strawberryJAMM's Security and User Experience WebLog

The delicate balancing act between intuitive user experience and secure software systems

Blogs

The Problem of Usabilty in Open Source Software

  • Comments 7
  • Likes

I just recently found this article on NewsForge about the problem of usability in Open Source Software (OSS). It was written by Frans Englich, who is a participant in the KDE project:

Open source usability is a technical problem we can solve on our own:

"Poor usability is a huge barrier to wider open source adoption. Our backends have matured and we consistently achieve technical excellence. Usability is the one area we have not yet mastered. For some reason, we treat it as a mystery instead of looking at it as a problem we can solve the same way we solve all other technical problems. " - Frans Englich, Friday July 09, 2004 (04:34 PM GMT)

I thought his initial characterization of the OSS community as a group of people who are wary of usability issues (and, by inference, other user experience issues) was both interesting and fairly accurate. In fact, I'd go so far as to argue that it isn't just OSS developers who look on with confused and dazed expressions when confronted with usability issues but that it is actually a characteristic of a significant majority of developers. Which is why product development teams frequently suggest or expect responsibility for the "mystical matter" of usability and user experience to be handed over, wholesale, to outside forces such as the "outside Companies, Experts and Laboratories" Frans attributes to the OSS developers or the members of a User Experience team that is rarely (in my experience) looked on as an integral part of the product development team.

Usability and other user experience issues aren't really all that mysterious - they just represent another technical problem that needs to be addressed along with all the other technical problems associated with a software application, such as Security, Performance, &c. I think we (Microsoft) are starting to grok that across the company - after all we already have "HOWTO's" for user experience issues in the Windows environment in the form of the Windows User Experience Guidelines and the Microsoft Manual of Style for Technical Publications - but it's going to be a long, slow journey that involves a significant culture shift away from developing technology for technology's sake and towards solutions for the user's sake, regardless of technology.

I just hope that our developers accept this culture shift and eventually acheive the goal of producing a truly usable, useful, operating system and software applications.  (Preferably before the OSS community does. )

Edit1: Corrected link to the Windows UX Guidelines. Thanks to uwe for pointing out the error.
Edit2: Expanded on and changed the tone of my last sentence.

Comments
  • Another great resource for me was http://useit.com

  • Oh, and another thing, when following your link "Windows User Experience Guidelines", I see the start page, but no links to follow?!?

  • "I just hope that we reach the goal of producing truly usable, useful, software before the OSS community does."

    The problem Microsoft has, unless it can successfully undermine the legal basis of OSS, is that OSS is under no such time constraint: its fixed cost base is much smaller than Microsoft's.

    It's also worth pointing out, that to the extent Microsoft's software costs money and OSS is available free of charge (modulo support), commercial software is expected to be of higher quality. After all, it's reasonable to expect that a $300 copy of Office has a way to justify the marginal cost difference it has compared to a free download of a competitor. (It's pretty sad that Microsoft's only serious long term competiiton these days is free of charge.)

  • If the community wanted to build an OS which may be perhaps verbosely renamed as the "Bloated Insecure Unstable Overpriced Media Player" (BIUOMP), they indeed would have done so. Linux was generally used by people who knew what they were doing and thus it's usability is a tad more esoteric at times. But that facet is complemented by additional power and stability. (Of course the times are chaging... it's now sexier than windows ever could be)

    If you want Linux to be the aforementioned BIUOMP then you can make it to be so. But it's flexibility and adherence to good software engineering form mean that it has the power to be better at everything that any Microsoft OS Platform tries to be. And with Solaris soon to join the OSS community, Microsoft's server product line is fast becomming an embarrasment for the company.

    The ridiculous situation your engineers have found themselves in in attempting to 'port' Windows to x86_64 bit architectures is typical of a badly designed unscalable project. With Linux it was usable as soon as the platforms hit the market. A look at the state of the precious Windows low level code would be humorous lession in how not to design software systems.

  • uwe: Yes, Jacob Nielsen's site is definitely one of the best places online for usability information. Jacob's book "Usability Engineering" is also a good resource because of its emphasis on "Discount Usability" - that is, the quickest and easiest ways to start dealing with the Usability issues in any project.

    Thanks for the note about the link - the actual content for the book had actually been moved. I've updated the URL now.

  • > The problem Microsoft has, unless it can
    > successfully undermine the legal basis of
    > OSS, is that OSS is under no such time

    Microsoft does not need to "undermine" any "legal basis" - all it needs to do is to consistently produce better software and add value to it in the form of functionality and support. The widespread belief that Microsoft cannot compete with "free" is ridiculous, because "free software" is not really free once you move out of your parent's basement and dump the Celeron "boxen" running Slackware. Microsoft does not "fear" free software, especially now that IBM, Novell and the like have jumped into the fray. Free software can't go anywhere in the business and home markets without money and support from corporations, and it's that association that makes it also a tangible entity that can be targeted by a competitive strategy. The "summer of love" is over.

    > It's also worth pointing out, that to the
    > After all, it's reasonable to expect that a
    > $300 copy of Office has a way to justify the
    > marginal cost difference it has compared to
    > a free download of a competitor.

    Surely you don't actually believe that OpenOffice is comparable with Microsoft Office? It may be, if you compare it to Office 97. Works 2.0 might be a better match. This never-ending hype and proclamation about how FreeProductX is about to kick the floor from under MSProduct has been going on for four or five years - at the same tempo - and for the life of me I still can't see how OO.org (or anything else) can be considered anything but a bloated, slow and crappy-looking clone of Microsoft Works. Word, Excel, PowerPoint and all the other Office products are far superior to anything these people can churn out, never mind the fact that they still have no Access clone *and* no equivalent to VBA, OLE or anything else that people who *purchase* Office rely on day to day.

  • "all it needs to do is to consistently produce better software and add value to it in the form of functionality and support."

    Which is exactly what Microsoft's competitors have said in the past, only to find out that MS would just keep trying, again and again, until they got it right. IMO, that is MS's essential corporate value: pick a plan and stick to it, come hell or high water. Products like DOS, which sold with virtually every PC shipped, helped this along by providing a reliable means to fund other products. Also helpful was that a lot of the early competitors (Lotus, WordPerfect, etc.) were strongly tied to one product. Therefore, any hit to the marketshare of that one product hurt the competitors' financials and thus their ability to improve their product. A nice and effective downward spiral...

    Now, fast forward 15 years and the big competition is OSS. Marketshare no longer impacts competitors' revenue nearly as directly and most of the development is either done by dedicated amateurs or corporations trying to support other interests (hardware, consulting, etc.). Even worse for Microsoft, there's no one corporate entitity to go after: putting Red Hay out of business doesn't end the threat of Linux.

    In short, the OSS community is in the same position as MS back in the 80's: they don't have to win to compete. Losing hurts, but doesn't end the game, and the question now becomes how long Microsoft can run before it stumbles and gives people a reason to switch.

    Maybe this will happen in 1 year, maybe in 5, 10, 20, or not at all, but no matter how long it takes, as long as OSS is legal, it will be there, waiting for its opportunity. After all, it's taken 20 years to get this far.

    "Free software can't go anywhere in the business and home markets without money and support from corporations, and it's that association that makes it also a tangible entity that can be targeted by a competitive strategy. "

    However, even if competitive companies drop Linux or fail, the product itself will still be around and able to be evolved and improved by someone. With enough people using Linux, Linux support will be a viable business for quite some time to come.

    "The widespread belief that Microsoft cannot compete with "free" is ridiculous, because "free software" is not really free"

    of course not, but is is cheaper? Is it better able to avoid vendor lock in?

    "Surely you don't actually believe that OpenOffice is comparable with Microsoft Office? "

    Of course not, but that's not the point. The real question is this: "Is MS Office comparable with OpenOffice+$300/user+being_locked_in_to_one_vendor?"

    "they still have no Access clone *and* no equivalent to VBA, OLE or anything else that people who *purchase* Office rely on day to day."

    You'd be suprised, basically all of that technology already exists in OSS form, particularly some of the component oriented and scripting stuff.