Lewis' Blog Tales from the trenches of information technology


On the ungrateful nature of users and the OSS development community

Download PDF

Nothing new, here, I'm afraid. It's a problem as old as the concept of "free stuff."

So what exactly is free stuff, anyway, and why do people seem to equate freely available to freedom from cost?

Recently, a contemporary of mine posited the idea that as the majority of Linux apps ported to eCS were built using GCC, they were somehow unstable, inferior, untested, or otherwise unworthy of use, and further, that developers porting these apps either did not care to - or simply did not - test them before releasing them on an unsuspecting public. This summoned the ire of one of the OS/2 community's most prolific - and talented, and approachable - porters, who reiterated what so many of us have said to and about this individual for some time: "There's the door."

(A brief aside to mention the actual challenges of porting these apps to OS/2 using GCC: The biggest issue we seem to face is fragmentation of memory in the shared arena. This fragmentation means that after a certain period of time, entire sections of physical address space become essentially unusable, and the only way to regain the space is to reboot. On a system like OS/2, which was designed for long uptimes, this is simply unacceptable, and obviously, downright frustrating when one begins to notice the effects of this fragmentation. In short, there was a time when apps running on OS/2 simply did not crash, and when an errant process did fail, it surely did not take down the whole system. Not so these days, unfortunately.)

Getting back to the thrust of this post, what exactly is the confusion between software which does not come with a licensing fee and software which is fully supported and somehow falls from heaven, in a retail box, with an endless warranty?

People such as the one cited above, instead of being appreciative of the time and effort that developers and otherwise talented people put into creating, maintaining, documenting, testing, enhancing, supporting (maintenance also differs from support; the former includes fixing bugs on a generic level and providing compatibility updates, whereas the latter implies providing specific attention to the needs of solving one particular user's problems - hence the difference between maintaining an application and supporting a user), and porting to platforms on which the original developer(s) either was/were not focused or not able to build, seem to think that they are somehow entitled to defect-free software with ongoing maintenance and support, just lacking the retail box. This, of course, could not be a less correct interpretation of the concept of freely available.

I am much more inclined to donate funds to developers who give up their otherwise spare time to work on their pet projects and who then, in turn, share the fruits of their labors with others - with no stated price tag attached to measure that value - than I am to plunk down $xxx for a retail package likely to contain more features than I want, the development company of which could really care less about my enhancement suggestions and critique, beyond fulfilling what is stated in the original purchase support period.

Now, I firmly believe that there is room in the universe for both models to coexist. Our income tax preparation software, for example, comes with excellent support. Support, I might add, which would literally be impossible for a single individual or small group to provide on a timely basis (and as income tax filing season in the US is essentially 90-115 consecutive days - not factoring in extensions - this is a heavy lift even for a commercial developer). However, I do strongly object to the marketing of some companies' office suites and the exorbitant prices charged for accepted minimal standards (i.e., nothing new and amazing) - a word processor is essentially a word processor. OSS to the rescue! Long live OpenOffice.org!

In his original post, the individual in question began by citing instabilities in recent builds of Thunderbird, calling them "totally unusable."

Where would I be without Mozilla? Well, on my platform of choice, eComStation, I wouldn't be very far along the Information Superhighway, that's for sure. Am I pleased with the rapid fire release cycle for Firefox, Thunderbird, and SeaMonkey (SeaMonkey is the most sane of the three, particularly in this regard)? Of course not. But is that the fault of the people porting the code to our particular platform? Naturally, only a fool would consider the concept (and whether the individual who prompted this diatribe is indeed a fool I shall leave to the reader to decide).

Is the current OS/2 port of GCC at fault? The people responsible for porting the compiler, perhaps? Serenity Systems / Mensys, the purveyors of the current incarnation of OS/2, eComStation? Or can we take all of this blame and heap it on IBM for leaving us in a lurch with what is now essentially a 15-year-old 32-bit kernel from a 25-year-old paradigm, which was never designed to do the things we seem to ask of it every day in our current computing universe of modern applications, which are designed not for 386 and 486 processors, but for multicore systems with more RAM than the average user could have afforded to buy (let alone afford the chassis in which to install it) when OS/2 was in its heyday...?

Marvel of engineering design as it is, the OS/2 kernel is not without faults. It has scaled quite well over the years, but it does have its issues with memory management and fragmentation, and these days, applications are using more and more resources. Fragmentation in the shared memory arena is a huge hurdle for us (the OS/2 user community), and not since my DOS days have I been so acutely aware of the limitations of physical RAM as I am currently.

All of this aside, I try my best to thank - via kudos, public and private acknowledgement, reciprocal services, and cold, hard cash - as many of these selfless folks who allow me to work in my preferred manner on a daily basis. I just wish that others would at least try to grasp the concept of TANSTAAFL, and realize that freely accessible software doesn't mean free software: someone has to work to make that happen.

Comments (2) Trackbacks (0)
  1. Lewis,

    I tend to ignore the ungrateful people. However, they can be annoying.

    I want to say two things.

    The first is when it comes to donations, I’ve been making an effort to wave my specific cash donations to OSS in other peoples faces. I hope it actually does some good in terms of getting other people to donate, which in turn leads to better software for myself. I don’t know if whatever good I do negates the wrong from letting the left hand know what the right hand does, but I’m always looking for ways to get more people to donate.

    Second, as a LibreOffice user, and one time NeoOffice donor, I will have to say that Microsoft Office is leaps and bounds ahead of Neo/Open/Libre office in terms of usability. I’ll also say that google documents is in some ways more usable then LibreOffice and its forks. Mind you, I’m not talking about the office ribbon, which I hate. My anger is over little things like no shortcut keys for “insert hyperlink” and “fill series.” I’m not even looking to make my reverse DNS excel macro work (server audit).</rant>

    Now I like and use Libre Office, and deal with its faults as opposed to buying MS Office. Its not even MS hate because I happily shell out $500 for Visual Studio. I’m actually amazed at how well it does work. My word format resume is maintained in LibreOffice, and I’ve never looked at it in Microsoft Office before sending it to a potential employer “just in case.”

    • Good comments, all, Justin; and thanks.

      The last version of Word I used (and own) was/is Word 5.0 for DOS (and OS/2; it’s a bound executable). 5.5 had a completely redesigned menu system, and most 5.0 users would tell you that they disliked 5.5 as a result. I was never a WordPerfect user (DOS or Windows), nor a WordStar user (though I have a friend who swears by WS to this day). The first GUI word processor I used was Lotus Ami Pro, and later, WordPro.

      I’ve supported both MS Office (from about 6.0 up through 2007; I don’t have any clients running 2010) and Open/Libre. As both of them (Word and Open/Libre) have what I consider less than intuitive menus, I have my issues with both. The difference is that if bothered enough, I can effect the development/feature set of OpenOffice or LibreOffice (the recent decision by the LibreOffice team to abandon the code specific to the OS/2 port means that I will stick with OpenOffice on my daily system). However, even if Odin (the not-so-nearly-equivalent of Wine for OS/2) does one day support running MS Office, the investment in a closed end solution where I have little need of developer support (it’s an office suite, after all) and no chance of effecting change to the application is enough of a turn-off for me that there is little value in comparison to the OSS alternative(s).

      For clients, this is somewhat different, though I have historically shied away from providing “desktop” support (I’m a network engineer; I usually refer clients to the developer for end-user questions regarding how to use a given application), I have had to support various “features” on such closed source apps as MS Office. I find that they work well, when used in conjunction with other such closed source applications (e.g., Outlook plays nicely with Exchange, and Word plays nicely with Outlook), but rarely do they adhere closely enough to standards to be truly useful in a heterogeneous environment (square pegs and round holes). And so it goes…

      Thus, when I frown on such closed source solutions – regardless their extended feature set – clients all too often accuse me of being anti-M$ or anti-whatever, when my real slant is in favor of more open solutions which provide a voluntary upgrade path by supporting and adhering to accepted standards. This, I believe, is the true value of OSS, and why such efforts need support and the people who contribute to such projects deserve all the reverence we can give them.

      Kudos for supporting the OSS stuff you do, both financially and by contributing your time and talent. I, for one, truly do appreciate it. 😉

Leave a comment

No trackbacks yet.