This seems like the right community to ask something like this without getting the angry “you shouldn’t ask such things” backlash I’ve seen hints of elsewhere: Other than Ubuntu’s now-mostly-discontinued Unity, are there Linux desktop environments out there that have tried to go beyond the bare-bones WIMP (Windows, Icons, Menus, Pointer) approach with any success?
While I absolutely understand the variety of reasons the existing desktops don’t branch too far from that trunk, it seems odd that so few current desktops would be at all unfamiliar to a Windows 95 user (taskbar, start button, system tray, desktop icons). It seems like the world must have learned something about interfaces since the Xerox Alto and I have to imagine that the Linux community would be where someone was building it.
Is there anything? I mention Unity, but even that’s not particularly different, of course.
Currently, the tasks we perform using a desktop has not changed too dramatically lately. Microsoft has a big influence, and they want everything to remain the same for as long as they can milk it. As soon as people realise that documents and files belong to the 80’s, and that there are much better ways to handle information, the dominance of Office will evaporate, and things will, thankfully, move forward.
There was some change, when mobile interfaces became popular. But due to limiting factors, such as dumbing down, e. g. typical mobile email clients having no folder and rule commands, as well as information silos, where your information is held by Google or other big outfits with ulterior aims, people still prefer to use the PC desktop for ‘accurate’ work.
Surely non business based or open source projects would be free of of wanting to keep the status quo? Yes, but they seem to attract more techies who accept poor user interfaces, and get bogged down in a text based world. And end up letting proprietary software lead the way.
So it’s good that you asked this question. I think that we should improve ‘what’ we do. And the ‘how’ will follow automatically.
People often suggest the Minority Report style where, in the film, they used hand gestures on a holographic display. Nice but the underlying tasks must move forward too. Or you will end up with a very cool interface to do some backward mundane task like editing text files, where the syntax must be exactly perfect or something won’t work.
To me, the desktop is just an interface to get to whatever application I’m needing. Once I’m in the application(s) I don’t see the desktop much at all.
As @Ianp’s comments makes me think, it would take some kind of change in how apps work to make a change in how they are accessed. (I guess that’s what he was inferring, sorry otherwise!) Otherwise, it might be like switching to a joystick instead of a steering wheel to control an automobile. Not much sense in that at this time.
Now, as my imagination is limited, I wouldn’t mind hearing thoughts as to how things could be improved, changed, or eliminated.
Yes voice commands have had a rocky ride from dictation systems to mobile phones. Where no training is required. Mine can understand my voice amazingly well. But for some people’s voices, they are a disaster. Perhaps different accents have not all been mastered yet. And in a busy office I’d not be too happy annoying my neighbours with my voice commands.
However it could be a life saver for a people who have difficulties operating a mouse or keyboard.
Gestures is another input method that have been around for a long time. They are excellent for productivity. A short setup is required. But then you can perform operations with just a small twitch of the mouse, instead of needing larger and precise mouse movements. Press the right mouse button and draw a simple shape such as a L shape over any window. It immediately closes that window or tab. Very good for closing lots of browser tabs quickly. Draw a line to the right for Forward and line to the left for Back. Draw a G and Gimp starts. Once you have got used to them you are very fast. And other people can’t even see you doing it. And you never want to go back.
You can try it out on Linux with Easystroke. Or see YouTube demonstrations.
Also for productivity, there are keyboards that have dedicated buttons for OK, Cancel, Undo, Copy, Paste etc. These also speed things up and allow 2 hand operation rather than making the mouse do all the work. I used one from 1983 to 1995. So it’s also not a new idea.
But as with all these things, if there is any tiny bit of learning required, the masses shy away from it. And things get dumbed down and removed. Because “nobody wants that”.
I does seem odd, though, that (as I mentioned) Unity seems to be the product that diverged the furthest from, basically, the Windows 95 “standard,” with more argument over whether window controls go on the upper-left or upper-right than whether this is the best design. And I do think that there’s a certain amount of (as Ian put it) “nobody wants that” conditioning, where any deviation is immediately decried as unwanted or “bloat.” Sugar obviously gets honorable mention for trying to eliminate the “desktop metaphor,” but they still ended up with the same broad components of an open-window list, space for background processes, menu of applications, and so forth; the layout is different, at least.
And I guess part of my reason for asking, besides being up for a change now that Unity (which I like) is largely discontinued, is the weird and prominent fraction of the Linux community–which doesn’t seem to be actively involved with Bad Voltage, but seems to penetrate a lot of other podcasts–that will cheer the innovation and diversity of the Linux ecosystem and wonder why more people aren’t involved on one hand, and complain about new projects like Unity (or systemd, for that matter) and rave about XFCE on the other. And it’s worth pointing out that I kept this centered on Linux because that’s what I use, but Windows and Mac OS X aren’t far from that Windows 95 Platonic ideal, either, Windows for a lot of similar reasons. So, I’m sort of looking for the complement to that community.
Voice always struck me as sort of a non-starter beyond extremely simple and disjointed tasks (as in, “what’s the weather” or “call the office”). If we think about how many times we interact with the desktop over the course of a work day, it’d be exhausting to talk that much, and would probably be a disaster for the people cursed to work in open-plan offices.
Gestures might be more interesting, especially as touchscreens penetrate the PC market.
Going back many years, many recommended Ubuntu to newbies. As it would ‘just work’ in many cases. So the switch to Unity became an exaggerated problem, as is was very poor for mouse users. Great for keyboard users. But for non IT type newbies not a great Linux first experience. Sure it was Canonical’s first attempt at creating a desktop. But I’m relieved they’ve changed. Not that Gnome is especially good. But at least Canonical can now transfer some of the Unity (keyboard) goodies to their distro, hopefully. And keep everybody happy.
The desktop background of most desktop environments is an oddity. It’s interesting to see Gnome dropping desktop icons. Sad for those that build processes around that. This is a symptom that they didn’t have a decent alternative. Anyway this desktop space could be used in a very different and more productive way.
This is the road that got us to Electron applications. The fact that you don’t see the Desktop does not mean that you don’t need it, because the desktop metaphore encompasses things like your address book, the clipboard, time tracking, local storage, most-used commands and so on. If you do away with that metaphore, you have to reinvent or reposition all these essential features. Sooner or later somebody will want a “window” or “graphical shortcut” or “widget” for these things, to be placed on a “common area”… very quickly, you are back to a desktop.
I think search based navigation qualifies and I believe that was in Ubuntu / Unity first at least that was the first place I saw it and since then it’s made it’s way into Windows.
Also, I think Linux had multiple desktops first and that’s in OSX and Windows now I think.
If you’ve got a good idea, throw it out there. I hope we get some sort of 3d navigation soon. Seems a no brainer once VR becomes commonplace but it wouldn’t have to be tied to VR. There have been 3d pointers such as Space Pilots available for regular computers for a while. It would be nice to see some creativity put to use here.
I would argue that the default GNOME 3 user experience, default Elementary OS desktop user experience, and a few others diverge more from the Windows 95-style user interface a bit more than Unity does. I’m not saying they’re objectively or subjectively better, just further away.
I’ve been using computers since 1995, so I’m most comfortable with terminals and the Windows 95 style user interface. I’m not asserting that Windows 95 is a high mark in user interface, just that after more than 20 years I’m so accustomed to it that anything else feels awkward even if I use it for weeks. I ran Elementary OS for two years before I switched (edit: switched back) to MATE desktop configured in a 95-esque style.
@toyg, I don’t think Electron has anything to do with that. The success of Electron is that it’s cross-platform but lets the user work mostly in NodeJS + HTML + CSS + Javascript, so you don’t have to know as much C, C++, Objective C, or similar to build something that acts like a native application. But if someone built a new Skype client or new version of the Atom editor in pure C++ it wouldn’t make any difference in the end user desktop experience. The buttons, windows, fonts, etc… interact the same on screen regardless of how they’re managed behind the scenes.
I was not implying that I think the desktop is irrelevant. And the things you listed that the desktop is involved with is something I take for granted. It’s like the 20 miles of highway that I use to get to work. It’s not something that I’m thinking of all day while at work. But take it away…
I don’t think the “native” matters one bit - people like Electron, as you say, because they can build interfaces in HTML rather than being forced to learn about the desktop infrastructure (widgets, services and so on). Nobody cares about looking “native” with Electron, they certainly don’t look native to anyone, the naming is simply a masterpiece of doublethink-based marketing. People use Electron because they are then free to write interfaces as they like, breaking all rules and guidelines that desktop metaphores have accrued over the years - with little thought for why those rules were introduced and what sort of service they enable (and more depressingly, simply to reuse code already deployed on websites). And no, all those buttons and windows do not behave the same in Electron vs real desktop tech: you lose most accessibility features, including things like full support for entering Unicode characters, right-click features, and so on and so forth. But you can have “material” buttons and that’s all that matters to developers.
@toyg
I agree with your comment. I misunderstood the intent of your previous post, sorry.
In defense of the use of Electron, it has good cross-platform support and it allows people whose developer experience is mostly in web applications build a desktop application much more quickly than if they had to kick all or most of their HTML5/Javascript expertise to the curb and build something in Qt or equivalent. I’m not saying it’s ideal or good, just understandable.
That said, as a web developer myself I see Electron as a last resort. If I can build the features my application needs using web technologies then I’d try to go for an offline first webapp (Progressive Web App). If that’s not sufficient I would just go native.
Sadly they prioritised the search-based navigation to the detriment of the mouse users. So it was a load clicks to get where you are going, compared to other desktop environments that have menu with direct access to favourite apps, as well as the app categories. So loads of people were really frustrated by Unity.
