OS/2 eZine

16 March 2001
 
Pete GrubbsPete Grubbs is a self-described OS/2 wonk, a former doctoral candidate in English literature at Indiana University of Pennsylvania, a former part-time faculty member at Penn State and is still mucking about with a copy editing/creation service, The Document Doctor, which tailors documents for small businesses. He has also been a professional musician for 20 years and is working on his next album, scheduled for release in early 2001.

If you have a comment about the content of this article, please feel free to vent in the OS/2 eZine discussion forums.


Previous Article
Home
Next Article



Back Again II for OS/2, Windows 98 and NT


Finding a New Path - Part II

Last month, I detailed my long and tortuous path which has finally led me to a more accurate understanding of IBM's corporate workings. I've also come to the conclusion that we OS/2 users have basically seven choices before us when it comes to our future operating system(s). For those of you who tuned in late, here's my list:

  1. Windows
  2. the Mac
  3. Linux
  4. eCS
  5. Don't change anything, apply Fixpak 15 and ride the pony till it dies
  6. Abacus
  7. A Whole New Thing that's really Partly an Old Thing

For various reasons, I narrowed my choices to numbers 4, 5 and 7. Apparently quite a number of OS/2 eZine readers have been giving this issue a good deal of their attention, because many of you wrote me to share your thoughts. One reader chastised me for not including Software Choice and the Convenience Pack in my list, even though Rob Basler looked at both e-ComStation and Software Choice in some detail in our January edition. Perhaps I was remiss in ignoring that option, however, it wouldn't have made my final three picks for reasons I'll detail below. Another disliked my lumping of the various flavors of Windows under a single heading since Windows 2000 is a clearly superior product than its predecessors. (While it's true that many in the computer world have been bragging up Windows 2000 as a huge advance over its earlier incarnations, considering their points of reference, it wouldn't take a whole lot for any other OS to look good, and we shouldn't overlook the fact that Microsoft has long been attempting to put all of their OS eggs into the same insecure basket. What assurance does any computer user have that the new, unified Windows won't be even buggier than the various predecessors it's built upon?) On the whole, however, the responses I've received since this editorial hit the web have indicated that many of you are thinking in the same direction I am. I've heard from readers who are getting ready to move to Linux, others who are enthusiastic about Serenity Systems and eCS, and at least one who is content to ride OS/2 as-is until it just don't work any more. Many of you expressed an interest in my choice. As one of my favorite fictional characters would say, "Ah, therein lies the tale . . ."

Back to the Future

One of the most enlightening messages I got came from someone who identified himself as a former IBMer. His most important observation (to my mind) deserves to be quoted at length:

"IBM is a FOR PROFIT Hardware company, at its heart, and will quickly (even more so now) change its (software) direction, and whatever is in its wake, is the victim of the currents."
I remember the sense of vertigo I felt when I first read that sentence; I also remember the tremendous 'whack' that reverberated through the room as my hand hit my forehead. Want to know why IBM can't market OS/2? Want to know why IBM has been and will continue to push Linux as hard as possible? Want to know why huge sections of IBM didn't care if its customers preferred Windows to OS/2? Look at Big Blue's corporate history and culture: IBM is a HARDWARE company. According to a recent article in ComputerWorld, one of the few bright spots in the latest downturn of technology sales has been the mainframe and server market. In recent quarters, IBM, the hardware company, has been making money selling its favorite product, big iron.

Now, its AIX brand of UNIX has been a solid, mature staple for years. If it is successful in marrying AIX and Linux, it can cash in on the recent Linux craze by making its big iron that much more attractive to its favorite market segment, the enterprise, and it can do so without taking Microsoft on directly. Imagine the marketing pitch IBM can toss: the stability and security of AIX mated with the cost, desktop applications and GUIs of Linux running on some of the best hardware on the planet. That sounds like a pretty attractive package to me. When you consider the number of Linux-native open source/freeware apps that are popping up all over the planet, that deal looks even sweeter, especially when you consider that IBM's software development investment in all of this is, comparatively speaking, minimal, and it gets brownie points for supporting open source. Friends, this looks like a winner to me.

Of course, it also means that, from a hardware company's point of view, there's even less reason to give a damn about an 'also-ran' OS like ours. Please don't misunderstand me: I'm not suggesting that OS/2 is dead. I'm not even saying that it's on life-support. What I am saying is that there aren't a lot of compelling financial reasons for IBM to do a whole lot with OS/2. As long as it has enterprise customers who are using it, IBM will continue to support Warp just enough to keep them happy, but I see even less reason this year than last to believe that IBM will put any serious effort into doing anything more than maintaining what it already has. It might release some new device drivers, it might tweak up Warp's USB support, but I have to question where any revenue would be generated if it did more than that. This is the reason that I'm not interested in Software Choice and the Convenience Pack. As the months go by, I can't see that it will offer the kind of advancements that I want, since there's no visible financial incentive for IBM to spend the money to put them in.

In other words, I think the Big Blue part of our ride is drawing to a close. So, that brings me back to my list and the first option I want to discuss,

5) Don't change anything, apply Fixpak 15 and ride the pony till it dies

As OS/2 stands right now, it's got a lot going for it including fully featured office software like Lotus SmartSuite and StarOffice, a flaky-but-functional browser, some great e-mail apps, and all of the stuff that's coming our way from *nix ports via XFree86, to name only the very tip of the iceberg of apps that are readily available without any further developments from Netlabs' Odin project. If the hardware you're running right now has all the power you need and you're as productive as you need to be, there's precious little pushing you to go in any other direction. From my perspective, however, I'm not happy with the prospects that I'm left with if I decide to go in this direction. Right now, looking at the end of the first quarter of 2001, I don't see any need or feel any pressure to want a 64-bit Virtual Turnip Twaddler and Radish Whacker, but I don't know that I won't be faced with an overwhelming necessity to make one work on my machine by the end of the fourth quarter next year. If I'm sitting astride a comfortable-but-dead pony, I'm pretty well certain that I won't have that option, and that's not acceptable. While I don't need to have access to bleeding-edge technology, I've got to have something that does have some growth associated with it, so I won't be staying with what I have now, which carries me to my second consideration,

7) A Whole New Thing that's really Partly an Old Thing

When I interviewed Adrian Gschwend at Warpstock last year, one of his comments stuck with me: "The whole OS/2 community still listens too much to noise from IBM. The important thing about the OS/2 community is that we are the community, not IBM. We don't have to blame them, we have to do something better than they." I've been muddling this over for six months, turning it this way and that, trying to really get a handle on it and I've come to the conclusion that Adrian is right but I still don't know how to implement this idea in the real world.

The current Linux model is the most obvious and, probably, the most fruitful way to put the idea into practice, but it requires something that we don't have: access to source code. As long as IBM can make money with OS/2 in its current situation, it has no reason to allow the kernel to become open source. Given the success that Netlabs has enjoyed with their many projects, particularly Project Odin, I'm convinced that the OS/2 community has the talent, know-how and willingness to push the operating system's development forward, but that's not going to happen without legal access to the code. Unless circumstances change more radically and more quickly than I can see, I don't think we're going to get that access any time soon. Perhaps, in another 3 to 5 years, IBM will open the code up. If their AIX/Linux gambit plays out particularly well, that time frame will shrink as more of their enterprise customers who have been running OS/2, leave it. Still, a for-profit company isn't going to give away years of R&D until they've squeezed that investment for all it's worth, and OS/2 is still making money for IBM. (One of my colleagues, Fernando Cassia, has a different take on this than mine, but I suspect that the core of our disagreement lies in the fact that I'm looking at a longer time frame than he.) So that brings me to my final option,

4) eCS

Without rehashing Rob's January analysis, let me briefly say that eCS has a lot going for it right out of the box. It's got the stability and security I've come to expect because it's based on the kernel I'm already using. It's got some nifty features added in, and it's got some darn bright people working on it. Unlike the Big Blue Behemoth, the people at Serenity Systems are going to be a lot more accessible to their SOHO customers and a lot more responsive to those customers' needs. eCS is also bundled with enough software to pretty much establish a networked office right out of the box, but none of these are the reason I'm going to use it. I'm going to bet on eCS because of Linux and IBM.

Yup. Linux and IBM. Confused?

Maybe this will bring the picture into focus for you. It's a quote from the eCS web site http://www.ecomstation.com/more_information.phtml on March 15, 2001:

"Later this year Serenity Systems expects to announce support for Linux and expanded support for Windows." I don't know what 'expanded support for Windows' means and, frankly, I don't care because the only Windows stuff that I'm all that attracted to are high-end games that probably wouldn't run well in a non-native environment anyway. It's the 'Linux support' that makes eCS a winner in my eyes. If that support means that I can run Linux apps natively on my OS/2 desktop (which I certainly hope it does), it means that all of the push that IBM has been giving Linux is going to result in more Linux applications with better stability and support. It means that I'm going to have immediate access to stuff that's closer to leading edge than anything we've had in years; all that from the comfort of a desktop I understand and a kernel that's more like an old friend than piece of software. If Linux becomes the powerhouse OS that its supporters believe it to be, there's going to be a lot of applications written for it that will run on my eCS machine.

Of course, if eCS is a tremendous flop, I'll still have a dead pony that I can flog for another couple years of service . . . and then, it's time to figure out that abacus thing.


Previous Article
Home
Next Article

Copyright (C) 2001. All Rights Reserved.