I remember my first encounter with a computer was in high school calculus, an Olivetti Programma 101. It was part of our curriculum to program the arithmetic steps for summation equations. A decade later, I was learning DOS on a military computer. Not the underlying technology, I became the training guru for our Enable office suite. I also wrote all the automation scripts in Enable for the forms we had to process. I still use a copy of Enable O/A in my XP Mode emulator on Windows 7.
I also wrote a lot of other instructional material in the military because my superiors said I had a knack for it. Then I got hurt and left the military. Back home, I started writing a book, but the pile of papers quickly grew out of control. When I got a battered old DOS machine, the writing became more serious. As better used computers fell into my paltry price range, I found distracting all the newness of Windows and graphical office suites. Going back to college a second time, I encountered WordPerfect in the computer lab and fell in love; nothing compares to Grammatik. I still run a copy of WP 6.1 in XP Mode just for fun.
At some point, I discovered Linux. I can recall the thrill of getting my mouse to work when I stumbled across a RedHat 5.0 book with the disks in the back. I always thought Another Level was a great desktop, if lacking the highly integrated functions of Windows. I surfed with Netscape Navigator 3.06 for several months before I discovered updates. Nor can I forget buying Applix 4, then the thrill of getting my hands on the retail box of WordPerfect 8 for Linux. Despite the occasional crash with it on RedHat 6.3, I thought it was wonderful, a real advancement over 6.1 on Windows. Applix 5 was cool, too. I still have the boxed sets for WP8, Applix 5, and RH 6.3. I'm currently running WP8 on Ubuntu Hardy in a virtual machine with VMware.
I was still working on that book, but progress was slow because I was eternally distracted with usability issues. I knew it could be better and the Open Source community kept promising it would be better. I kept believing.
Meanwhile, I got involved in volunteer computer assistance work. Not so much tech support but I focused on user training. I could fix hardware and software issues, but I spent the vast majority of my time helping people understand what computers could and couldn't do. I always took the line people were the reason computers existed, not the other way around. I'm not into the joy of computing so much as how it makes some jobs quicker and simpler. Computers have always been just tools. For someone aspiring to write and teach, it was my best and most important tool, but never more than just a tool.
Over the years, I've discovered far, far too many of those most capable with computers were the least capable with people. In the commercial software world, you don't have to care about the customer, only convince them you do. You have to respond to user input to some degree or lose that market share. There are people in management and marketing who understand this, and won't pay developers who don't at least work toward producing software they can sell to the consumer. The software is superior for the most part at things that only marketers notice. In Open Source, the roles are reversed. The coders are demigods and those who serve as their management and marketing are dependent, and much closer. The software is superior only at things coders and techies notice.
Meanwhile, the user is totally left out in the cold by both types of software production.
Open Source software is seldom better and often inferior in all the ways most important to ordinary users, while commercial software meets most of their expectations. I understand the sociology and economics of the software market, how it can both lead and respond to business at the same time. I understand the necessity of both standardization and competition, how people as a whole get used to whatever dominates the market they work. It's no mystery to me Microsoft seems to tolerate a certain measure of piracy simply because it keeps them dominant. It's free advertising at the grass roots level. Just because you have a better idea in terms of how computers themselves work does not mean it's better in terms of how people work. Changing the computer landscape won't happen by attacking user patterns; that's only a symptom.
You see, users seem to think support is important. They'll sacrifice some things to get that. It won't matter how pure and elegant the technology is if they can't get help with things they don't understand. That is pretty much my whole mission in volunteer work. It's always a compromise; I help the user negotiate the possibilities. For example, commercial software companies often provide fixes it seems they never bothered to advertise. Open Source seldom offers fixes to the stuff that gets your attention. When they do, it's not a fix, only wholesale replacements — “fixed in the next release.” That is anti-user, because users like stability. Too often, that replacement comes with too much relearning of new habits, or a whole range of new breakage. While that's true to some degree from both sources, Open Source seems almost hostile to user input on the issues.
Very few Open Source project managers understand the concept of stability of a product and fixing the good features already included. Once users incorporate software into their work routine, they don't want significant changes. They aren't computer technicians. They cannot be techies if they are to accomplish anything else. It's enough work just getting used to computers as part of the routine; computers cannot become the whole routine. Wholesale replacement needs to be far better than the previous stuff with no substantial difference in how it works. Users don't care what constitutes techie habits. They want technology harnessed to their habits. They'll compromise some, but frequent wholesale changes are not compromise, they are user abuse.
Open Source is the worst about that sort of abuse. Too often intermediaries (distributions) have full replacement control; they are serving entirely different agendas from both the developers and users. Getting the projects to talk to distros is nearly impossible, except when they both agree you have to replace everything. In addition, it's not just backward compatibility, but there are the maddening complications of getting a new version without having to change everything else. I can replace something the size of the Nano editor without much trouble. Older libraries compile newer versions for quite some time moving forward, but the whole desktop? Moreover, the entire system of Linux software provision shifts constantly with vast numbers of players, so that trying to keep track of it is impossible if I am to have any time left for helping people, which is the whole point.
Right now, the most stable distributions aim only at the corporate server market, not the user desktop. The most common user software in Linux is where most of the bugs are, and is the stuff least likely to be fixed, especially in the stable distributions. Rolling release and frequent punctuated release are both anti-user. The hard work of getting something truly stable, and then keeping it working long term is antithetical to the ways of the code gods.
Why do you think people still keep trying to run MS Office 97 on their newer versions of Windows? What they use most needs to change least once it does a respectable job of meeting their needs. It's the same reason MS Security Essentials is now a dominant anti-malware product; it's the least intrusive while getting the job done. You simply cannot make high security a priority if it's too much work. If you want people to use the stuff, keep it within their attention channels. Otherwise, stop pretending the world needs what you have.
It's bad enough the commercial world is loaded with anti-user behavior. For longer than a decade, I've used the free email service offered by GMX.net, based in Germany. Excellent stuff and support is great, if you don't mind chatting with them in German (I don't mind). I've only needed support twice over the years. So why does their UK/US branch, GMX.com, stink so badly? You can get a free account, but the service can be very unreliable. Support response is almost nonexistent. They seem to have taken a cue from Yahoo on this, but without the advantage of being so popular and not nearly so usable. There are others; Sega seems to hate their customers, and Sony Entertainment tried to defend their rootkit ploy. Yes, the gods of commerce too often want that total control which intrudes deep into the user's personal life. Still, if it gives them what they want, people will use it.
The consumer software market is accountable through sales. Accountability in Open Source depends entirely on the random character development of the folks who write the code. The user has no leverage. You count yourself lucky if the worst you get from someone is, “Write the code yourself. I don't care what you want.” If we could write our own code, we'd have no use for your project. The problem comes when the common trend of development departs significantly from what the user needs and expects. Even when Open Source developers decide to listen, too often they leave themselves open only to fanboys who already love the same things. It's almost incestuous in that sense, because common users find the whole system utterly closed to them. The developers and fanboys are in their own world. The barriers to entrance are excessively high and the insiders become prickly when users aren't willing to invest that much. Open Source insiders don't understand that we'd like to use our computers, not have sex with them.
Therefore, we have a very large community deluded into thinking they can sell Open Source on what they value, as if the world should only rightly value the same things. Open Source developers seem to expect reprogramming users should be as easy as computers. It doesn't work that way and the arrogant snippy comments on places like Slashdot demonstrate why there will never be a Year of the Linux Desktop. Demanding every Internet user gets a “net user's license”? It defines the elitist snobbery that makes Mac-heads look reasonable and friendly.
When KDE 4 came out, and the developers refused to listen, I started losing interest. When GNOME 3 came out and the developers became hostile to user input, I gave up. XFCE? That's where I got that snobbish quote above about “code it yourself.” Minimum expectations of computer users require a fully integrated desktop experience, but none of the other Open Source desktops on offer comes even close to that level of integration. There are fans for just about every window manager and desktop, but those people have in common almost nothing with ordinary users. For a brief period, I could sell people on Linux, but that was at the end of KDE 3 and GNOME 2, back when it seemed developers realized they had to compete for their place on the user's desktop. That's all gone now.
Telling me I should be bowing-and-scraping grateful because it's given away freely, and the source code is open for inspection, doesn't mean much if I can't use it. Have you ever heard of Google? They try to straddle the line dividing Open Source and commercial profit. If the service or software they offer freely is a pain to use, their advertising share goes down. While Open Source typically is not ad supported, most ordinary users really don't see the difference. Google has something to lose, but Open Source is by nature wholly unresponsive. Sometimes you can get the developer to respond to a genuine bug report; I'm grateful for those projects. Some of the project managers actually like people. It's why they do what they do. Timothy Pearson's Trinity Desktop Project merits honorable mention on that score. The good guys are always woefully under-supported.
I've spent countless hours with clients and while you can play amateur analyst all you like about what makes people tick, real people will change if what you offer is actually better for their needs. A significant portion of my clients switched to Ubuntu 10.04, but they refuse to use 12.04. Current Linux offerings are no better on any of the measures users notice; the offerings are steadily getting worse. It has nothing to do with what the developers and fanboys think is good stuff. Real users want what they want. I don't try to tell my clients what they want. I do my best to listen and give them what they ask for, including many things I think are utterly stupid. They are in the driver's seat; it's their hardware, time and money. I don't sell Linux any more.
I'm very grateful for VLC, FFmpeg, GIMP, Notepad++, jEdit, Cygwin for my CLI fix, including Lynx for most of my Net surfing, The Sword Project, and many other heroes of Open Source. Each project includes folks who respond, either by fixing things, describing workarounds, or reasonable explanations for things. They are accountable, either directly or through their fan base, which works just as well.
So, I still use some Open Source software, but I gave up on Linux. It's okay for security and absolute control, but otherwise way behind the competition. I run a couple of virtual Linux machines because I prefer it for when my curiosity draws me to unknown places on Internet. However, most of my work is done in Windows 7 with real office suites.
I don't care if all the new kids on the block are all about the smartphone interface. There are millions of us yet needing to get work done, not simply consume a stream of bytes as entertainment. Nor do I care that Windows 8 could easily become the next ME/Vista debacle. Hanging onto Windows 7 for as long as possible offers far better prospects than anything else. So long as there are people who do work on their computers there will be an under-served market.
For myself, I finished the book but outgrew it. Instead, I've written several more and published them. My budget is nearly zero most of the time; my hardware is mostly donated. The software I use came with the hardware or someone gave it to me. I'd be a prime candidate for Linux if it ever met my needs because I know how to make it work as well as it can. Instead, I'm using a significant amount of commercial abandonware, lovingly maintained for download by fans from yesteryear. My publisher insists on Word format, so I use MS Office 2000 as the last step in processing books for submission. For everything else, it's a copy of Corel's WordPerfect Office I bought used. Open/Libre Office is not even in the same class, offering a large number of interface glitches, all of them apparently longstanding bugs they haven't bothered to fix. Need I mention it has no viable grammar checking?
If someone ever makes Kmail so it's not buggier than all other mail clients in existence, when it's also possible to use it on Windows, I'd love to switch. The concept is marvelous. For now, it's Opera. I'm keeping an eye on the KDE ports to Windows, but I don't really expect much, because the entire KDE project has been the buggiest nightmare ever to intrude on my dreams. From what I can see, the distance between new features and debugging grows exponentially.
I'll stick with what does the job for me as I have an awful lot of stuff I need to write. Meanwhile, I can dream that the Open Source community will stop deceiving themselves as well as stop trying to deceive the rest of us when the community can't be bothered to care what matters to us. You can have what you want for yourself; I have no complaint with that. Enjoy it. Just stop the propaganda.
Ed Hurst is Associate Editor of Open for Business.