Same interface? [LWN.net]

Same interface?

Posted Aug 24, 2011 13:58 UTC (Wed) by Kluge (subscriber, #2881)
Parent article: The year of the Linux tablet?

"There is clear value in having the same interface - and the same applications - on both types of device."

I disagree. Tablets seem to be optimized for content consumption, not creation. And a touch interface probably can't have the necessary precision for a laptop OS. While the OS on PCs and tablets should be compatible, I see no reason that they should offer exactly the same interface.

to post comments

Same interface?

Posted Aug 24, 2011 14:14 UTC (Wed) by corbet (editor, #1) [Link] (5 responses)

Um...but that paragraph was talking about handsets, not PCs...?

Same interface?

Posted Aug 24, 2011 17:20 UTC (Wed) by Chocrates (subscriber, #67068) [Link] (4 responses)

I still think its a valid comment. I use my phone for a certain set of tasks. I use my PC for others. And if i had a tablet id use it for another set. While some of these overlap, a web browser for one, I would greatly prefer an optimized interface for the hardware, with simply some way to share data between. Preferably with compatible document types if its applicable. I don't need my tablet to look like a big phone.

Same interface?

Posted Aug 25, 2011 8:41 UTC (Thu) by sebas (guest, #51660) [Link] (3 responses)

Exactly. :-)

You want your devices to work consistently, but take advantage of the special hardware you're using. For that reason, I think one interface for tablets and laptops (as an example) is suboptimal, because you end up compromising on both sides. Some of the complaints about gnome-shell and Unity are I think symptomatic for these compromises.

In Plasma Active, we're sharing components and infrastructure across devices, but create specialized interfaces suitable for the type of device (screen, input methods, etc.), and the use-cases that go along with it to get the best user experience out of your hardware. The end result will (hopefully :-)) be a system that works consistently across devices, integrates well with your other form-factors and is easily adaptable to new types of devices.

And yes, the plan is to also support smartphones at some point, but given the "closedness" of this segment of the device spectrum, we are concentrating on tablets first -- a lot of that work also benefits an eventual phone interface at some point in the future, and not to forget the desktop and netbook interfaces.

Same interface?

Posted Aug 31, 2011 15:48 UTC (Wed) by nix (subscriber, #2304) [Link] (2 responses)

Yes yes indeed. It's sensible to have a menu locked to the top of the screen on a handheld, just as it's sensible to have gestural interfaces on a handheld. The first is broken by a big screen, the second by machines where the screen is not held in your hand. Thus, different interfaces are absolutely required.

It's a shame the GNOME people didn't notice this. You can't just give the two things the same UI without frustrating a lot of people.

Same interface?

Posted Aug 31, 2011 17:58 UTC (Wed) by tshow (subscriber, #6411) [Link] (1 responses)

Top of the screen menu is also particularly lousy with multihead setups.

Same interface?

Posted Aug 31, 2011 22:03 UTC (Wed) by nix (subscriber, #2304) [Link]

Quite so. Top of which screen? Sod's Law says, whichever screen you weren't using. (Not that other programs are exactly blameless in this area: you'd go green if you saw the appalling hack I'm using to make libsdl-1.2 work on my dual-monitor setup.)

Same interface?

Posted Aug 25, 2011 7:56 UTC (Thu) by ebassi (subscriber, #54855) [Link] (5 responses)

I disagree. Tablets seem to be optimized for content consumption, not creation.

that's what people in the industry kept saying when the ipad came out, and I used to believe in that. "it's for consumption on the sofa". which is definitely true, media consumption on a tablet is an incredible experience; people should just try it it's awesome.

this stuff, though, was also said of the personal computer thirty years ago (except not with the "consumption/creation" terms): "it's too small, not powerful enough, not connected enough to do real work on it. at most you can play with some game".

let's just remember that the tablet market doesn't really exist: right now, it's the ipad market; and Apple has released media creation applications on it, with different user interaction models. others will start copying them soon, and miss the point by about a thousand miles: it's a new device, and it's finding its footing, and exposing new interaction models along the way. copying Apple, or trying to port desktop applications as they are to it it's not going to work.

you cannot write an Excel sheet on a tablet. only Microsoft thinks you can, but they also thought that the biggest appeal for Windows Mobile 6 was being able to edit a spreadsheet on a QVGA 2.4" phone screen (according to their adverts), so they don't really count.

the fact that you can't edit a spreadsheet on a tablet could mean that you just simply won't be able to do it, or (and I believe it's more likely) that the spreadsheet user interaction model has to be changed to take into account a touchscreen and direct manipulation of data. what we call a spreadsheet on the desktop might be a completely different experience on a tablet and you could still transparently migrate documents from one device to the other.

so I don't believe any more to the "it's a media consumption device" mantra. it's just what people say before others come up with something really cool.

Same interface?

Posted Aug 25, 2011 14:48 UTC (Thu) by tshow (subscriber, #6411) [Link] (4 responses)

> so I don't believe any more to the "it's a media consumption device" mantra. it's just what people say before others come up with something really cool.

Speaking as someone who's done some iphone/ipad development, it really is a media consumption device first. The problems:

- screen keyboards suck for extended use
- bluetooth keyboards/mice are a negligible presence
- there are few hardware buttons, and you may not have access to them
- touch interaction is weak

The weakness of touch interaction is a particular problem. There are three main things that make the touch interaction suboptimal:

First, you have no context. There is "touch down" and nothing else. No "left click" vs. "right click", no pressure sensitivity, no angle detection, no "hover", no cursor position when there isn't a finger down. The only context you can have is multitouch, but that only improves things slightly, and it causes its own problems:

- two points approach each other, meet, and move back apart. Did they pass through each other, or bounce off each other? There's no way for the hardware to know.

- two points approach each other, meet, and then one of them disappears. Which one disappeared? There's no way for the hardware to know.

- you can't easily have multiple gestures occurring at the same time unless they're happening in compartmentalized parts of the screen that you've pre-defined

Second, your finger obscures whatever you're touching. Unlike an indirect device like a mouse or trackpad, with a touchscreen you can never see the point on the screen that you're activating.

Third, touch coordinates are very loose. Even if you touch the screen very gently, there's a 5mm or so diameter circle that's touched, and the touch coordinate the OS reports could be anywhere within (or near!) that point. On a first generation ipad, I've had touch coordinates come in completely offscreen (ie: out in negative coordinates, somewhere on the bezel) when touching a spot that was near the edge of the screen.

Things get better if you have a stylus, but not enough people have a stylus that you can actually count on it.

There's a reason the gestures are all large movements.

So, you've got sloppy, imprecise "mouse" input, weak text input, and some vaguely useful (but mostly unsuitable for content creation) inputs like the accelerometers.

You *can* produce content with current tablets; it's like producing fine art with crayons. Feasible for the talented, but a chore and done mostly for the novelty of the process. Much more of a chore than doing the same thing with (say) a laptop. Perhaps future tablets will tighten up their input systems, but until that happens tablets are going to remain mostly unsuitable for all but the most trivial of content generation; the kind where poking a few onscreen buttons does all the work.

Same interface?

Posted Aug 25, 2011 16:44 UTC (Thu) by dmarti (subscriber, #11625) [Link] (2 responses)

A tablet can offer a good interface that mimics single-function music devices: iPad Music: How To Make Drum & Bass w/Korg iMS-20 App. (If I were in the business of making gig bags for guitars or other instruments, I'd put an iPad-sized pocket on the outside.)

Same interface?

Posted Aug 25, 2011 17:44 UTC (Thu) by tshow (subscriber, #6411) [Link] (1 responses)

I think that's where tablets and ipods and smartphones are going to be for the next few years, at least; a consolidation of all the little gadgets and doohickeys (phone, calculator, camera, music player, portable game machine, pager, password vault, video player, wristwatch, GPS, portable storage, map, notebook, wallet, keys, ID...) you used to have to carry around separately.

That's why Apple's stuff is all aggressively single-tasking; their base assumption is that the device is a swiss army knife of gadgets, and like a swiss army knife it's only good for one thing at a time, and it only has to be ok at any given thing as long as it does everything passably.

We're really feeling this in the game industry, because the iphone seems to be gutting the handheld console market. Why carry around a 3DS or a PSP when you've already got your iphone?

One of the places I think free software has an opportunity in this space is with more technical stuff; if a tablet plus a small arduino dongle could give you (say) a credible portable high-res oscilloscope, for instance, there's a set of people who would find that invaluable. Or something you could hook up to weather monitoring hardware. Or other arbitrary "unsanctioned" devices.

Same interface?

Posted Aug 25, 2011 18:01 UTC (Thu) by dmarti (subscriber, #11625) [Link]

+1 Insightful.

For a lot of devices, it's cheaper to put a mini web server on it than to design the needed physical controls and display. You could sell hardware, publish the REST API to control it, and optionally sponsor some of the tablet app projects that use the API. Test equipment and weather stations are two good ones, also lab equipment such as PCR machines, high-end printers, A/V gear, etc.

Same interface?

Posted Aug 30, 2011 15:01 UTC (Tue) by carlleigh (guest, #79395) [Link]

That is why when I'm out on the road working out of my office or home I want a touch screen tablet. I'd also like it to support 3G and voip or mobile phone.

At home I want the same device to hook up to my 24 inch display either by HDMI or micro USB. No Bluetooth I like speed and efficiency. Bluetooth requires several transceivers and I'd like to save the power when I'm on the road for a more energy efficient micro USB keyboard.

I want my tablet to do everything. I've got a spreed sheet program which works reasonably well. I don't need Microsoft...........

Nobody really needs Microsoft. (Tablets must really have them shaking!)

CPU and GPU Processors for small devices will improve quickly. Batteries will get small and cheaper. What do I need the old workstation for when my tablet frankly can do it all. I love empty pockets one device to rule it all.

5 billion of the worlds people really can afford all the eye candy. Powerful Desktop. Tablet. Mobile Phone. Land Line for data. etc .etc. etc. But they can afford what will be coming in the tablet space and wireless networking.