too early and then I would have to reach a long way to actually get to the target. May just be my muscle memory was attuned to Windows/Linux.
Re:Libinput's bus factor is 1 (2019) by ragona on Friday May 22, 2020 @11:56AM (#23273427)
I absolutely agree, but it’s a personal preference thing.

Years ago I had this argument with a coworker, and I tried to take the position that the Windows curve is better for games. He politely disagreed, I implied that he was a casual gamer, and he gently told me of the year his Quake III team won the world championship. We spent the next while watching demo videos of him doing 360 noscopes with the rail gun against a target in the air.

I handily lost that disagreement.

Re:Libinput's bus factor is 1 (2019) by robin_reala on Friday May 22, 2020 @04:17AM (#23269775)
When I used a mouse on macOS, the first thing I did was install the Microsoft mouse drivers,[1] which had Windows acceleration curves built in as an option. Unfortunately they stopped updating them beyond 10.7 Lion.

Now I’ve just gone over to Apple’s trackpads for 100% of my cursor needs.

[1] http://download.microsoft.com/download/B/1/0/B109F931-70E2-4...

Re:Libinput's bus factor is 1 (2019) by teekert on Friday May 22, 2020 @09:46AM (#23271994)
Yeah, I did the same, I always wondered if that was by design, like: "You don't need an ugly mouse next to your macbook, we'll make sure of that". But the touchpad was also indeed that good that I didn't feel I needed a mouse. Opposite on Linux/Windows where I love my external mouse. I'm a bit on the fence about what I prefer, Linux with mouse or mac with touchpad. Both (as in MacBook touchpad on my Linux laptop) would be nice...
Re:Libinput's bus factor is 1 (2019) by toast0 on Friday May 22, 2020 @11:20AM (#23273051)
Mouse acceleration has always been different on Macs, just like font rendering, keyboard shortcuts and menu bar position.

Apple is very opinionated about all the parts of User Interfaces (including inputs), it's great if you agree with their opinions.

Re:Libinput's bus factor is 1 (2019) by olq on Friday May 22, 2020 @11:13AM (#23272963)
I use Steelseries ExactMouse to get rid of that odd acceleration curve. It works for any mouse brand and it still works on Mac OS 10.13, maybe newer OS'es too.

https://downloads.steelseriescdn.com/drivers/tools/steelseri...

Re:Libinput's bus factor is 1 (2019) by Gravityloss on Friday May 22, 2020 @07:27AM (#23270905)
The mouse on Apple has been horrible since at least 2005 or so (and I haven't used older macs than that). Low update rate, laggy, imprecise. Windows and Linux have had it right for a long time, Windows probably since 2000 or so.

The trackpad, it's the opposite.

Re:Libinput's bus factor is 1 (2019) by LeonidasXIV on Friday May 22, 2020 @05:37AM (#23270219)
I have the same with the mouse, had to install http://triq.net/mac/mouse-acceleration to tune mouse acceleration to 2.3 (!). Now it is fine, though it forgets the setting every reboot (despite showing up as 2.3), so I have to touch the slider every time I restart. Fortunately that's not too often.
Re:Libinput's bus factor is 1 (2019) by Nursie on Friday May 22, 2020 @04:53AM (#23269976)
I boot between Windows and Linux every day, same hardware (logitech G703 mouse)

I have never even noticed a difference, let alone registered a preference.

Re:Libinput's bus factor is 1 (2019) by clarry on Friday May 22, 2020 @04:12AM (#23269746)
My xinitrc starts with xset m 0 0. I want input from my mouse to map 1:1 to pixels. Nothing more.

(I think libinput breaks that -- or at some point did break it -- and I solve it by removing libinput. Coincidentally, I started having issues with my trackpad when someone somewhere made libinput the new default. I fixed that by removing libinput)

Re:Libinput's bus factor is 1 (2019) by Ballas on Friday May 22, 2020 @05:06AM (#23270038)
I think you can still achieve the same using xinput to set the appropriate accel profile and speed. (And coordinate transformation matrix should be identity matrix)
Re:Libinput's bus factor is 1 (2019) by bscphil on Friday May 22, 2020 @04:35AM (#23269872)
I'm not clear on what exactly you're advocating. If it's just pointer acceleration, Linux operating systems have that already. Every input driver I've ever used on a Linux laptop (for over a decade) has supported it. It works very well - it would be difficult to use a touchpad on a typical laptop otherwise.

Is there something else you'd like to see?

Re:Libinput's bus factor is 1 (2019) by throwaway8941 on Friday May 22, 2020 @02:50AM (#23269322)
Please, god no. If anybody really does implement this, at least be kind enough to leave the old behaviour behind a configuration switch.
Re:Libinput's bus factor is 1 (2019) by rowanG077 on Friday May 22, 2020 @03:09AM (#23269425)
This is the single worst input decision ms has ever made. The first thing you do is to turn that hellish thing of.
Re:Libinput's bus factor is 1 (2019) by IshKebab on Friday May 22, 2020 @03:12AM (#23269437)
Windows has by far the best mouse input of any OS. What exactly do you want to turn off?
Re:Libinput's bus factor is 1 (2019) by badsectoracula on Friday May 22, 2020 @03:39AM (#23269584)
The acceleration (aka ballistics). I want 1:1 mapping of mouse motion to cursor. It is easy to disable on Windows though, just uncheck the "Enhance pointer precision" (award for the most misleading title because it actually makes precision worse).

It is annoying in other OSes that you have to jump through hoops to disable that stuff.

Re:Libinput's bus factor is 1 (2019) by bscphil on Friday May 22, 2020 @04:46AM (#23269925)
> "Enhance pointer precision" (award for the most misleading title because it actually makes precision worse)

I would argue that it does enhance the precision if it's correctly implemented. (I haven't used Windows in a decade, so I can't comment on that point.)

Example: on a laptop, move the cursor to the bottom left corner of the screen. In one very slow continuous swipe, drag your finger from the bottom left to the top right of your touchpad. How far across your screen do you get? In my case, I get only about 1/3 of the way across the diagonal.

Assume I have no acceleration feature on my touchpad driver. This means I have two options to make the touchpad usable. Either (a) I can get used to dragging across the entire touchpad 3+ times to get the cursor across the screen, or (b) I can turn the touchpad speed way up. Suppose I do the latter.

Now suppose my cursor is 10px away from some target and I want to hit that target with a quick movement on the touchpad. Because of the very high speed required by the previous change, my accuracy is greatly reduced. I have trouble hitting small targets with tiny movements on the touchpad.

On the contrary, acceleration gets me the best of both worlds. With one quick swipe, I can get the cursor all the way across the screen. This is the "expected" behavior most people have for a laptop the first time they pick it up. On the other hand, when I'm trying to hit a small target with a slow cursor movement, I have a much easier time doing that. This feels like "enhanced precision" to your average user - the cursor moves more slowly and accurately when I need it to.

Re:Libinput's bus factor is 1 (2019) by barrkel on Friday May 22, 2020 @05:36AM (#23270213)
Mind that you're replying to someone talking about using a mouse, and you're talking about using a trackpad.

Different input modalities deserve different affordances.

Re:Libinput's bus factor is 1 (2019) by Someone on Friday May 22, 2020 @11:47AM (#23273330)
I think that applies to mice, too. The ‘natural’ range of mouse motion is larger than for a trackpad, but not that much larger. The width of a typical mouse pad is about 8-10 inch, so if you use one, the range of motion is a bit smaller.

Apple’s Magic Trackpad is over 6 inch, the latest MacBooks have a trackpad of over 5 inch.

Re:Libinput's bus factor is 1 (2019) by barrkel on Friday May 22, 2020 @01:11PM (#23274438)
Mice don't have boundaries, trackpads do. The most awkward thing on a trackpad is initiating a drag and hitting an edge.

I think mousepads are mostly an artifact of pre optical & laser mice, if someone uses one today it's either because they have a glass desktop or they want wrist support.

Re:Libinput's bus factor is 1 (2019) by Someone on Friday May 22, 2020 @04:08PM (#23276355)
Your desk may have plenty of space, but arms do have boundaries. If you have to move your mouse a foot to move the mouse pointer to its destination, you’ll lift the mouse halfway through that motion, and if, on modern displays, you don’t have to move a foot, pointing at specific pixels becomes hard for lots of people.

Even the original Mac, with its 512 pixel screen width, had a setting that enabled mouse acceleration (fairly rudimentary. It doubled distance travelled when the sum of horizontal and vertical displacement exceeded 6 pixels).

I think the discussion is more about tuning the acceleration curve than about whether to have one or not, for both mice and trackpads. There probably isn’t a single answer that suits everybody, either.

Re:Libinput's bus factor is 1 (2019) by clarry on Friday May 22, 2020 @02:25PM (#23275282)
> if someone uses one today it's either because they have a glass desktop or they want wrist support.

Or they just want a good surface where the mouse glides smoothly, without any rubbing or scraping or pits that affect tracking..

I've never had a desk that felt good without a pad. I imagine glass would be smooth as long as it's clean.

Re:Libinput's bus factor is 1 (2019) by maccard on Friday May 22, 2020 @06:56AM (#23270706)
They aboslutely do. I've never really thought about it but this is obviously why a Mac trackpad feels smooth as butter but even booting into windows on the same hardware makes it feel awful.
Re:Libinput's bus factor is 1 (2019) by spuz on Friday May 22, 2020 @07:58AM (#23271137)
Unchecking "Enhanced pointer precision" doesn't disable acceleration entirely. Actually I doubt you would enjoy a 1:1 acceleration profile unless you have a very low resolution monitor.
Re:Libinput's bus factor is 1 (2019) by clarry on Friday May 22, 2020 @02:32PM (#23275364)
I'm on 2560x1440. In normal desktop use, it's about 17cm to get from left to right edge. Roughly 400 dpi. For graphics work and gaming, I reduce my mouse's DPI further.

1:1 is all I ever wanted.

Re:Libinput's bus factor is 1 (2019) by rubatuga on Friday May 22, 2020 @12:47PM (#23274081)
Actually true 1:1 is very enjoyable. I use it on a 1920x1080 windows computer. You need some registry hacks to get it working.
Re:Libinput's bus factor is 1 (2019) by IshKebab on Saturday May 23, 2020 @09:27AM (#23282753)
I've tried 1:1 and it is awful. To get from one side of the screen to the other you have to move your entire arm all the way across your desk. Not very practical.
Re:Libinput's bus factor is 1 (2019) by rubatuga on Friday May 29, 2020 @09:37PM (#23357680)
fair, although you can modify the DPI of your mouse to get it to where you are comfortable with the speed
Re:Libinput's bus factor is 1 (2019) by barrkel on Friday May 22, 2020 @05:36AM (#23270210)
I think a subtle acceleration effect may be worthwhile, but the actual acceleration curve is far too crude, and I too turn it off.

Acceleration works much better on trackpads though.

Re:Libinput's bus factor is 1 (2019) by rmtech on Friday May 22, 2020 @10:51AM (#23272727)
Because it makes the mouse input worse. I always turn it off.
Re:Libinput's bus factor is 1 (2019) by devit on Friday May 22, 2020 @05:45AM (#23270260)
I think it's better to simply buy a very high resolution laser mouse (which are commonplace now) and map sensor displacement to screen displacement in a trivial linear way.
Re:Libinput's bus factor is 1 (2019) by smichel17 on Friday May 22, 2020 @06:12AM (#23270421)
This is what I did, originally for gaming, but now I can't imagine gong back to mouse acceleration. A DPI button on the mouse lets me quickly toggle the sensitivity if I need more or less control for a particular task.
Re:Libinput's bus factor is 1 (2019) by ubercow13 on Friday May 22, 2020 @06:21AM (#23270477)
Yep. There is a reason why everyone plays FPS games that way, and it applies just the same to the desktop