User Research Gone Astray: The Case of Windows 8 Explorer | Hacker News Reader

User Research Gone Astray: The Case of Windows 8 Explorer

thomaspark.me

45 pointsparkov40 comments

The article asserts that MS completely got it wrong but it doesn't even attempt to persuade the reader with arguments as to why. To me it isn't obvious that adding a ribbon with the most frequently used features to the top of the explorer is a bad approach.

I think the ribbon was a mostly a success in MS Office, Notepad, MS Paint, Wordpad and so on. So why not also use a ribbon in the explorer?

Bingo... This is what I was thinking about while reading the article, He cites 1 other blogger who has complained about this. Personally, I generally use the shortcuts for copy-paste but I do appreciate having the ability to avoid right-clicking to do some of the other tasks

You appreciate not having to right click? Right click is on your mouse, where as in order to use the ribbon you have to go up to the top. This seems like pulling the most useful feature out from under your finger to duplicate it somewhere further away. How does this make sense?

Because generally you don't merely right click. You right click and then have to navigate a small menu of options to find the one you want. The big button on the ribbon is a huge target and is generally easier for users to hit and click.

And for some reason, I personally find right-clicking with a track-pad much more difficult than right clicking with a mouse, although I haven't taken a lot of time to analyze why as I generally use a desktop machine with mouse.

Or right click and have to navigate a huge menu of options, if enough third-party shell extensions have had their way with your context menu ...

If you live in a non-english speaking country, you'll learn that Microsoft does us the favor of "translating" the shorcuts. So, in portuguese, CTRL + S means "save" in about every program you'll encounter, but for Microsoft products, it means "sublinhar" (underline).

CTRL + A is not "select all" but "abrir" (open).

Well, that's just messed up... thanks a lot Microsoft...

The before and after pictures are supposed to speak for themselves, I assume. I know that for me, just seeing the two pictures makes is very clear "how they messed up." I'm willing to agree that not everyone is so offended by huge, ugly, attention grabbing, screen real estate hogging interfaces.

Exactly what I thought, it's almost as if they implicitly say the only good way to make a file explorer is to make it just like osx' dolphin, where you need to use drag motions for any operation. Sure, let's get rid of the buttons and the context menu and make our users open a couple windows so they can drag stuff around.

The ribbon is a weird thing for me because despite trying to adopt it for quite a while now, I still don't like it, yet I'm not egocentric enough to think that alone makes it a bad design. The evidence that it isn't is all around me... everyone I know who is a "regular person" and uses Windows loves the ribbon.

Damn it.

At least Microsoft generally gives you the option of configuring the ribbon manually in a way that suits your needs (especially in Win 8). This is much better than when I run into UX/UI elements that do things I personally don't like when Apple is involved, because then I'm pretty much stuck with it.

The article assumes that Microsoft's goal is to empower and delight the user. This isn't really the case; their goal is to reduce support costs for corporate IT departments. Many of Microsoft's strategies and tactics that seem nonsensical seem more sensible once you realize that corporate IT managers, rather than end users, are their customers.

I think that's a bit of a facile assumption. Sure, Enterprise is huge for Microsoft. But so is the consumer space. The vast majority of new computers sold are still Windows boxen, and at $250 a pop home users are still a massive market for Microsoft. Sure, many or their choices since the NT integration with XP have been driven by Enterprise needs, but not at the expense of home users.

I'm far more ready to believe that they simply messed up than that they are willing to make huge usability and design sacrifices just to appease IT managers.

Also, remember that we are probably the least important people to Microsoft. I'd say less than half of us run Windows on our personal machines. Our design sensibilities don't matter to them; they're designing for our mothers and technophobic friends. In a word (3), the facebook crowd. We use keyboard shortcuts, they consider having to right-click to be a burden and a skill that needs to be learned. We compare the relative benefits of various anti-malware tools, they renew their subscription of Norton that came with their computer three years ago. We lament the general craptitude of cmd.exe, they would be scared if they ever saw it. We talk about the inferiority of snap-to-grid fonts versus anti-aliasing, and they wouldn't know what kerning was if it bit them in the ass.

Windows is not designed for the tech and design crowd, so while we might think of the ribbon as a design travesty, the people it's designed for are the ones who have three toolbars strewn across the top of ie6.

Let's just say that my analysis is not uninformed. There are countless examples of how Microsoft puts IT managers' needs first -- if this were not the case, IE6 would have been stricken from the face of the earth five years ago.

You're also quite mistaken to think that Microsoft (who, you'll recall, is still run by Steve "DEVELOPERS! DEVELOPERS! DEVELOPERS" Ballmer), doesn't care about what nerds think. It's no accident that they have a whole division to sell tools and technologies to developers that's separate from their cash cow operating system and productivity divisions.

>You're also quite mistaken to think that Microsoft ... doesn't care about what nerds think.

We aren't the target market for Windows. Yes, Microsoft wants to get more developers on .NET making Windows apps, and to the extent that it needs them to use Windows they'll make it at least usable for that purpose, but they aren't about to make Windows harder to use for your grandma just to court us. We prefer keyboard shortcuts (which they have) to messy buttons all over the place, but if buttons make the product more usable for their core market then that's what they'll do.

> $250 a pop

Home users typically buy OEM windows, which is quite a bit cheaper: http://arstechnica.com/microsoft/news/2009/09/newegg-reveals... suggests ~$150.

(This is irrelevant, until you get into revenue discussions.)

Right, and corporate IT managers want interfaces like Metro which cannot be turned off and require touch to be used effectively.

When the F will people start seeing things from a non-power users perspective before writing such a rant?! At the very least, back up your post with some sort of an objective argument.

I know a lot of people who spend minutes looking for simple options that are hidden in menu entries. I think the ribbon is going to be of good use for such people who'd appreciate the most common use-cases/functions within the reach of a mouse-click.

for such people

Except those people are quickly dying out - literally.

Also, by any metric, the visual layout and appearance are beyond terrible. There doesn't seem to be a logical grouping at all.

If this is the answer then someone must have asked the wrong question. (a button labeled "easy access"? seriously?)

"Except those people are quickly dying out - literally."

Eh, I strongly disagree. I still have the misfortune to every now and then watch people who are supposed to be from the 'internet generation' (in their teens to early twenties) struggle with just-beyond-basic computer tasks - like opening a file with a different program when the extension association has somehow been screwed up. When I extrapolate their demographic traits, I can come to no other conclusion than that the vast majority of the general population does not have the computer skills that designers and software developers assume them to have.

To the contrary, I'm looking at this very much from a non-power user's perspective. I think a ribbon is a reasonable application here, showing copy, paste, etc. by default. But look at all of the other functions located throughout the chrome. As the "lambasting it" link says, many of the options that are visible by default are used much less than 1% of the time, according to Microsoft's own data. "Invert selection"?

But more to the point is how the user research data is turned into a redesign in a seemingly mechanical way. "The button was used 10% of the time, so let's make it bigger next time." Without careful consideration, this can be a big mistake, and can miss opportunities for helping users with their tasks at a deeper level.

I suspect you missed the point of the article, which was that the telemetry measured the wrong level of abstraction in users' tasks.

Problem is the ribbon is so full of buttons, that people will spend just as much time looking in the ribbons, as they did in the menus.

>"It seems that essentially, every single command that customers have used or requested has been moved into a ribbon or wedged into some corner of the chrome. And many are rightfully lambasting it."

Being surprised by Microsoft employing the ribbon and persistent UI elements in a redesign of any software product is nothing short utter cluelessness.

The story of the ribbon: [http://www.youtube.com/watch?v=Tl9kD693ie4]

I followed Jensen's blog postings all throughout the development of the ribbon; but it's all highly unconvincing to me, because it fundamentally is at odds with how I interact with software. Menus work for me because they are lists of text; when I'm looking for a command, a list of text is exactly what I want. If it's a command that I use often, it doesn't belong in a menu - for that, something like a ribbon might be a good idea, though a toolbar uses less space and doesn't have tabs, so I think it's better - but for commands I don't use often, lists of text are very close to ideal.

>"but it's all highly unconvincing to me, it fundamentally is at odds with how I interact with software"

The same argument could have been made about GUI's v command lines or command lines v punch cards.

Any new UI architecture will be fundamentally at odds with how you interact with software.

Sure, any new UI modality will be at odds. But for mouse-oriented UI, I haven't seen anything better for a situation where you have many, yet infrequent, commands. A search interface seems like it could work well, but you'd have to know (a) that the software supported the command you're looking for, and (b) what aliases the software supported for that command. The Windows start menu search kinda works, most of the time; but that's because you generally know whether or not you have a particular piece of software installed. But when you don't know 100% if the software actually has the feature, and you're looking for it...

So in the absence of an effective search, it seems clear that a list of commands, grouped into categories, is best for finding infrequent functionality - particularly when a mouse is the modality.

One thing I am absolutely certain of: searching for a command in a ribbon is harder than searching a menu. I've had opportunity to do both in unfamiliar UIs, and searching ribbons is definitely harder.

I'm not sure the comparison is quite the same. The key difference is that with a CLI or punch card interface, you need to know the command syntax in advance - it can't be discovered by interacting with the system. With a GUI, you can discover the appropriate commands to make the application do what you wanted to do.

The menus vs. ribbons debate essentially boils down to one of aesthetics. But I'll personally agree with barrkel in finding it much easier to locate the function I need by browsing through one-dimensional lists of text than by browsing through two-dimensional grids of pictures (of varying sizes and shapes).

The Ribbon may make locating a small set of basic features slightly easier for inexperienced users, but it makes far more difficult for everyone else to work with more complex functions - the very thing for which software adds the most value.

I find it painfully ironic that someone supposedly in charge of user interface and making "user experience better" would put in background music that he then has to talk over in a powerpoint presentation.

Edit: Also, I'm a little sad that you're being downvoted for a comparatively informative video.

I don't understand the complaint. The new interface clearly presents what options/actions are available with text names and pictures. Compared to hunting through various menus with only an action's name to determine what something does, this seems an improvment. I would agree that the Ribbon looks visually busy, but isn't it easier to scan through the Ribbon than search through multiple menus to locate an action, or to remember a option's location on the Ribbon than in a menu? Besides, a user knowledgeable of Explorer short cuts can hide the Ribbon for more usable space.

Useful information consolidated at the bottom of the new Explorer compared to the previous screenshot where the same information is either not present, or present in multiple places, feels like a good improvement as well.

With the push towards tablets and touch, doesn't the new Explorer make sense compared to interacting with the Explorer through menus, right click, or some other context sensitive input? I say this considering the Build keynote, where (I think) they mentioned that they believed in a future of even regular monitors being touch-enabled. Having main options clearly present and touch friendly works towards this.

No, it's easier to scan menus, because they are left-aligned lists of text. Ribbons have tiles of various sizes depending on measured importance of the action, and that busyness makes it harder to scan.

The lower screenshot looks like someone visually puked all over the top of it. It's got knobs and gizmos hanging out like a big mess of wires, something you'd expect to see in a stereotypical movie genius's garage, not a user-friendly UI.

(Touch UI is a complete red herring, IMO. You want different UIs for different input modalities. That's exactly the opposite of making your mouse UI look like a touch UI; that approach is just as bad as making your touch UI look like a mouse UI, which is roughly what Windows kept repeating and failing with tablets and "Windows for Pen Computing" - the failure here literally goes back decades.)

and that busyness makes it harder to scan.

I couldn't agree more. No idea why you're being downvoted, you make excellent points.

Are we missing the fact that you can click the up arrow in the top right and the ribbon disappears? That is essentially the "power user" button.

I wish they'd fix all the annoying little things they broke in Windows 7 before moving onto anything new.

In Windows XP, copying files over a network worked. In Windows 7, some "improvements" seem to result in abysmal file copying performance on a very noticeable proportion of machines/networks. The Web is full of such stories, and the best anyone has come to fixing them is basically random tweaking of low-level network parameters that no user should ever have to go anywhere near on the off chance that turning off something supposedly beneficial will fix whatever incompatibility or feedback loop is crippling performance on any given system.

On Windows XP, you could navigate folders in the tree in explorer by clicking once to expand and open a folder. In Windows 7, I have yet to find any way to avoid double-clicking or aiming for those dinking little triangles to expand each folder, which is not terribly efficient on a modern large, high-resolution display. Also, why do they hide the triangles unless you're hovering in the correct area with the mouse? Might a user not want to know which folders they can see in the tree also contain nested subfolders? Again, Windows XP did this fine, and the Windows 7 behaviour is objectively inferior in this respect.

ClassicShell fixed much of the oddities of Windows Explorer's tree view for me; it brought back the +, and makes it stay even when you're not hovering with the mouse.

There's a difference between making something more (vs less) accessible based on usage patterns, and making things impossible to do at all.

Disclaimer: I work at Microsoft, and as long as they don't touch cmd.exe or powershell.exe, the Windows org can do whatever they want to Explorer...

I think the point of the ribbon is to get rid of rightclick menus and the menu bar. This is certainly a requirement for making a touch-first UI.

That said, it will necessarily make mouse based interaction more clunky. But they are committed to touch first, so this is not negotiable.

My issue with the new explorer UI is more specifically about the choices they made within the ribbon. Why put something that provably is never used in there at all? Why not put context-aware stuff in a context aware tab like the image tab in Word (only visiblw when an image is selectad)?

Why put something that provably is never used in there at all?

It's important not to confuse "never used because it's useless" and "never used because nobody knows about it".

Of the items in the Windows 8 "Home" tab that fewer than 84% of people use, half of them appear to be new features and half are features that exist in Windows 7 but don't appear in the toolbar or any context-menu, have no keyboard shortcut, and aren't documented in the help files. People who used these features in XP could be forgiven for thinking Microsoft removed them in Vista.

Why not put context-aware stuff in a context aware tab like the image tab in Word

They do. In their blog post they have examples of context-sensitive tabs for libraries, drives, and images.

Everyone knows better than Microsoft, of course.

The most important UI element in that photo is the little hat in the upper-right corner.

Souping up the file browser in your next operating system is like putting New Balances on your next horse.

Very very good article. Shit, in fact, the fact that I hurt my foot. But, other than that, not shit, but an excellent article. Not shit at all. Congrats.

Reply on news.ycombinator.com