Jeff Duntemann's Contrapositive Diary Rotating Header Image

Minty Failness

I gave it a good shot and I tried, honestly I did. But Canonical’s Unity UI simply doesn’t work for me. It’s obvious that Canonical is trying to create a single UI that will serve end-user computing from top to bottom. It’s just as obvious to me (now that I’ve had six weeks or so to play around with a Droid X2) that there is no single “end-user computing” anymore. Desktops are fundamentally different from smartphones, or anything else (tablets, possibly; we’ll see) that is primarily tap-and-consume. I’m having no trouble working the Android UI on my phone, and Android habits don’t intrude on my desktop synapses. I’m not confused or in any way slowed down by the differences between the two, no more than I’m confused about the differences between a shovel and a rake.

So if Unity is all I get under Ubuntu, Ubuntu has to go. Others seem to agree with me, and at times the discussion gets disturbingly violent. Online I’m seeing that huge numbers of people are fleeing Ubuntu for Linux Mint, which I’d barely heard of a year ago. I have to smile a little bit, because Linux Mint is Ubuntu, basically pulled back to a variation on the GNOME 2.3 interface. The upcoming release (Mint 12) will move to GNOME 3, which worries me a little (I like GNOME 2) but I’ve seen word that Mint 12 will allow users to have something very like the old UI–which is precisely what Canonical did not do with Ubuntu and Unity. It was Unity or the highway, and boy, it’s bumper-to-bumper out there.

There’s an enormous issue of why we’re suddenly tossing older and much-loved UIs away without nary a glance over our shoulders, when there’s no compelling reason to adopt one of the new models. Programmers like to create Shiny New Stuff, fersure. I in turn don’t like to change the way I interact with the machine I use, unless such changes make me a lot more effective. So far, the costs in relearning ordinary tasks far outweigh the fairly paltry benefits for me.

I’ll take up that issue eventually. In the meantime, I’ve hit the highway, and installed Linux Mint 11 Katya in its own partition here on the quad core. The OS looks great and works the way I’m used to working. I have some minor quibbles, like the failure of the Software Manager to tell me when it’s done installing something. Ubuntu does this well, but Mint installs and gives no sign. This was critical when I installed WINE, since (because WINE is not an app, strictly speaking) it’s tricky to determine if WINE was fully and correctly installed. Because running Software Manager again and selecting WINE still indicates “not installed,” I think there’s something wrong.

Small stuff. The big deal is that Mint doesn’t work well with the integrated graphics on my EVGA NForce e-7150/630i Core 2 Quad motherboard. The default graphics drivers worked, but looked clunky and don’t support effects. Installing the recommended proprietary NVIDIA drivers produced weird graphics failures, including windows refusing to render once they’re over a certain size. (Some windows would not render at all, and simply remained blank and white even when first instantiated.) Using the supposedly experimental NVIDIA 173 drivers worked better, but still fails on certain apps, especially Stellarium, which worked exactly once and then comes up with a blank, black window every time. I’m not willing to give up Stellarium, so at this point Linux Mint is on hold while I wait for Mint 12 Lisa.

Linux Mint has supposedly become the 4th most popular OS on the planet. It’ll be interesting to see if that continues to be the case once they cut in the mandatory GNOME 3 upgrade. I’ll give GNOME 3 the same consideration I gave Unity, but I’m also looking closely at the Xfce UI and Xubuntu. It’s going to be an interesting year in the Linux world. I’m keeping all my old Linux installer .iso files, trust me.


  1. Tom R. says:

    What’s your take on KDE Jeff? I have been using Mepis Linux off and on and have been happy with it and KDE until the newest release. I was not too happy at first with the change that came with Mepis 11, but I have learned it is not as bad as I first thought. At least you can still “get under the hood” and configure things pretty much the way you want.

    I am still not sure why EVERY distribution has to change desktops every couple of releases. I recently read one comment where the user decided to do away with the desktop idea entirely and just use their window manager of choice. I did that quite by accident years ago, but never thought of staying there. Maybe it is time to go back!

    1. My view of the current KDE product is very much what Chuck says below: It slows the machine down, and the V4 changes basically obsoleted all the skills I had built for the earlier versions, which actually worked a lot like Windows. I poked at KDE4, had a hideous experience with the first releases, and basically dropped it. I know it’s better now, and may have to take another look now that GNOME is going the same way and Ubuntu is going to Unity.

      I also agree with Chuck on most of the rest he says in his very good (long) comment. Do read the whole thing.

      1. Forgot to mention this: There’s a package available providing KDE 3.5 for modern versions of Debian and Ubuntu. It’s the Trinity Desktop:

        I haven’t tried it yet but it’s rising fast on my to-be-researched list, given that I had more Linux Mint failures last night. Will describe as time allows.

  2. Chuck Waggoner says:

    So far, my explorations into Linux are limited to audio and video applications for a couple of part-time projects I am involved with. Earlier in the year I was very frustrated with a couple of days trying to cope with Unity. I finally dropped back to Ubuntu 10.04-2 LTS to get Gnome back. (By the way, I have found KDE to be a CPU hog, and lacking interoperability with audio/video software, as it does not adhere to accepted programming standards, and thus some software will not function properly with it. Much too bleeding edge.)

    I read that entire discussion link between Shuttleworth and his community guy, including all comments following. Tom R above indicates the underlying point in all of these discussions–a point which is totally missing from these interchanges between “the community” and the paid developers: “power users” are a tiny fraction of the people who use desktop computers these days. Regular computer users–which is most of us–have developed ways of accomplishing their daily work that are efficient for them, across a broad range of software and tasks. Then, along come developers who want to pull the rug from under everything! Unity is a completely different way of working. So were the Office 2007 menu changes. This affects productivity across EVERY SINGLE TASK performed by computer users (which is everybody with a desk job, every student, and computer tasks in most homes).

    Working stiffs who have to turn out 50 to 100 tasks a day, do not have the time to learn TOTALLY new procedures for navigating to documents, opening them in half-a-dozen different software programs, figuring out how to modify them with completely new menu structures, then new procedures for printing, saving, and filing them again–or sending them on to other people.

    Such a ‘development’ is a catastrophe to workers. I hear that Win8 is going to be such a complete paradigm shift. I can tell you from having worked with several large international organizations with hundreds of thousands of computer seats, that is just not going to fly. They work their people hard, and presenting them with the need to do everything differently is going to be rejected very quickly with a brick wall and a sledge hammer!

    Personally, I do not care what will satisfy IT people, programmers, and “power users” which is all that discussion concerned itself with. One guy said that Unity is capable of tweaks that make it more usable, but I know that the time I have spent over the years in shaping my methods and operating habits with computers is time I no longer have at my daily disposal. I MUST turn out more work than ever (typical of most people), and the time I spent customizing my OS interface is time I will never again have at my disposal in the foreseeable future. Changing operating paradigms is an absolute impossibility for most working people these days, and I will strongly resist having to do so. It appears Shuttleworth is totally and completely blind to–or truly ignorant of–what most people use computers for. All the stuff they used to do with paper, typewriters, and telephones, is now done entirely on computers. If he does not perceive the productivity business demands of workers today, he and Canonical will be history in very short order.

    Whatever the future of computers is, success is not going to be built on some ‘smart’ guy’s vision of re-inventing the wheel. It will be constructed around small things take minutes a day off each task workers have to do–without the need for ANY retraining.

    Here is a typical example of how guys like Shuttleworth–and MS in particular–are actually shutting people out of computing. Earlier in the year, my father in-law and his wife had to buy a new laptop for her, as the old one died. New laptop has Win7 on it. Neither of them can figure out how to do anything with that computer. They are approaching 90 years of age. Choosing “Classic Windows” on the desktop apparently only changes the colors and graphic shapes; it does not return the computer to the Win2k/XP menus and ways of doing things. Consequently, I have not received an email from her in over 8 months. She just stopped using the computer. I used to get several emails a week from her, and so did all of her close relatives and friends. This is a very sad development, IMO. But it will not be tolerated by business.

    1. Boy. I wish this had been published on a major site or in an IT magazine, because it encapsulates a lot of what I hear from non-developer practitioners these days. Programmers like to have projects to tinker with, which is where most of this trouble comes from. (I understand that restlessness completely.) Alas for them, our UI concepts are mature, proven, and represent countless hours of well-honed skill on the user side of the screen. As you say, improvements need to be bottom-up, and incremental. Top-down basically throws molasses in the works, and makes people less productive. Occasional computer users like your in-laws may never build back the skills and confidence they once had and may just feel that the boat has left with them still on the dock.

      This is completely unnecessary. It’s why, when (say) someone at our church needs a computer, I buy them an XP-equipped used Dell desktop on eBay for $100 rather than recommend a new $500 Best Buy box with Windows 7 and a crushing load of crapware. I set up an LUA and with relatively little tweaking, they’re back in business. Social engineering remains a problem on the malware side, but at least they don’t have to learn how to compute all over again.

      1. Tom R. says:

        Chuck Waggoner’s great comments and your reply, Jeff, got me to thinking back to the analogies used in the early days of Personal Computing when every computer and every piece of software worked differently. Back then it was said that computers should be like cars where one does not have to relearn where every control it and how every function works. The gas pedal is on the right, the brake is to it’s left and the clutch. if any, was to the left of that. Oh, there was some variation, but it was relatively minor and if one could drive a car with an automatic transmission you could pretty much drive any car with one. Computers did evolve into the common desktop metaphor that was really invented by Xerox and then popularized by Apple and later Microsoft. Eventually Unix followed suit with windowing systems that were similar.

        Having said that it now appears that the concept of a uniform way of doing anything is going by the boards in an attempt to “differentiate” each product from any other similar product. Even cars have gotten more different in the way they work. My newest vehicle is over 10 years old now, and the oldest is well over 20! I have to relearn where things are like the headlight switch, the windshield wiper controls and a number of other functions when driving the one I use the least.

        In even newer cars it is sometimes hard to know how to start them and even worse how to shut off the engine!

        I don’t know what it is exactly, but I think this common theme of change for change sake is being pushed into everything these days. As an engineer I would have to blame it on marketing!

        I do appreciate Chuck’s point of view in particular since for about 20 years I designed applications that ran on a large distributed mainframe system with thousands of users. The philosophy was that a change that could save a few seconds for each of those users was worth a lot of development time, but any change that required a significant re-training of those users was not something to be considered lightly.

  3. Bruce C. Baker says:

    UI design is approaching the point where developers have nothing left to tweak except the size and shape of the tail fins.

    1. If only. The problem is that they’re not just tweaking the tail fins. They’re playing with the whole damned body, including the dashboard. Total makeovers may be fun for programmers, but the rest of us have to learn the whole damned thing all over. It makes me sound like an old guy, but I’m not sure that incremental changes in UIs are worth the slowdown they cause to users, unless they increase productivity radicaly. And I have yet to see one that does.

      This issue reaffirms my long-held opinion that desktop software is mature. We have things that work, and have not yet discovered a mechanism that makes them work so much better that the old ways need to be abandoned, at the cost of much time and hair-tearing.

      1. MGalloway says:

        Speaking of UI changes…I keep wondering what’s going to happen when Microsoft releases their Kinect for Windows SDK (after the beta) and what sort of changes that have on UIs. I can foresee lots of useful applications but also the potential for a lot of interaction disasters.

  4. Paul says:

    I couldn’t agree more about the new UI coming out of Ubuntu/Gnome, and that is finally what prompted me to finally jump ship as well. It goes back a bit to what Microsoft has done in different areas, dumbing down the UI to the point that it hurts productivity.

    Moving off of Ubuntu earlier this past summer I finally landed on Sabayon for two specific reasons; it lets me choose my desktop, and it is also on a rolling release schedule. My current desktop is LXDE, though XFCE was also a strong contender for me as well. The rolling release schedule has been nice so far in that I get to stay with the latest versions of installed software, not just getting security updates only.

  5. Ray says:

    For teaching purposes, it might be better to go with a stripped-down minidistribution where Lazarus is already installed. Something like Knoppix or Puppy Linux. it’s too much to ask a newbie to go through the hassle of setting it up the way is now.

  6. Craig Anderson says:

    I’m a software developer and so not mainstream. I just switched my notebook workstation from booting Windows 7 to Lubuntu 12.10. I prefer LXDE to Unity and Lubuntu runs nicely on my older computers. I work in a Windows server shop and support for Windows shares and RDP to Windows server works well.

Leave a Reply

Your email address will not be published. Required fields are marked *