E14 : Trackpad Discussion Only

ok ive up on trying to use the touchpad with my thumb resting on the bottom. Is there a way to set up the touchpad macbook style, were clicking anywhere is a left click and clicking anywhere with two fingers is a right click?

 

that twofingerscroll program can do two/three finger tapping.

'course it isn't two fingers at the same time it's more a ta-tap really quick like.

If you just do two/three fingers all at the same time it'll trigger a right click and then immediately afterwords it'll do a left cick)

 

but is there a program that does this for clicking? Also itd be great if i program triggers left click when i click anywhere, cause clicking currently is completely useless since it only works on the bottom portion of the touchpad but the thumb method is a total failure here.

 

no. it's tapping.

clicking won't register everywhere since this isn't the mac trackpad

or the new synaptics track pad.

 

can anyone get the three finger swipe to go back on a webpage to work?

 

All this finger talk has made me want to show my love for this track pad.




- Myspace Layouts,Graphics, and Comments!

 

Anyone know how hard it would be to implement with the current SDK? Seems like it would be fairly straightforward if the SDK tells us when the touchpad is clicked. Seems like a great method for using the touchpad, and the most natural way to use a clickpad

 

well I think it's that there's no sensor for the actual button except down at the bottom on the left and right.

I mean you can't click at the top, that's where the hinge is.

 

well you can click in the middle of the toucpad, not just at the very very top cause of the hinge. And, you shouldnt need a sensor for a button....if the button is clicked, the hardware will register the click regardless of where your finger was when you clicked it, shouldnt it?

 

The drivers don't *directly* pipe whatever they get to the SDK listeners. Instead, it parses the data and applies a few corrections. Unfortunately, that means
A) We don't know the location of all fingers, only that more than one is pressed.
B) We only know when there's a left or right click, not when the button is clicked. I'm assuming the driver recognizes the location of the finger on the pad when the clickpad is depressed and only sends the click event for backwards compatibility with non-clickpad hardware/software.

Now I'm not saying I dug extremely hard for this - there may be a way, but it probably isn't documented very well if at all. The SDK is a nice touch, but it's no replacement for actually functional drivers or, even better (at least for the hardcore tweaker), a full low-level access to the data, maybe with the option to apply corrections automatically.

 

When you people double tap on your trackpad, do you get an annoying click? I get an annoying little click, almost like the trackpad is kinda loose

 

No clicking here. You're not pushing so strongly you're clicking the pad, are you? Just try pushing towards the very highest edge and then again near the buttons (but over the line). It's normal for a strong push to click the pad towards the bottom since this is a single, large physical button with two zones acting as left/right buttons.

 

I assume most of you have set your palmcheck to max settings? Could someone screenshot me his sensitivity settings? I've been trying to find that happy medium lately.

 

Yeah it feels different than a click. It feels like the whole trackpad is kinda shaking, and makes this loud tinny noise.

short video of what i'm talking about: http://www.youtube.com/user/Logryk#p/u/0/GiBpWWjwuW0

 

Darn, thats prety unfortunate. I just enjoy the tactile feedback so much more than tapping, but i guess i have to tap now since resting thumb on the bottom is useless.

EDIT: stupid question, but what does "two+1 finger" mean?

 

You put two fingers on the trackpad then tap with one finger to complete the gesture.

 

also, whats maximal finger movement distance?

 

Your video makes me think the clickpad is somehow loose. I'm afraid that means RMA unless you can live with it

 

Always easy to find flaws in any notebook. Examples?

After using my G50VT for almost a year and my Envy for a few weeks, I'm now seeing the flaws and downsides in both. The G50VT's screen quality is just fine overall. But compared to the Envy 14, viewing angles are meh and the keys are clattery. One hinge's plastics look like they're starting to come under some strain. Temperatures fluctuate way too much even during regular webpage browsing. On top of that all, it's damn heavy. But it's a gaming notebook, so can't complain!


My Envy's touchpad clatters too though only when i tap it hard. Keyboard flex is minimal. Sometimes Catalyst doesn't work properly or turns itself off for no odd reason but this is a once in a while occurrence Battery life is mediocre. Two-finger scroll even with the Google touchpad software is sometimes iffy--I might be trying to scroll down and the touchpad doesn't detect anything until I scroll down again.


But despite all this, I like both machines and I especially love my Envy. Introducing an new version of the Google touchpad software with flawless scrolling and new gestures software would certainly improve the user experience, but once again, that would just be a nice touch. All other factors like the keyboard, the beautiful screen and its resolution, and the build construction really add up together nicely for me.

 

Well I hopefully can get a part of your wishes granted

Attached to this post is the first version of my take on a new clickpad "driver". Currently supported is:
-Tapping, tapping and dragging, sticky dragging (all already supported by the drivers, but re-implemented here)
-Two-fingers right-clicking (and FAR more reliable than any other implementation so far, personally speaking)
-Two-fingers scrolling with momentum (also more reliable in my opinion)
-Top-left double-tap disables the touchpad (also supported by the default drivers, but had to re-implement that too)

Coming up when I get some time will be:
-Three-fingers tapping and swiping
-Customizable hotspots with two/three finger actions (stackable)
-Possibly using the clickpad outside of button areas, if I can get it to register (I have a few ideas)

I'm hoping this is bug-free but of course report any issue you might have with it. I'll do my best to support it and will be setting up Google Code to publish the full source.

You need .NET Framework 2.0 to run the program; shouldn't be a problem on the Envy 14. The program may also be compatible with other Synaptics touchpads, but I offer no support there.

Enjoy and provide feedback please

 

I can't get that zip to work right (it won't extract)

reupload it to something like mediafire

 

Cant unzip it for some reason...

 

Fixed, only 7-Zip wanted to unzip it for some reason.

 

A proper implementation of two finger right click, nice work. A cool feature would be if you could middle click by tapping on the upper right hand corner of the touchpad.

 

A proper implementation of two-finger right click would be another nice way to stick it to Macbook Pro users. Seriously. It's completely phail on Macbook Pros atm.

 

I have no problem with the right-click I made (contrarily to the previous implementation, which would right then left click and clear off any menu/selection). If you do, be sure to tweak the settings to your likings.

 

After installing your program, I cant reenable tapping the synaptics drivers. When I do, it resets a few seconds later and disables it. How do i fix this?

 

Make sure the program isn't running. If it is, it will interfere with the drivers.

There's no actual "installation", all it does by default is disable the gestures, which can be re-enabled by checking back the box in the drivers. If it's shut off, it cannot do anything to the drivers.

If you don't see it in the taskbar, make sure it's off in task manager. If even with that it doesn't work, try using the reset button on the Synaptics drivers, but I'd be very confused as to what could cause such a behavior.

 

I just restarted my comp, seems to be working fine now. It seemed like your program was trying to write to registry for some reason so i thought it perhaps made some permanent change but guess not

 

It's only writing to the registry to start at launch/remove that key. Otherwise it'll read the registry to know whether it is already setup to do so, and that's it.

 

The first implementation was able to do this, but it was sort of buggy: I am able to move the cursor, then go into scrolling mode without any issues in this version. However, when removing the second finger after scrolling in order to get back to moving around the cursor, I find that it is locked.

Possibly an improved version of pinch-zoom? It's WAYYYY smoother on macs than with synaptics (though this is probably due to the OS in some way, so it may not be possible). Thanks very much for updating this, I appreciate it!!

 

Hmm... I hadn't thought you'd want to scroll then lift a finger to move around again. The way it's designed right now indeed locks it down (I tend to do swipes when scrolling so I never had an issue with that).

Unfortunately pinch zoom would most likely require knowledge of where each individual finger is, which the SDK does not provide. I don't think it's possible.

 

a suggestion about momentum. when doing multiple swipes with two fingers, it queues multiple momentums and keeps going outside of the users control.

in the beginning of the handle packet, if there is a finger present, it probably makes sense to cancel the previous momentum. this would also prevent multiple momentums from happening.

i re-wrote two-finger-scroll envy last week (using vs 2010, so it might need .net 4.0) and i'm attaching it for a contrast of implementation/behavior. it was written for personal use so is not very configurable, it just behaves how i want it to.

it has:
-inertial scroll using robert penner's easing algorithms. (depending on how fast you were scrolling, it will go farther. touching the pad will stop the inertial scroll)
-three finger swipe back/forward,
-1/2/3 finger tapping (double tap and drag for 1 and 2 fingers)
-two finger scrolling.

i'm not really developing it anymore as it fits all my current needs (unless someone knows how to invoke aero peek from a c++ app). if someone wants the source, just ask.

two licensing things:
robert penner's easing algorithms license
the icon license

link to app:

edit: removed link, see later post.

 

NOTE: Include this file in the same directory as the software:

http://www.dll-files.com/dllindex/dll-files.shtml?msvcr100

Testing out your version. Two-finger tap to right-click runs better than on the MBP! Repping you for that as that's an amazing development (SUCK ON THAT APPLE FANBOYS!). I can't seem to get inertial scrolling to work. When I try to open up the popup, it zooms through what looks like 1,000,000 options then closes immediately. Also 3-finger middle click needs some fine-tuning.
Also need to work on my palms causing the touchpad to jump.


EDIT: Got inertial scrolling to work SOMEWHAT. Soon as you click on the dropdown menu, you have to make sure to capture one option. Or use your arrow to move down a few at a time.

 

Yep. As a tester, here's my assessment:

+Two-finger right click is fantastically implemented. But I may turn it off as it sometimes conflicts with two-finger scroll. If you could put in an option to turn off two-finger tap, that'd be great. The other "X" options go backward or forward.

-Still need something to prevent palm from hitting touchpad and moving the cursor

-Inertial scrolling shows promise. I'm tweaking with the settings and getting it closer and closer to MBPlike settings. Will post SS later if I get there.

-Major issue bug. When I try to do basic stuff like change some settings in Synaptics or click on a tab in slickdeals.net for some reason it gets registered as a double click even though I tap just once. Try it out to see what I mean.


EDIT2: Trying to find out which easing algorithm works best. I love the inertial concept, but i may have to go back to the other 0.1 version as this one interferes with palm settings. Almost seems to turn down palm check.

 

Awesome. After this program the trackpad is insanely good. Three finger gestures finally work all right. The only problem I have now is the click on drag using the physical push of the keys getting stuck which I use for gestures in Chrome or the occasional highlighting. For example I will be left clicking by physically pressing the key and using another finger to move the cursor and then the cursor will just fail to move all of a sudden and wont until I let go. Especially annoying when dragging a file somewhere and it gets stuck over a folder I dont want it to get dropped into.

Anyone have any ideas? I realize the click and drag is there too, but edge motion is also now gone.

 

yeah i noticed a serious problem with the palmcheck recently, possibly because of installing this program (not sure). my left and right palms hitting the touchpad triggers both the scroll and the right click, its unbelivably frustrating. It happened like 4 times already just writing this paragraph.

 

tapping will take precedence over scrolling. if you tweak the tap threshold and distance, to smaller values (depending on how you tap), it does alright. i'm at 150/20.

for inertial, i use easeOutSine and a 800 (ms) duration. its fairly natural to me. i gave up trying to mimic a macbook (or iphone) since macbooks scroll at pixel level while windows has mouse wheel level (which is 1 line minimum).

the palm check thing is probably true. i don't check any W or Z values. i don't know how synaptics maps the palm check values to Z values.

i don't have your major issue. you have to turn off tapping in the synaptics settings. i don't lock out other software from receiving taps (since i personally don't have anything else running that reacts to tap events), so every tap can be reacted to by multiple software.

anyway, added max w and z threshold configuration, this will probably help with palm check:

edit: had a bug, removed a line of code i wasn't supposed to.
edit2: added max z to scrolling too. i don't think w will help here.

link

 

Will try tomorrow while at coffeeshop. Getting bored in my apartment and it's LATE!

 

looking at the packet data that my palms can generate on the pad, i don't think its possible for someone to completely prevent the palm from triggering a tap or scroll. if the palm hits the edge of the pad (or is lightly pressed against the pad), it will have W and Z values exactly like a finger would generate.

with the max w and z thresholds, it triggers a lot less often than it did before.

 

Patent of the future:


Miniature sensors that lie either between the keys or underneath the primary typing keys (e.g. F, J, space bar, etc) which trigger every time something comes within a cm of hovering over them. It could be modeled off of the way the webcam detects light and adjust the backlit keyboard's light.

So outta curiosity...how is the MBP's trackpad good at avoiding palm swipes?

 

Damn. I'm going to have to revert back to v.02. Seems like setting Z all the way u prevents me from using the bottom two deadzones.


EDIT: Having problems now. Soon as I turned off the inertial version and returned to v.02, I can't seem to get tap to click to work anymore.

EDIT2: Rebooted, but now it seems like my deadzones are gone. Did the inertial version disable them?!

 

well didn't one of the things disable tapping altogether when you installed it? (It was just disabled in the touchpad settings, not like disabled in the driver)

 

Nope, had tap to click enabled in Synaptics. Rebooting fixed that problem but not the "dead" deadzones. It was only after I started up the inertial version did the deadzones return to normal. Funny part is the Z settings were back to default. I'm going to have to wait and see until someone perfects this!

 

Using reil's program Im still having issues with jumpy scrolling which I hope someone can figure out, palm tapping, and the physical click and highlite seizing up in the middle of a drag. Any fixes appreciated.

EDIT: Watching my click in the moodpad I can see why it gets jumpy and its all on Synaptics' end because I can see the registered click jumping around from finger to finger for some reason.

 

Seems like reil's solution is already pretty capable. It'd be preferrable to only have a single application like this, so I'm afraid I won't be developing mine any further. It does what I needed it to do and it's less thorough than reil's, plus I'm too busy right now to add anything to it.

 

apple has access to the raw pad data and has dedicated engineers to develop drivers and the application and lots of usability testing. we have a synaptics sdk which doesn't surface a lot of information and people with some spare time. its no question that they will have a better experience. i use a macbook at work (though older), its just a lot better.

i've had issues with the synaptics control panel not saving settings when i wake the laptop up from sleep (before i ever wrote my app and was using two-finger-scroll-envy). i have to reboot sometimes then change the setting for it to stick. i'm not sure if anyone else has experienced that. this may be related to your inability to activate certain synaptics features. for myself, i just disable everything manually in the control panel and use my app. i don't use deadzones or anything. i also use tapping/dbl-tapping for everything and no buttons.

i don't think its possible for my version to disable anything. i don't recall ever using any set property anywhere in the code. i decided that since my use case was to just disable everything in the synaptics control panel and i have no other packet handlers running, i can just leave it all alone. my application is highly likely to ignore any settings set in the synaptics control panel as long as the packets gets sent through the packet handler.

Friendly0Fire, i don't think i have a more thorough version. you did a better job for a complete experience than i did since yours sets the user up during startup and considers more potential configurations than mine does. the only thing i think i did better on was the inertial scroll (and that's only because i referred to my iphone when trying it out) and its the thing i wanted most. i hope you don't stop development since i don't intend to develop mine further unless someone mentions a feature that i would really like (like three finger up or down for aero peek).

 

It's just that right now we have three apps doing just about the same thing with small differences for each... In this specific case, diversity isn't good as they're all mutually exclusive. In order to create the best experience available, we'd really need to have a single solution that can do just about everything properly, otherwise we're just splitting the workforce needlessly.

For inertial scrolling, I honestly have never used an iPhone/OSX (and I don't intend to, either), so I've no clue what their inertial scroll looks like. My current method is simply to take the last scrolling distance reported by the SDK and multiply it by a value <1 recursively until the distance falls below a certain threshold.

For Aero Peek, I'm not exactly sure what you mean, but couldn't you simply move the cursor to the bottom-right corner through SendInput's absolute mode? I know it's a rough solution but at least it'd work.

 

i guess we have different reasons for doing this. i'm doing it because i'm a java dev and not very good at c++ and doing this to try some things out. also, all of the software at work has to run on linux, so i haven't done windows application programming for a number of years. whether it gets fragmented or not, shrug, its not a big deal to me. if someone makes a version that is more useful than mine, i'll delete mine and use theirs. this was just for fun/learning.

apple does some things very very well. their UI is so polished that i can see why people constantly compare to apple products. the inertial scroll on the newer macbooks is way better than what i did but is probably also patented.

i put my source here if you want to take a look:
envyTouchPad-src.7z

i take the last delta x and y and use one of the easing algorithms by robert penner to degrade them over a period of time.

i could also press windows+space to trigger aero peek, but moving the mouse or holding those buttons are pretty ugly hacks. i haven't found any resource on explaining how to programmatically invoke aero peek. i honestly would never use this feature anyway. its like flip3d, it'd just be kinda cool to do but not very useful.

 

How about having each 3finger swipe activate user set key commands with the option to hold the command until another key is pressed? Possible?