9 hours ago, AndyC wrote

*snip*

No, but that's probably because the mouse pointer is naturally very close to the thumbnail by the time it appears (because it had to hover over the button to trigger it). As a result you're typically making a shorter, slower movement and are thus more likely to accurately hit the correct target. By contrast the IE tabs or the W8 app list potentially require a much longer range movement and thus accurately targeting quickly becomes more difficult. You may have heard of this before, it's called Fitts Law.

Hmmm... Or they could just make the thumbnails bigger. Maybe make the panel scrollable. You might have heard of this before, it's called Common Sense.

*snip*

Am I? Or did I explicitly state:

"

5 days ago, AndyC wrote

*snip*

What you're missing is that if you were designing an OS from scratch today with absolutely no legacy cruft whatsoever, even if you went with a "desktop" style paradigm with overlapping windows etc, you still wouldn't design it in a way that requires users to "open" or "close" applications.

"

No mention of mouse and keyboard there either. Next...

*snip*

What you're ignoring is that just because you take an app-centric view or move the application lifecycle to one that is more system managed - it doesn't mean you have to give up a desktop environment. Nobody is (or at least nobody should) say that all future applications will exist in a Metro-like environment.

No they just kill the development of APIs used to make the vast majority of desktop apps. It's a sure sign that the desktop is going to be well supported in the future.

 If you look at the last 15-20 years of OS development you'll see this is exactly the way everyone has been trying to go. It's precisely why the Apple Dock was created originally in a way that barely distinguished between running and non-running apps and why Apple at the time were often trying to explain the "with OS X you won't need to close apps" line (admittedly they someone pre-empted their OS being truly ready to take that approach. It's what things like Restart Manager and the OS X equivalent were meant to enable. Even your much loved Window 7 Taskbar is about blurring the distinction between launching and switching apps, albeit within the confines of what traditional Win32 apps could do.

I know this is quite a stretch but could Apple's Dock and W7's task bar just be better ways that user can launch to their favorite applications and documents quicker? You know without going through some heinous list of applications or sea of tiles? Maybe also forgoing the laborious chore of having to type the name of the application or document? – if you can remember what the name is. And perhaps a simple way of seeing what you have running without having to step outside the context you are working – you know beyond a crippled task list that is nothing more than an MRU list limited to your vertical screen space? Oh and just maybe a way to close applications and documents that are providing clutter to your active operating environment?

The only way in which Metro represents a shift in thinking is that Microsoft have learnt the hard way that they simply can't introduce bits and pieces into the OS and expect apps to embrace them in the same way that Apple can. The amount of software out there today that supports Restart Manager, for example, is statistically insignificant - because the sheer weight of developers wanting legacy OS support steers them all away from using new features (no matter how many times the idea that apps can "light up" by using new functions on a new OS, it's just a bit too rare in reality). In contrast all the big OS X apps are first party, so new OS functionality gets embraced fast, which in turn pushes third party developers to keep up. The end result is taking a much bigger, more holistic approach to bringing Windows up to date and dragging developers along with it and focusing that effort on the sector of the developer market (that which is predominantly consumer focused and thus carries less back-compat baggage).

And there it is... "dragging developers along". Having us spend the last 4 or 5 years coming up to speed with WPF and SL only to drag us on to the next thing. Now that's how you build good developer relations and make us heroes to our customers who have forked over hundreds of thousands of dollars for a "modern" app! I'm glad you're ok with that. I always thought you build something that was better and people naturally gravitate to it; kinda like Android....

It's not "wrong", it's just dated. There are many good reasons that moving in favour of a RESTful approach is preferred these days, the fact that such designs work better in suspend/resume scenarios is by no means the biggest.

Dated? Really? I guess you haven't worked in too many scenarios where the cost of the connection is too high for the frequency of data updates. Of course users don't need buttery smooth and accurate HMIs... They can live with less.

Oooh look, the goalposts move again. What do driver and hardware issues that cause resume to fail have to do with the speed by which a working PC can resume?

Well you did say "Do you honestly think even current PC's have trouble resuming from sleep?" If that's not trouble I don't know what is.

Clover Trail is the Intel SoC design. It may use x86 instructions, but otherwise it's a lot closer to the ARM chips you'll find in iPads and Android tablets. SoC, aside from supporting newer power modes like "connected standby" are far easier to handle from a driver PoV, because there is obviously a lot less variation in the support hardware than in a traditional PC.

Oh Andy I'm so very glad you've come to your senses and agree that the Build tablet is closer in performance to "W8 grade" hardware than you originally gave credit.

I sure hope that those SoC don't SuC too bad after their quick resume from sleep. Seems like a compromise to me...

Hearing != Listening.

Here's one that I don't prescribe to but you might better relate: Ignorance == Bliss

It won't compile but don't let that stop ya.