Saturday, March 30, 2013

Standardization as a Road to More Choice

by Guest Writer Michael Glasser

If you are reading this article from a Desktop Linux system take a minute and look at a few programs on your system: maybe your web browser, your primary word processor, your email program, and whatever else you use often. Look at the save and print dialogs, the term used to Quit (or Exit) a program, the terms used for Options or Settings and where such options are located and what hot keys are used. Chances are you will find much inconsistency in the programs you look at. Even within single programs there is often a great deal of inconsistency. 

Is this a problem? Many users say it is not; they believe they “get used to” each program and speak of how it is not hard to figure out the multiple styles found on their systems. Even if they do not realize it, though, inconsistencies do lead to problems – they lead to reduced productivity and efficiency and increase in user errors. This is backed up by HCI/UI research, is accepted by pretty much every relevant expert, and such ideas have been expressed by the teams that produce KDE, Gnome, Ubuntu, FireFox, and many other open source projects. Having a distro that works as a unified system is important if you care about the work you do.

What I would like to see – what would benefit desktop Linux users – would be a way for distro developers and users to get more choice than they do now. Imagine if Ubuntu had  minimalistic styles for their dialogs designed for great ease of use while PCLOS had more robust dialogs that allowed for renaming and deleting of files from the dialogs. Novice users would be able to trust that “settings” for their system were always in the same place and could look for the same term; more advanced users would know the hot key to get to the same feature. If the users did not like the choices made by the distro developers, they could swap these things out on their own, and their choices would affect their entire system because developers had accepted whatever standards KDE, Gnome, and others had agreed on.

More choice. Greater productivity. Improved efficiency. Reduced Errors. Nothing is lost.

I am not going to pretend this would be easy or perfect. Nor would I want this forced on distro managers or users or developers (and there are good reasons in each of those groups why they might sometimes wish to go outside of the standards). It should be a choice. 

Having such choice is something I have been seeking and predicting for many years, and looking at Kubuntu or PCLOS (or many other distros) of even three or four years ago compared with what they offer today we see that they do act more and more like complete systems and not like a bunch of unrelated software jumbled together... what a customer of mine said felt like a system held together with duct tape that he was afraid would fall apart at any moment (even though he never had a system crash and had no more application crashes than on his old OS, maybe even fewer).

The open source community is getting better and better at allowing users to have distros act like unified systems. There is, however, much room for improvement. I am not a programmer. I would love to hear ideas on how to help move the open source ecosystem to better allow for this increase in choice.

-- Michael Glasser

Enhanced by Zemanta


  1. Glasser aka Snit does nothing without some type of ulterior motive. He's probably hoping for people to leave comments so he can engage in his usual nonsense as he does on Usenet.

  2. Well aware of who he is. Which is why I invited him to write. I'd be a hippocrit if I hadn't done so.

  3. My "motive" was to share what I did, above... and if people have questions to answer them in the comments. If you are open to it I would be willing to write other articles. I have no desire for the COLA "advocates" to bring their flame wars here or anywhere else - frankly I wish they would stop following me around trolling me. Enough.

  4. My main issue with this article is that the author is pushing the idea that an entire distro should somehome represent a system. This is simply an unrealistic goal. A distro is and always will be a huge collection of third party programs and suites.

    The *actual* "systems" are the desktop environments like Gnome and KDE, each with their own set of rules and ways of doing things. Applications developed with KDE in mind will fit in the KDE evironment; applications developed with Gnome in mind will fit in the Gnome environment. It's a choice in itself, and a valid one indeed.

    In fact, if the Gnome and KDE teams were to agree on a set of rules to make programs fit well in either environment, one would have to take away the choice for KDE and Gnome to be totally different environments. That's simply a fact, as you cannot have one without the other.

    The only thing distro developers should take care of is to offer a sane set of default applications in their default install (if they have one - Gentoo for example doesn't) and then let the user decide how they exapand/customize their installation.

    As for the "improvements" mentioned in this article, I suspect this observation is based on anecdotal evidence. In fact, Gnome and KDE applications have never been more different as currently. With the rise of GTK3 even a lot of cosmetic hacks (such as oxygen-gtk) no longer work 100 % properly (if they ever did - after all they are just cosmetic hacks).

    As a final remark I would like to say that the introduction of the article could've been "If you are reading this article from a Windows system..." as well, in particular with Windows 8.

  5. I am not suggesting taking away any choice... I am suggesting *adding* choices for users. KDE and Gnome could behave in very different ways - I am not suggesting otherwise.

    As far as Windows 8 it is a mess. MS has lost their minds as far as I can tell. Apple, Google, Ubuntu, and others get the idea that it makes sense to have to different but related OSs for desktops/laptops and for tablets. MS wants one OS to rule them all - but it works well on neither. They two types of systems are used very differently: one with a touchscreen and, mostly, a virtual keyboard and held horizontally, the other without a touchscreen (or not one that is going to be used all the time), a screen that is vertical, and a "real" keyboard and mouse.

    I suspect they are trying to deprecate their desktop UI and move desktops to act more like the newer tablets with touch screens, but who wants to be reaching their hand out to their desktop screen all the time? Makes no sense.

    If you are looking at me as someone who is going to defend MS with Windows 8 you came to the wrong person. :)

  6. "I am not suggesting taking away any choice... I am suggesting *adding*
    choices for users. KDE and Gnome could behave in very different ways - I
    am not suggesting otherwise."

    But if KDE and Gnome behave in very different ways, there will *always* be the "problem" of applications developed for one environment looking out of place on the other. That's simply a fact, and you cannot make this go away without forcing Gnome and KDE in the same direction.

  7. You may be my 'envoy' to COLA and solicit someone else to write. They will need to use their 'real name' however like yourself.

  8. I would love to see then "follow me" here - what I do *not* want is the insults and the attacks and the trolling. Keep Linux advocacy to discussions of Linux and open source and the movers and shakers in that world. While Tommy, below, disagrees with me he is discussing Linux and I have no problem with that. Disagreement is fine - trolling is not. That is how I see it anyway. :)

  9. Why not? Many programs work on multiple environments and also do a good job of fitting in with those environments. The developers of OpenOffice and FireFox and GIMP and other large open source projects have talked about this (and have done a reasonably good job - not perfect but good).

    I have a lot of faith in the open source community. I think they can solve problems, even hard ones you do not see a solution to. I believe different teams with different visions *can* work together to make the open source ecosystem better while still maintaining their own visions and allowing users to not just have the choices they have now but to increase those choices, as I describe above.

    Right now to do this well a developer needs to have two versions of their product - a KDE version and a Gnome version (not to mention a Windows and OS X version if they want to have their software run on other OSs). Heck, some also have iOS and Android versions. I think the KDE and Gnome teams can work together to make this easier for developers to do (perhaps even having a switch that allows the menu names and locations and hot keys and the like to fit either DE). Would this be easy and perfect? Of course not... but as I have said, I have a lot of faith in the open source community to solve even hard problems. And this problem is clearly not as hard as having, say, Firefox work on and "fit in" with not just some Linux systems but also OS X and Windows... which has already happened (even if not perfectly). Same with Chrome. And LibreOffice. And GIMP. And VLC. On and on and on... these open source projects prove you can have their products "fit in" with more than one environment.

  10. The 'cage fight' mentality tends to diminish if not go away entirely and politeness resumes when people are forced to use their real identitities. It's simply a good practice. One doesn't walk around in public wearing a ski-mask does one? ;) That aspect of COLA will always rule out my participation. People there are not acting in good faith when they use aliases.

  11. I can certainly see that but I can understand people not wanting their identities known. I have had people from COLA contact my family, employer, co-workers, and more with the most repulsive lies and attacks - in one case with publicly stating the goal was to have me fired (which did not work, but still, I can see why people would want to prevent this).