If you've never used a Windows touch screen PC, I can imagine that you can't imagine why you'd want a touch screen. If you have, then you know that after a day or so, you just start reaching for the screen intuitively, and it's a rude reminder of the bad old days when it's not a touch screen. This is what Apple missed: none of them use a touch screen, none of them bother to check out Windows, so they can't imagine how important it can be. The Touch Bar is a poor substitute for it, and calling any laptop "premium" in 2016 without full touch is marketing, not reality.

"This is what Apple missed: none of them use a touch screen, none of them bother to check out Windows, so they can't imagine how important it can be."

I surprised you don't envision that there must be a huge product development testbed at Apple, where they go over all these options in detail. They sold 4-6 million macs per quarter for years. They've been a product-development company ever since their inception. Sometimes it seems as if people think Phil Schiller and a summer intern are making these decisions on the basis of gut feeling.

This is like when Apple pulled back the curtain after the "Antennagate" brouhaha a few years ago, and it revealed things like a huge EM test facility, a CT scanner, etc., etc. (http://www.macworld.com/article/1152771/wireless_lab.html)

The power of Windows 8/10's UI has been severely compromised by the need to support both touch and mouse at the same time. The requirements/strengths of these two input modes are in conflict and trying to combine the two results in an interface that's not great for either:

* while the mouse is precise and allows for high information density, touch requires big targets since a finger is not so precise - so Windows 10 has big targets that waste lots of space relative to a UI that is designed solely for mouse/trackpad use

* touch allows interactions like swiping, pinching, and multi-finger gestures, but because Windows 10 has to support a traditional mouse as well these are rarely used or aren't used effectively

A touchscreen does not work well with an interface that's designed primarily for use with a keyboard mouse, likewise a keyboard/mouse can't replicate all touch interactions effectively. Trying to design for both in the same UI creates a lot of compromises that will delight neither group.

To me, it's about moving in this direction. Apple shows no interest in doing the work in MacOS to go here, and it's puzzling. Microsoft has iterated and iterated on it until it's pretty good. Not perfect, but they've already moved through a lot of the pain of making touch work well, and they're not stopping. When Apple figures out in 2018 that they need to ship touch screen laptops, they'll be years behind in UX on them... it's not as if they'll get it right on their first try, either.

I've had Windows touchscreen laptops for ~2 years and the only time I even remember it's there is when trying to point at something in conversation and accidentally clicking on it. Are there really use cases that matter for a touchscreen on a standard laptop? I understand it on a convertible, but not your standard clamshell.

I don't know how common these use cases are, but I find touch support useful while doing some web dev stuff. For instance, in Chrome, pinch to zoom is more precise than ctrl-scroll / ctrl-+. It preserves the existing screen dimensions (ctrl-scroll / ctrl-+ makes the actual font size larger and alters dimensions).

I'm not so vision impaired that I need to do this on a day-to-day basis, but it's great for zooming in to see if I need to adjust a CSS border by 1px or something.

I haven't tested this thoroughly but I believe Chrome also supports the same set of touch events it does on Android. So if you want to make sure the mobile version of your site is touch-friendly, having a touch screen is less clunky than trying to emulate it with a mouse.

The use case that matters for me is: when I'm using them as laptops. Most of the time, my Windows laptops are connected like desktop machines: external monitors, mouse, keyboard. When I take my devices off the desk and go to a cafe or my living room, I use the touchscreen all the time. I don't reach for my mouse to click a button when I have a finger to do it. I probably don't go more than two minutes without touching the screen in that scenario, even while writing code. Scroll up/down is much easier with a finger. Pinch-to-zoom is obviously easier.

I have little experience with vertical touchscreens. Some friends cheap PCs and my tablet attached to a Bluetooth keyboard.

I discovered that I was naturally touching the buttons in dialogs and web pages and the checkboxes and radio buttons in forms. It's faster than reaching for the mouse or the trackpad. With the trablet it was the only way and it was good. No gorilla arm.

Touching links in browser can be fast too, but the target must be big enough. HN is particularly bad at that.

Basically everything that is easy to do with the touchscreen of a phone and a tablet is easy also with the screen of a laptop. But the screen must not move when you touch it.

>This is what Apple missed: none of them use a touch screen

Some Apple employees might use iPads or iPhones.

> I don't really understand why one should look down and search for what to press instead of simply using a keyboard shortcut via muscle memory.

From a personal perspective, I agree. I think we generally forget though, we are vastly in the minority and this technology is designed for the layman

Basically my litmus test for this stuff is "would my technologically illiterate mum use and understand this. Will it enrich her computing experience". If yes, it's not a dumb idea. We'll see how the market reacts though

Yup, but why would a video editor buy a laptop that's limited to 16GB of RAM these days?

Because when your disk IO subsystem can do 2+ Gigabytes/second, RAM is much less of a factor.

Not really, unless you want to use it for a swap file. I'm currently sitting at 12 GB of RAM in use, with about 21 GB committed. I'm on half dev mode (watching netflix while hacking a bit)

Right now I only have 1 VM with one IDE open (thank god for vagrant), but I can have as many as 3 VMs and two IDEs (mixed .net and java shop here). Pushing 24GB of RAM in use is not uncommon for me (including the 50 or so chrome tabs and what not).

Could I make do with 16GB? Probably, but I'd have to change my development habits, and I don't really want to.

I was mainly referring to the video editing use case, where video could be streamed off disk quickly enough. For your use case, 16 gigs may indeed may not be enough.

What do you do with 50 tabs? I usually keep 10 open at most.