bztzt

It's a technically deeper version of the talk in that link.

Win8 does implement fuzzy touch targeting (looking at the geometry of your entire finger contact region instead of a single point, and picking a best-match UI element) on the desktop. Office 2013 (desktop) also invested in a bunch of touch improvements. It hasn't been enough. (plus, they already tried this approach with touch on win7 and it didn't seem to be working - which, among other reasons, was why they changed tack for win8 in the first place.)

You're right that there's nothing inherent to the tablet form factor per se that makes it less vulnerable to malware. It's just that the different input mechanism and smaller screen size makes backward compatibility less useful, and it's easier for platforms (including Windows RT) to build security into their app models if they don't have to worry about compatibility.

Definitely agreed. Another interesting problem for the near? future is multi-device window management - can you use the collection of devices and screens (whether phone/tablet/desktop/whatever) on hand at a given moment as a kind of ad hoc multimonitor workstation? What about when multiple users are involved? (of course all of this is a systems/OS and privacy/security challenge, as well as an interaction design challenge)

I sometimes get the impression most people now think of window management as either a solved problem and/or one rendered irrelevant by the rise of "mobile", but it's quite the opposite IMO.

Incidentally, early prerelease versions of Windows 8 used the "Open Share" icon: http://www.neowin.net/images/uploaded/195351i26jesbggjt18g0i... ; wonder why they changed it.

since day one of Windows 8 on tablet/PC, but it wasn't brought to the phone until 8.1.

omg, something finally supports XAML 2009. =O

The search charm (which typing on Start is a shortcut to) does search settings from throughout the system, including Control Panel in the desktop, but I'm not sure if third-party extensions are included in the search.

I once worked at Microsoft on a product that was actually getting almost entirely positive feedback (while still in preview) for once and ... it was kind of frustrating. I started scouring the internet for "<product name> hate" (and did find haters, some with interesting povs) :)

Live Mesh also still exists - it became the SkyDrive sync client which is now built into Windows.

Groove actually still exists, though it's gone through a bunch of branding and strategy contortions:

Groove 2006 -> Microsoft Office Groove 2007 -> SharePoint Workspace 2010 -> SkyDrive Pro -> OneDrive for Business (the current product)

Correct that you can't do this on Linux. The Microsoft CLR (.net runtime) itself is still proprietary and available only on Windows-based platforms.

At a very high level, sure.

Windows 8.1 can run multiple instances/windows of the same app, although the app needs to opt-in to implement support (which the built-in PDF reader, IE and mail client all do).

But that's an example of why inheritance-as-subtyping is dubious - sharing implementation doesn't -> sharing behavior, and for subtyping common behavior that can be relied on by clients is what we want.

I actually think a lot of media has both sets of problems at the same time. They will take press releases or whatever and reprint them coated in identical thin layers of snark and generic cynicism, without providing any real criticism or context.

IE8: XP/Vista/7 IE9: Vista/7 IE10: 7/8.0 IE11: 7/8.1

Good post (even if I'm not sure how much I agree with the mouse travel argument - I find moving to the edge still quick even on largish monitors, but it probably depends on your mouse/speed acceleration settings) but I'm not sure there's anything specific to open source here - understanding how and when to apply lessons from the past is tricky everywhere in software. I think design understanding is probably reasonably well dispersed by now and open source not at a marked disadvantage from the rest of the software world, unlike the 90s/early 00s.

Sure, and that's the case on Windows too. You can launch and switch exclusively through the start screen and never use the recent-apps swipe. It's just included in the tutorial popups for whatever reason (maybe because Windows users are expected to be more keen on multitasking, given its desktop heritage and since that's a point of differentiation for Windows vs. the iPad).

Settings is the place for configuration and management of your computer (and the current app). Shutdown/restart is a management function so it goes in Settings.

The problem with Windows 8 on traditional PC form factors is it deliberately ignores ("blows up") the model that's been fixed over 20 years in users' heads and replaces it with a new model they've attempted to design from scratch from first principles. It's not that the new model is especially complex or badly designed on its own terms, it just clashes with the model people already have in their heads.

That is also why people often say the Windows 8 UI works well with touch, but is unusable with mouse and keyboard. The inherent usability characteristics are actually not hugely different between input methods, it's just that with touch there's not as much competition in people's heads from an entrenched existing model.

Both do make sense insofar as they are consistent with the design model/philosophy of their respective OS shells.

In Windows 95 the model is that Start is your starting point for everything you do with a computer. Moreover, managing and configuring a PC was regarded as a core activity for all users, so management functions like shutdown/restart and the control panel got a prominent top-level place in Start.

For Windows 8 the design mantras were "content over chrome", "users should be able to enjoy just using a PC without feeling like they have to manage it", and "each view in the UI should have one clear purpose". Therefore Start was dedicated to "destinations" like apps and content, while management and configuration functions like control panel and shutdown/restart (meta-functions that are about the PC or app itself, not what you do with it) were moved to the Settings charm (could've been called "settings and management"), sort of like tucking away wires and cables.

Of course, making sense in terms of your design model is no guarantee a decision will make sense in terms of the user's model.

The Windows "tutorial" item in question is a shortcut for switching to recent applications. The equivalent action on iOS is either a four-finger swipe to the right, or else double-tapping the home button, then tapping on an app. These are obvious to 2 year old babies and cats?

Thurrott has been bizarrely bipolar about Microsoft for a very long time. He is pro-MS "on average" but will careen wildly between unwarranted hyperbolic praise and unwarranted hyperbolic invective. Sometimes about the exact same things within the span of a week or two.

I would still question the value of folders. It makes the homescreen customization experience more confusing and frustrating for everyone in exchange for functionality that's only really useful for a small % of enthusiasts.

Microsoft, infamously, has been willing to lose billions year after year on new products in the hope of future success. They spend as much as anyone on basic research and incubations with no certainty of productization. For better or worse they've also taken on major redesigns of existing products focused on future scenarios at the risk of damaging short-term success (e.g., Windows 8, office 2007). None of those things fit his narrative about narrow short-term profit maximization.

(The example of the above he does mention - Xbox - he dismisses because it's not that profitable! So when something is profitable it's a sign of narrow short-term profit focus, but if it's not it's irrelevant because profit is what matters. Huh?)

Microsoft does seem to have trouble turning its R&D activity into breakthrough products, but his attempt at a narrative explaining why just doesn't fit the facts and doesn't make sense.

I think part of it is the new style of desktop/shell itself just hasn't yet had enough time to mature.

The article is wrong, Sinofsky came on at the very end of Vista development and had little to do with getting it out.

Porting all of the Windows desktop is essentially negative effort, in that the new parts of the shell are (intentionally) so technically intertwined with the pre-existing parts that it's easier to just port everything than try to separate them.

To me what Myhrvold is vaguely describing as the "IHC" in this memo seems more like a set-top box or game console than a phone or tablet. I think this may have influenced Microsoft's decision to pour resources into the Xbox starting in 2001. They got it right that there would be an "IHC" but misjudged what sort of thing it would be.