Cman

@aLEXXOid, you misunderstand what open source is. Open source community can always do better than a single time-constrained developer because it is more resources, testing, hardware, etc. Nothing to do with how talented or not a single author can be. Second, it is not about getting a ready made solution as you stupidly suggest. A piece of software may not work for all or may have boundary conditions that affects some people. A single developer cannot devote time to fixing all of it especially things that affect a small number of cases although the aggregate of such problems may affect a large number of people. But a community can fix such small items and boundary cases much more efficiently for better software in the end. My first post here was a number of suggestions on how to make it better. But I understand the resource constraints of a single developer and don't expect all of it to happen but that is a problem with this closed source work. Third, the author is violating the principle of open source software licensing by basing his kext on third party code and source code provided under open source license. Perhaps, your ethical framework does not see that as a problem but it is a problem for many and why links to this site may be discouraged at some sites. I have no problem if the author did not violate any open source licenses and had a proprietary piece of software to do as he pleases. Then your comments would perhaps make sense. My post above gave that choice to the developer. People have been more than patient with the author's foot-dragging on open source to the point it now seems dishonest and misleading. I am even sympathetic to the author's need to be monetarily compensated for the work but it cannot be at the expense of others who have made open source software available. This is like the unethical developers that take open software and sell apps on App Store in violation of those licenses. Good for you that you are enjoying the fruits of others that have followed the spirit of open source licensing that this work is partly based on and so don't feel the need to take a principled stand or have empathy for those for whom this does not work. But that is more about your selfish perspective than what my post was about.

Here is another example of annoying unintended consequences with feature creep/poorly-tested interactions. With my settings (where I have tried to turn off everything with multi-finger except 2f scrolling), if I slightly touch with two fingers and hold for a very tiny amount of time (not a click or tap), the driver sends some codes to the program that has focus. In Xcode for example, this results in the navigation window toggling (equivalent of Cmd-1/Cmd-0). So with inadvertent touches, my Xcode navigator window keeps opening and closing while working on it. It seems almost impossible to just have a simple trackpad with two virtual buttons that are completely dead except for clicks (and do not move the pointer) and allows 2f horizontal and vertical scrolling. Nothing more. Is that too much to ask without having to fight with settings for multi-fingers, rotations, inertial flings, sensitivity, palm rejection, etc, etc, etc? Very un-mac like, in my opinion.

I appreciate the author has done a lot of work on it but I would strongly recommend either the author makes it a commercial ware and spend sufficient time on it or make it open source and release the source so bugs and limitations can be quickly fixed without depending on one person's limitations and tendencies for feature creep without stabilizing what already exists. The current situation is the worst of both worlds. One person keeping the source code and not having time to fix it. There are far too many problems with this driver (while it works fine most of the time, it is very annoying to use as it keeps showing its problems). 1. The driver crashes and freezes 2-3 times a day requiring a reboot for a Elan v4 touchpad (the hardware is working just fine as the laptop is used for other operating systems). It seems to be related to some weird interaction between the settings (the virtual buttons and multi-touch while scrolling) that I cannot reproduce reliably but if properly instrumented could be isolated with help fro the community than depend on one busy person's time. 2. The tap is way to sensitive to function as a click even with all the settings. Even if the tap is disabled for a click, placing a finger on the virtual button and tapping still acts as a click. This is an example of bad interactions between multiple features that takes time to isolate and fix, not something a single individual with limited time can fix in any reasonable time (especially when they don't have access to different hardware). 3. Just changing some setting can make the kext crash on boot, again related to weird interactions between the different features. This is not isolated quickly except by the community working with source code or the developer working in it full time. 4. Keeping the source code closed violates licensing of open source code used as the basis for this kext. No one can really enforce this of course but it does make the developer seem less than honest. I can understand the developer hoping to make money via donations as long as possible by keeping control and source closed. Then make it commercial software without violating the open source licensing of code it is based on. The current situation is very dishonest. Many people have pointed this out. The dishonesty brings into question whether the source code can be trusted. Malware inside drivers have access to just about anything on the machine. 5. Sorry if the above seems harsh especially to those who may be using this driver successfully but they are only one step away from acquiring a new laptop or new hardware where it breaks and having to wait for the developer to find time. No, I am not going to send you a Skype ID and participate in your beta tests. That is not a solution. Thanks and shame in the same sentence!

V4.2 on a high resolution screen 1920x1080 is too squirrelly. Just tapping the touchpad often makes the cursor jump. Have a Elan v4 touchpad. The touch on the virtual buttons area makes the cursor jump even more. So, it is very difficult to accurately click a required area or button. Which setting will control this to dampen the movement on a tap or click? As I mentioned in PM, there should be an option to make virtual button area insensitive to finger drags. Only taps or clicks. I have also sent an algorithm by PM to separate taps from touches based on time of contact and pressure which no one outside Apple seem to have implemented. Hope this can be implemented in a future version.