76

After Plasma-Desktop, Plasma-Netbook, let's have a Plasma-Tablet.

Even if tablets have existed since long ago, now that Apple has showed us his iPad, it is most likely that kind of computer is going to expand massively in the next years.
We developed a Plasma for the regular PCs, and a Plasma for the netbooks (which are obviously not going to win the war of the medium devices, if they have to fight the tablet).

As there is already some free tablets, I think KDE should develop a new Plasma layout for that kind of devices, more touchscreen-oriented, including a virtual keyboard, and all that stuff.

I was thinking about a device in particular : the Touchbook was developped by a very wised french from the Allways Innovative company. The revolution is that the Touchbook is both a netbook and a tablet (you can take the keyboard off and use it as a tablet).
Using also an accelerometer (the same kind of stuff you find in a iPad / iPhone and which allow you to interact with the tablet position in space (holded horizontally or vertically), it is most likely the best concurrent for the iPad, which is quite a failure anyway, but will probably expand as the iPhone did, using the Apple fanboy.

The Touchbook could be one of the target of an eventual plasma-tablet release.
Then would come the iFreeTablet, which will be presented tomorrow.
These devices deserve a full-featured KDE desktop, designed for the tablets. And I think KDE deserves to get released on theses devices also : in the name of the KDE domination over the world, KDE needs to be accessible via the tablets, as well as on the netbooks and the regular PCs.

I think that the plasma people have this in mind already. Plasma-netbook is still at its early stages and will be easily adaptable to a tablet device. Wait until after tokamak4, I am sure they will present something on this topic.

BTW, in my optinion the notion ink adam will be a good contender. Although they focus on Android, it remains a open device where we can easily install kde

I like this idea, although it would need to be taken farther to make it effective.

Plasma, already, is a great interface for this. We have keyboard plasmoids, plasmoids with sliders, pretty much everything about plasma is very touch-friendly. I have no doubts that plasma could easily make an excellent touch interface. I personally am already imagining a spec for the interface.

But we would need to re-make some things to be touch-friendly, and we would need to implement multitouch in X (projects exist, just getting them more mainstream)


If Plasma-Tablet was made, I would make the argument that traditional applications should be completly removed in favor of a completly plasma-powered interface. If that is done, then we would make massive improvements to the gethotnewstuff interface for plasma downloads, and possibly create a plasma-tablet-app type class of plasmoid specifically for multitouch-enabled standalone tablet applications.

It would be quite a bit of work, but this work I am heavily in favor of, and I would be a member of any team with this goal.

I disagree that other application types should be "gotten rid of" - one of KDE's huge strong points, Krita, would be best suited for a tablet interface. I think, however, that perhaps the widget classes in kdelibs could be edited for more finger-friendly buttons in applications.

I recently tried out the "Skulpture" widget style - I discovered that you can change the "padding" of widgets. For example you could make buttons or scrollbars much wider. I played around with it for a while and yes - it's definitely touch friendly given the proper widgets!

So yeah I'm with Madman, no need to throw away the apps.

My opinion (that's all it is) is just because current apps aren't touch friendly. Granted, tablet-pen friendly, they are. But existing programs tend to pack a lot of functionality into small areas you just can't effectively finger around in.

By the time we pad the buttons, re-size the widgets, modify the scroll bars and then hunt down the bugs or hideous deficiencies after doing this - it would almost be more worthwhile to choose several of the programs we would use most on a tablet and re-write their interfaces from scratch. Various KDE programs would also get a bad rep because they might not be optimized for tablet use.

If we did simply refresh the vanilla interface for tablet use, we aren't getting the creative potential out of the system. Sure, it would work, probably reasonably well, but if you look at Apples iPad we would be making an interface that feels 5 years behind the curve.

If we use plasma, since it's obviously KDEs most durable asset, a complete interface overhaul would be more worthwhile as we could better use plasma and its features for much more interactive and kinetic interfaces. And by creating a unique breed of plasma-programs, we can separate desktop and multitouch paradigms. Desktop apps don't get bloated with multi-touch features, and tablet-only programs get the most out of their interfaces without being held back by mouse-only users.

How many users are using tablet and touch-enabled devices?
How do they use them?

I think this change could be important enough for checking with users the use cases and usability tests. I've been using a tablet and a netbook in conjuction with a desktop for 4 years (using MS Windows, mind) and, having used KDE4 as well, I think that not every app is useful on a tablet or netbook in any case. So there's no need to completely overhaul all of Plasma to make it touch friendly. Let's begin with the relevant parts first, these which users use and would actually use in these use cases.

Kver wrote:By the time we pad the buttons, re-size the widgets, modify the scroll bars and then hunt down the bugs or hideous deficiencies after doing this - it would almost be more worthwhile to choose several of the programs we would use most on a tablet and re-write their interfaces from scratch. Various KDE programs would also get a bad rep because they might not be optimized for tablet use.

What? You talk as if this would be done individually for every single button everywhere - it wouldn't. It would be done once in the widget theme.

I guess I wasn't clear - may bad. My line of thinking was more that, after making those changes, the line of programs that we have wouldn't "fit" properly. Currently, many of our programs have interfaces with tightly grouped buttons, when we modify the style enough to make a usable touch variation of oxygen - we would see massive growth pains in the existing interface.

For example, if we have a toolbox with 18 rows of buttons, even increasing the height of buttons 10px and padding them an extra 2px, we quickly add 432px to the height of that toolbox. Coupled with other interface elements such as pallets, tool options, etc, we very quickly out-grow the screen.

That being said, if we did decide to modify the oxygen style, we would need to go though all of the individal programs and try to find new ways of re-compressing the interface to account for the lost real-estate. Very quickly we get to the point of maintaining tablet versions of programs anyways - and the interface would still feel half-baked.

Aswell, a traditional interface can't really make full use of multi-touch capabilities. Sure, easily we could have pinch-and-zoom or the like, but there is a whole new set of widgets that are simply better for touch-screen. Stacks, for example, are not found on traditional apps, but make several appearances throughout the iPad. If you look at the iPad version of iWork, they did not just scale up the buttons. If you look carefully, you'd be hard-pressed to find more than 20 active buttons on any given screenshot of the interface. Instead, they hid much more of the interface in favor of more dialouges and liquid controls.

If you look at Krita, it has over 33 buttons in the toolbox alone. 18 buttons on the toolbars, and various buttons for the tools themselves. If every button was to get 5px of padding or scaling there would be no room left to work on the image itself - and the buttons wouldn't even be that much bigger.

If I had my way with "krita-tablet", the interface would fill the screen with the image, and use drawers that group tools, and selecting a tool would also access it's options. Only one toolbar out of the entire interface would be permanent, and even that could slide back. Maybe one more drawer would maybe have access to some of the tools. If I had the chance to make the dedicated interface, I would set up a way you could customize drawers and popups with entire chunks of interface - the same way we currently customize toolbars with buttons.

I guess, long story short: If we just scale up the existing interface, it will become too bloated for actual use. And if we customize programs for the scaled interface, were maintaining tablet versions of the programs anyways. And if we maintain tablet versions of desktop interfaces, we lose out on many of the more tablet-centric widgets. The best way, I beleive, is to simply build a tablet-specialized interface. I say we build it on plasma, because plasma is flexible, and plasmoids can very easily be embedded into eachother, drawers and panels.

Kver is right, we can't use the actual KDE apps in a Tablet version of KDE.

Luckly, Plasma is extending and getting much much love from all the devlopers. For instance, we will get somthing like a "Kopete-Plasma" which will probably be great for the touch-screen device.
Plasmediacenter, growing slowly but certainly, will solves the problems for all the multimedia tasks that the tablet may be used for.
And as we have seen on the planet, KDE is currently working on kdewebkit in order to optimize the touchscreen uses : it's made for plasma-mobile right now, but may be easily extented to a plasma-tablet.

Once all thoses tasks will be covered by Plasma and X2, we may need a portage of KOffice to Plasma... but that's not likely to happen.

Kver wrote: The best way, I beleive, is to simply build a tablet-specialized interface. I say we build it on plasma, because plasma is flexible, and plasmoids can very easily be embedded into eachother, drawers and panels.

This posts were written in 2010. It's now 2019. KDE has matured considerably. I like the performance on Kubuntu 18.10. The only issue is auto screen rotation on a 2 in laptop (tablet). Elementary OS does it. The screen can be rotated manually. Not sure why it can't be done automatically. Smart phones do it all the time. WIndows (God Forbid) does it easily.