A GPU driver for a new undocumentated architecture is "trivial" compared to properly doing an input driver for a touchpad?
You make "properly handling any type of free-form user input" sound like rocket science, when it's the regular multi-touch kind of driver we have on billions of mobiles, and laptops...
The "properly" here is just the sauce for macOS handling, not some inherent difficulty of "handling any type of free-form user input" on a 2D surface...
It's because there is no great trackpad support outside the Mac. But there is good GPU support. So it seems, getting great trackpad support IS rocket science currently, for whatever reason. (I also wonder, what is going on there? I guess people on the other platforms just don't care about this detail, otherwise they would be on the Mac anyway).
There are a lot of crappy trackpads on crappy PC laptops, but the ones on the high-end laptops are quite serviceable even if not the ideal version that Apple created. With a serviceable trackpad, the other benefits of having a PC over a Mac tend to far outweigh the trackpad (for people that value what a PC brings).
Yes, I understand that, I am just not getting my head around why for example Windows is so bad with a trackpad either. I mean, how tough would it be for Microsoft to nail this if they cared?
They probably don't see it as their job. The fact that trackpad quality varies between makes/models implies that it's both a hardware and a software problem, the latter probably being drivers. Maybe there is something in the OS that hampers it though, I don't know. Would be interesting to hear from engineers at the OEMs. Frankly, when I read these comments about the problems of PC trackpads, it's like a foreign language to me. Problems with gestures, multi-touch or palm strikes.. I can't recall the last time I had a PC laptop with those problems. My XPS 2-in-1 9310 sits in a bag across the room and am typing on a four-five year old Lenovo right now, neither of which have any such problems. What I notice on the wife's MBP trackpad is the feel (glass) and the (good) lack of physical movement in trackpad - this lenovo is clearly a momentary switch and at the very upper end of the pad the force to needed to click is tougher, but it hardly matters. A tap accomplishes the same thing that a physical click does and is probably how I engage it most often. I'm just not sure it's as big a problem with decently-built PC laptops as people make it out to be. PC hardware has a lot more going for it that overshadows the delta between a 99% trackpad and a 94% trackpad. Things like a touchscreen, active digitizers, tablet modes, escape keys, function keys, facial recognition, variety.
I will never need a touchscreen in my laptop. That's what my iPad is for, which I can connect to my laptop, by the way. Function keys I've last used in the previous millennium, and I've got an escape key, thank you very much. My laptop automatically unlocks via my watch. But yes, I expect facial recognition to come to the Mac very soon, too.
And I've never used a PC trackpad with satisfaction rate of > 30% ;-) They are all shite.
I guess it is just a matter of preference. As I said, if you care about the trackpad, you will never touch a PC laptop. If you don't, you think you have a 94% trackpad, and that it doesn't matter.
I imagine getting the trackpad right is either a giant pile of heuristic rules, or a nice tagged data set for ML, or both, maybe with a side of online learning from usage. A trackpad that automatically does the right thing for me coding, my 12 year old gaming, and my 8 year old with sloppy fine motor control does seem like rocket science to me.
I'm not surprised at all. I bet there are plenty of multi-billion dollars companies using that project who never contributed a cent to its author yet sent their developers on GitHub to complain about the state of the project.
It happened to a few maintainers I know. These companies save millions in development cost on the back of a few developers in developed countries. This is open source, I get it, but this isn't fair either.
Look at that comment:
> ....yikes. Sounds like a fork needs to happen. And github should really look to provide a 'risk' rating to projects from a maintenance PoV... a project depended on by 4.5m users with 1 maintainer should visualise as a high-risk dependency.
It's not up to GitHub to do that, it's up to the individual or the company to audit a project and its dependencies and be prepared to be able to maintain it themselves. Or pay the only developer to do exactly that.
Another comment (quoted by another guy):
> Babel maintainer here
We are probably not going to fork core-js because we don't have enough resources to maintain it.
And people expect a single developer to have the resources to do exactly that? look at babel and the list of sponsors, yet they don't have the resources:
Another part of the loop is that many hiring groups also expect to see some public code out there. Maybe it's some library or widely popular project that helps your hiring chances or maybe it's just a toy project. None the less, you better have something sitting out there active for free.
Then, once you're hired and they're paying you, they certainly don't want you wasting their resources (paid time, IP) contributing free software.
The entire tech industry is morphing more and more into a dumpster fire itself.
In Japan there is sometimes a button for cyclists/motorcycles to push to get the lughts to change. I assumed this was because there is a sensor in the road to detect cars, but now I'm not so sure.
If the pedestrian was about to get green (or blue) but a car was detected and the pedestrian was forced to wait then that's not fair is it?
We already prioritize cars enough, pedestrians usually have to press a button to trigger the crossing so they're always waiting but cars don't get punished for getting to the lights a second after the sequence starts.
Close to where I live there are a couple of funny examples:
In one place the lights are red by default for everyone. Cars are detected from 100m or so away, so they barely need to slow down. Pedestrians need to press a button and usually wait for a while.
At another crossing the lights are green by default for the bigger road. A sideroad has car-detection that is ofter rather quick. Pedestrians and cyclists need to press a button and usually wait for a long time.
Just because you don't sit in a car it's ok to wait that much longer.
A random pedestrian is just as likely to be helped by a light change as hurt by it. You provide an excellent example of motivated zero-sum reasoning though!
Google is a black hole that sucks in the best talent. I'm sure once you're there it's a singularity of extremely talented engineers. So if anyone were to have good security practices, it would be Google.
That's fine and all, but so is every major tech company in the public cloud space. And humans make mistakes regardless of how talented or brilliant they may be.
Do you have any details of which ones? I have not discovered any so far - everything seems to work just as well in Firefox (e.g. gmail, youtube, search, maps, adwords, adsense, analytics, cloud console, webmaster tools are the ones I requently use on Firefox with zero issues)
