|[Home] [Credit Search] [Category Browser] [Staff Roll Call]||The LINUX.COM Article Archive|
|Originally Published: Tuesday, 12 October 1999||Author: Mike Corns|
|Published to: corp_features/General||Page: 1/1 - [Std View]|
Linux: An Open Source Lab
"This article isn't about why Linux is technically wonderful or about the "evils" of profit. The community abounds with technology wonder stories - some exaggerated, most not and the costs though important, are not the compelling issue. It's a survey of some of the more common criticisms of Linux and Open Source, the rebuttals of those critiques and a discussion of the new awareness this investigative process awakened in me as a decision maker and strategist to the real importance of Linux and the Open Source Evolution. "
Two months ago, my Linux exposure consisted primarily of knowing that it began with an "L" and ended in an "x" and that Linux was a hybrid formed from some guy's name. In the course of obtaining sufficient understanding needed to get everything I wanted out of my new Linux-based LAN environment I became immersed in the Linux "community".
Everything about the Open Source community and Linux seemed antithetical to my experience with "state of the art" proprietary hardware, operating systems and development tools. Then the real epiphany - it occurred to me that an antithesis for the current "state of the high tech arts" might not be such a bad thing.
If you are a decision-maker painfully familiar with the high costs of ownership and the technical demands created by mainstream, "normal" technology, if you have watched more and more of the same product and service delivery methods produce less than startling impact on your business other than steadily increasing costs. Then you should give Linux and the Open Source movement a hard look.
This article isn't about why Linux is technically wonderful or about the "evils" of profit. The community abounds with technology wonder stories - some exaggerated, most not and the costs though important, are not the compelling issue. It's a survey of some of the more common criticisms of Linux and Open Source, the rebuttals of those critiques and a discussion of the new awareness this investigative process awakened in me as a decision maker and strategist to the real importance of Linux and the Open Source Evolution.
I carefully chose "evolution" rather than "revolution". Open Source is not a brand new idea (http://www.oreilly.com/catalog/opensources/book/toc.html). Linux, arguably the preeminent Open Source banner carrier, is based upon tried and proven technology seated upon an improved new foundation, (http://www.oreilly.com/catalog/opensources/book/linus.html) .
However, critics, myself included, of Linux and Open Source have contended that several crucial elements are missing from the financial and manufacturing model embodied by Linux:
These criticisms taken together and at face value insured that there was no way I was going to seriously commit to Linux much less Open Source - even if the technology was good.
- The "oldness" of the Unix model on which Linux was based.
- The absence of centralized support organizations.
- The "otherness" of the very idea of free and open sources.
- The zealot, super hacker culture.
- The expected descent into schismatic warring by various Linux distribution factions.
- The improbability of an upstart, pseudo-socialistic adventure placing any serious pressure on the established players in the proprietary software community.
My journey into Linux challenged those views. Sharing elements of that journey with you should, I think, challenge your views if they are similar to my pre-Linux perspective. I do not believe that the proprietary software community is bound for extinction. I do believe it is hurtling towards a period of significant realignment of its manufacturing model. I understand the reluctance and incredulity in some quarters surrounding Linux and Open Source. I make my living today exclusively using proprietary technology. Nearly twenty years in managing proprietary software product development, punctuated with a few stints in consulting and contract programming, exposed me to a wide array of technology and management challenges.
Nothing however, was quite as unsettling as the idea of "free software" and "open source". Every time I considered the idea I vibrated emotionally between fear for my livelihood and incredulity at the ridiculousness of the notion. In particular, it seemed clear to me that the technology could never overcome the latent suspicion associated with the phrase, "you get what you pay for".
Still, when faced with purchasing a server license that would cost nearly as much as the machine I had just purchased and with sufficient experience with the other server OS's to know that nothing was easy - I took the Linux route.
The price was right, so I bought a book, that included a Linux distribution CD, purchased a 400 mhz Pentium II PC and began my journey into the "community". Along the way, I've been frustrated, impressed and finally, energized by the trip. Technically, I'm extremely satisfied with the results and the required effort to obtain those results. But, what of my concerns about the viability of the Open Source business model? As I explored these criticisms, in concert with making my new Linux technology work, my view of them changed dramatically.
Critique: "Linux is based on old technology."
Why is it that incremental improvement is so bad as Linux detractors, and this criticism, suggest? Most high-reliability technology is the product of improvement iterations - not ground up redesign.
You should ask yourself if the "quantum leaps" in OS technology in the last five to eight years produced real value for businesses? Are quality, stability and reliability synonymous with these improvements? Moreover, it's not that Linux has suddenly appeared on the Intel desktop and mid-range server market as much as those servers and workstations have become increasingly capable of fully exploiting the power of Unix from which Linux is derived. Other OS's have been trying to keep up with desktop hardware advances and arguably not doing all that well. Unix has been waiting patiently.
Critique: "I want a central source for information and support"
The Linux community is awash in sources of information. How-to's, news groups, IRC groups, and User Groups. For a few examples see:
http://www.linuxdoc.org/HOWTO/Virtually all of this information is free for the taking. In fact, one of the ongoing challenges for the community is to consolidate, catalog and improve access to the wealth of information that is available.
The problem with Linux information is that it is like drinking from a fire hose. But, as criticisms go, that's a pretty good one. What's more, after going through the single front door of many large proprietary software vendors web sites, you find that finding information inside that door is no simpler than finding information on Open Source products.
Critique: "This is some blip on the high-tech radar screen. Some fad soon to fall by the wayside as the old order of business and service reinstates itself."
I doubt it. The Linux community is the first to fully mature within the global information community created by the Internet.
The Internet changes, entirely, the viability of the Linux/Open Source model. Old models of support and distribution have already begun to change in directions enabled by the Internet. Is it surprising that the manufacturing model, and that is what Open Source is about, would change as well?
Critique: "How is software owned by no one going to be maintained and improved? Where is the incentive?"
Imagine a technology marketplace where service and support vendors scramble to get your business. Where they pounce on fixed priced requests for support and bid against one another for larger support, enhancement and product development opportunities.
Imagine a competitive, $$$-driven market based on improving what you already have rather than a market where each vendors is trying constantly to convince you to switch to their flavor of the week product.
Imagine not being forced to frequently absorb all the costs associated with switching or upgrading to radically different, "new and improved" technology. Imagine a reference database where quality profiles of individuals and companies was maintained by the customers of those entities - not by the suppliers.
Now, stop just imagining. Two exotic start-ups, CoSource.com (http://www.cosource.com) and Question Exchange (http://www.questionexchange.com) provide a glimpse of what could evolve to support Open Source products.
CoSource.com provides a broker service for collecting Linux enhancement requests and the price the requestor is willing to pay for that enhancement. Requests can be accumulated producing the potential for significant funding to be presented to the developer.
Question Exchange uses a similar approach in creating a forum for paid technical support for Linux related issues. Both offer the ultimate in free market competition. Neither could be feasible without the Internet and the access to source and experience Open Source permits.
Critique: "But this Open Source thing just doesn't make sense. Open Source is scary. You mean just anybody can hack my code?"
Well, yes and no. Think about it. Can't just anybody hack your code now? How much do you spend on anti-virus technology? Do you spend that because binary, proprietary code is uniquely secure? How do you know if someone has hacked the code? You can't ever look and build from the source - the definitive documentation?
Maybe you have an escrow agreement with some of your suppliers. But, how many companies validate that the escrowed software matches their production binaries or that they or anyone other than the vendor could actually BUILD the escrowed sources. Or, even that the escrows CAN be built - by anyone at all? Generally, you never want to look at the source. But, why does having access only to binaries make us feel so much safer?
The point is that Open Source products like Linux are not only, by definition, escrowed - but, you, also, get to touch the source and if you want, you can build it and validate your ability to produce a correct binary version. That Open Source accomplishment alone should catch the attention of anyone familiar with the challenge of making portable build environments.
Critique: "I like being able to place requirements and bug reports with a single company."
You want to know your problems are being addressed, that someone hears them and acts on them. Okay, that's fair. But, how do you know that with a proprietary vendor? Wouldn't it be useful to have the option to actually pay for a fix or enhancement in a competitive software bazaar? Payment would only be made if the fix worked and was incorporated back into the publicly available source baseline?
Are you really so satisfied with the responsiveness of the proprietary vendor community, that you are prepared to rule out Linux and Open Source just because the Open source model is different from the proprietary model?
Some criticism has been made of the relative slowness of the Open Source Linux community as compared to proprietary vendors in responding to system improvement requests. First, I suspect that critique warrants further examination. Second, consider the responsiveness obtained from an industry without corporate walls or substantial funding. What will happen when monetary incentives and the Open Source community model are fully integrated?
Critique: "The development and support community for Linux is weird. Are they artists or programmers?"
If a renewed call for personal accountability is one hallmark of the late twentieth century, what better implementation of that accountability could be had than you knowing who the developer(s) are that built your software.
Ok, you say, but free. How can something free, or orders of magnitude cheaper than it's predecessors, sustain itself in a free market? Now that's a funny question in the face of the seeming economic paradoxes of the internet. Yet, the internet is producing extraordinary revenue growth with much more by all predictions to come.
All I can say is go look at CoSource.com and Question Exchange again as possible answers. Or, recall the internet economic and technological buzz circa 1994 and compare it with the internet buzz now. Remember when the word was that the Internet was going to collapse and that there was no way to make money on it?
The Linux folks ARE different. They want their names on everything they do. Would you rather get millions of lines of compiled code from anonymous, faceless corporations instead of these uber Geeks who want bylines and recognition for what they do right as well as for what they do wrong?
Critique: "The support model will lack the ability to respond to my time constraints. How can we meet aggressive deadlines and tight schedules supported by a massive, geographically dispersed, loosely managed development model like that used for Open Source?"
Are your schedules being met all that well now and are we talking about just Linux and Open Source as entities facing this challenge? Is that far-flung development model unique to Open Source? Don't many major software companies resemble that structure? Doesn't it describe more and more large software development projects of all kinds? Doesn't it seem like the Linux people are managing it pretty darn well?
Are you more comfortable betting your business on one vendor with proprietary protection of their software asking you to "trust them" to deliver what you need on time? Many of these situations give you little recourse if they don't deliver other than to switch technologies - which is usually impractical if not ludicrous - or lump it which is what we usually do.
Would you feel better with dozens of vendors each with access to the same code base? Each vendor working to improve your technology rather than displace it with their proprietary variation? Each motivated to do the job better and faster than the other vendors to be successful -- to get your money?
Critique: "In the 80's the Unix guys beat each other to death. I think the Linux community will do the same."
Well there is some infighting in the community regarding which distribution is best. But, this wave is different from the earlier Unix folks. These folks are managing themselves and they have less of a vested commercial stake in most of what a distribution contains. They generally are arguing qualitative nuisances not major, functional differences.
Compare that with the OS and Browser Wars. Moreover, the Linux folks have established forums like the Distro Wars site (http://distrowars.linuxmafia.org) and canons governing the ethical methods of advocacy (http://www.linuxdoc.org/HOWTO/mini/Advocacy.html). They have learned from the mistakes of the past and they have the Internet to create an organizing medium that allows cohesion unprecedented outside of a corporate structure.
Critique: "There are really only a handful of people that know how to support an Open Source product. What happens if they quit or decide to do something else?"
Proprietary companies seldom have MORE people than absolutely necessary to support a product. The myth of depth being the hallmark of proprietary companies really needs to be challenged. Generally, you make money by doing the most with the least - isn't that manufacturing 101. If one of their folks is promoted, quits, dies or is fired the same problem occurs in a proprietary enterprise, however, you may never know about it.
Open Source is no different other than how your technology is or isn't maintained is a bit more visible. Of real concern is the ability of Open Source to continue to grow unless it makes some money. That leads to the next, and last, criticism.
Critique: "We live in a capitalist economy. Ultimately, it takes money to manufacture, market and support technology."
Absolutely. My mortgage holder will not accept good will as payment. Unlike some of the folks in the Linux community, I see events like the Red Hat IPO as not only good but, essential if Open Source is to succeed.
The Open Source movement is developing a new competitive emphasis for our industry - not advocating the elimination of competition or monetary reward. To date, emphasis has been placed in the proprietary world on competing through "innovative obsolescence". By that I mean technological innovations have tended, or at least purported, to make competing software offerings obsolete. Innovations have also tended to obsolete some aspects of earlier versions of technology from the same vendor as well. The effect of this has been to balkanize technology along lines defined largely by proprietary ownership rather than by significant technological superiority and to create significant "organizational pain" in order to switch from one technology or version to the next.
We are entering an era where information technology pervades every aspect of our society. The switching costs must slow without completely giving up technological innovation. Reliability, integrity and qualitative improvement of the technology must progress much faster. Open Source facilitates this maturation process by shifting the emphasis from functional differentiation as the key competitive objective to quality differentiation as the primary competitive battlefield. In concert with this shift new businesses will emerge that make money by being primarily better at doing what they do rather than being primarily different.
But, there are problems with the current Open Source model. Why should I innovate and then give away my rewards for that innovation? I think the answers for these problems are a short distance away.
Perhaps what is needed is a variation on the current licensing for Open Source products, the GNU General Public License (GPL) (http://www.fsf.org/copyleft/gpl.html). A modified structure that would permit a period of time where an innovation's economic rewards are secured for the innovator, where innovation must occur in a manner that evolves rather than obsoletes existing technology and then, after that, the innovation moves into the current Open Source model.
I am confident that these problems will soon be resolved. Like the problems that plagued the early internet and the web, there is too much to be gained for the Open Source problems to go unresolved for very long. In the interim, wise technology consumers and producers will get involved with Open Source in order to educate themselves to this new manufacturing and consumption paradigm and train their organizations for its emergence. Many organizations are recognizing this fact. See the recent spate of articles and announcement from heretofore traditional, proprietary oriented sources:
- Apple heads down open-source path (http://www.zdnet.com/pcweek/stories/news/0,4153,1014146,00.html)
- Plotting an Open-Source Path
- AOL opens-source code for Web Server
- INFO World Opinions article
- Open source gathering steam
- Sun Community Source License Principles
Linux provides a useful means for initiating involvement with this mega-trend. Along the way involvement will produce value for the organizations that support it and place them ON the Open Source train rather than behind it or worse yet in front of it.
Are Linux and Open Source different? You bet. We've been churning technology now for twenty years without achieving the reductions in costs or improvement in fundamental value delivery we expected when the "PC-revolution" began. If you subtract new functionality delivered through the Internet from consideration, how many real breakthrough business improvements can you name that have been delivered through the desktop or the server markets in the last nine years?
Maybe the next real breakthrough won't be entirely technical. Maybe it needs to come in through a different door. Isn't that what innovation is all about? As for me, I plan to keep my eyes and my hands on Open Source and Linux as an Open Source indicator and introductory technology, it's important. I recommend you do the same.
Comments? Email the author of this piece.