Advantages of the Linux Community from a Buyer PerspectiveSep 25, 2000, 16:08 (6 Talkback[s])
(Other stories by Paul Ferris)
No-Size-Fits-All! An Application-Down Approach for Your Cloud Transformation REGISTER >
by Paul Ferris
This document is aimed at people new to Linux (often referred to as GNU/Linux), who are wanting to learn more about it from the consumer perspective.
Often, people who are new to Linux have questions like: "Where do I buy stock in Linux?", "What is all this licensing stuff about?", and "How come I've never heard of it before if it's so widely used?". This document is not aimed at developers, or business people looking for a new edge in software development. For those people, I highly recommend Eric Raymond's fine essay "The Cathedral and the Bazaar," as a great place to start.
This document is aimed at people who want to know who is creating Linux, why they are creating it, and what the benefits of community style development are from an end-user perspective. It is a general reference with links below to some of the more valuable web sites that cover the subject in more detail.
The Linux Phenomena
Since about 1998, the media's attention to Linux and the growing need for software diversity have driven a lot of people to look for alternative computing solutions. These people have often ended up coming to the Linux community to see if there's something in the world to answer their growing concerns about cost, freedom, security, privacy and quality.
As people continue to roll into the Free Software camp, there is a lot of confusion caused by community-familiar issues. Things like differences in licensing, arguments over software design and so on have a tendency to cause a sort of culture shock to manifest for quite a while.
There are still a lot of people that haven't heard of this alternate way of running software. As of when this document is written (September/2000), the number of people coming to Linux is rising and the number of debates hasn't really died down (and likely should never die down).
There hasn't been a whole lot of marketing by the Free Software or Open Source communities to the end users of their products. From time to time, end users will write glowing praises as to why they themselves use Linux. This is all well and good, but it doesn't say much about the positive community aspects. This document strives to bring these qualities to light.
First, a bit of a definition here of the so-called "Linux Community" itself. There are many ways to view the Linux community. The community could be viewed as a static entity, composed of a certain number of active developers who write the software that make up a typical Linux distribution (Red Hat Linux or Debian GNU/Linux are Linux distributions, for example).
There is more to it than just Linux itself. A Linux CD-ROM contains many software packages, created from many different projects. As an example, the free web server product Apache is often included in a typical Linux distribution. Apache is developed by its own community. Apache runs across multiple operating systems, including many proprietary systems. Another example is a file-sharing product that among other things, allows PCs running the Microsoft Windows operating system use Linux file servers on a network. That project, Samba, has its own community. The Samba product runs on many other systems (besides Linux) as well.
The GNU utilities included in most Linux distributions are central to the operation of the system as a whole. These utilities work on many other operating systems besides Linux, and have their own developer community.
These items, among many others, are loosely referred to here as part of what people perceive as "Linux," but are actually developed by separate communities.
There are also corporations who have hired some of these community members or in some cases the corporations themselves have been started by them. These people contribute to Linux, or their respective development projects, and are paid to do so.
Finally, there are non-developers who daily spend time supporting Linux via internet relay chat, newsgroups and bulletin boards. These people do the user community a service that is not so glamorous, but essential to the acceptance of a free product. They give their time as a way of "giving back to the community." They rarely receive mention, but it is arguable that they are as important as the developer community. Usually when a new Linux user has a question about Linux it is a member of this community that answers it.
All of these people comprise the loose community that creates and supports Linux. When a document refers to the "Linux community" as a whole, it is referring to these people, intentionally or otherwise, usually viewed as a static group of people.
In reality it is a bit more complex than even this definition. The people in these groups are in a constant state of flux. People come and go as work loads allow. Developers enter and leave projects as they see fit. The biggest mistake to make about the Linux community is to somehow assume it's a corporate entity.
So, in reality, it's more of a process, almost a living, breathing thing that answers a need by reflex. Although some of the development for Linux is paid for (as in the corporate contributions mentioned above), almost all of it is being donated. This scares some business people wanting to invest in Linux, because they wrongly assume that no money can be made from it. It scares some corporate managers as well, as they cannot believe that a product such as Linux exists. They cannot see how such a product can have useful, valuable features, if it is freely available.
The reality is that Linux exists contrary to these expectations. Each of these development and support communities have members that have helped create or maintain Linux. This answers the first question: "Who is developing Linux?" "Why?" is usually the second question.
To answer that question, one needs to understand that each project mentioned above had different reasons for creation. The GNU project began as a deliberate effort by one man, Richard Stallman, to create a totally free operating system. The Apache project began as a bunch of patches to the original freely licensed web server product put out by the NCSA. The Linux kernel (just the main brains of the operating system itself) rose out of the need by a lone programmer in 1991 who was fed up with the limited proprietary alternatives of that time period.
All of the projects are extremely successful in their own right. In many ways, the development Linux and the projects that comprise it continues out of a desire for high quality, user maintained software. When a consumer buys a "Linux" CD these projects are magically folded into something incredibly powerful -- an operating system that is as good as (some would say, better than) any alternative operating system product.
There is still some room for improvement, for sure. As of the writing of this document, there is still a need for a diverse set of shrink wrapped applications for Linux, and the technical aspects of the product still often require more study than the average new user has in mind.
Overall, the quality and performance per dollar aspects of Linux are very good. Linux, on its own merit, matches or exceeds almost any and all proprietary alternatives for commodity server tasks such as web serving and file serving. It makes a nice desktop as well, and it's a very good product for embedded applications (hand-helds and cell phones). There are many exceptions to these rules, but for the new user, these claims hold true in a lot of situations.
What are some examples of advantages to the community style of development? This question is actually the true focus of this document. The advantages of community style development are profound and important. They have to do with much more than mere software features, application availability, stability, or cost.
The control of Linux development is decoupled from the marketing, training, and support of the product. This is an enormous advantage, even though the gut reflex of many industry pundits is that of disbelief, examples of these advantages abound.
A favorite example of the author revolves around a discussion with Dan York, the president of the Linux Professional Institute (LPI). LPI is an organization responsible for setting the training goals that are to be met for certain Linux certifications. These certifications are comparable to other qualifications by proprietary vendors, such as MCSE(Microsoft) and CNE(Novell) certifications.
Dan relates the following anecdote: "I was at a seminar in Europe where the session speaker asked the audience 'Why do vendors create certification programs?'
"I was absolutely stunned when someone from a propriety vendor stood up and boldly said 'We do it to push product.' I mean, I think most people made that assumption, but to hear someone from the company state this plainly was just incredible. That's not why we (LPI) are doing this."
The LPI has many goals, but making money for training centers or Linux vendors isn't one of them. Their end goal is to build an the army of support engineers that can install and deploy Linux solutions in companies world-wide. "LPI is about changing the world." Says Dan. "The other thing about LPI is that we are all about the freedom of choice... not only of your Linux distro, but also about the way in which you prepare for the LPI exams... you can choose between many different books, training partners, online courses, etc., etc. We don't mandate/authorize/approve/endorse any one way.."
What may not be clear is that the LPI is a disconnected entity from Linux itself, and the eventual revenue stream of the companies that are actually implementing the training. Its goals are set by a review team that is decentralized, and therefore, has very little impact on marketing or development of Linux.
The end result: training and certifications that actually mean something more than good marketing material. Training that is a good investment for the company needing the expertise of the trained individual.
As an aside, Linux underpinnings are very similar to those of most modern Unix products. This means that the training and investments for Linux are a two way street. People using high end enterprise level solutions such as HP-UX and AIX, for example, often find Linux administration to be similar, if not identical. The same is true in reverse.
For these same reasons, Linux coexists very nicely with enterprise quality operating systems. This isn't to say it doesn't exist well with legacy low-end proprietary solutions, such as Microsoft Windows products. It's more to point out that Linux is very good at utilizing true industry standard protocols, such as those that most enterprise class Unix systems depend upon.
This allows midsized to larger companies to save even more money in the long run, as the training is a better value company wide. Single-sourced, non-standards compliant proprietary solutions do not offer this kind of low end to high-end interoperability.
Please also note that not all corporations take training to this level (treating it as marketing). There are good corporate sponsored training facilities that offer sensible goals and honest intentions. The point is that by being controlled by a decentralized entity, Linux certifications have an ethical higher ground to operate from, and with little or no marketing influences at all.
The support avenues for Linux also have positive benefits. There are varying levels of support available. Traditional commercial support can be purchased , such as that offered by LinuxCare, or from individual Linux distribution vendors such as Caldera and RedHat. Free support is available, such as that offered by newsgroups and internet relay chat (IRC) groups, such as Open Projects.
And again, there are numerous advantages.
If a corporation produces a piece of software with bugs, and they have paid support lines, software quality begins to take on hues of "conflict of interest". The least negative light that can be cast upon it results in a pall of nonchalance about the most devastating of system crashes. System crashes in the proprietary system world have been made light of many times. The example that often comes to mind is the one where the CEO of a company is demonstrating new software in a public forum, and it crashes, requiring a reboot.
The author has known true-life examples of this same software companies products being implemented as machine controls for manufacturing processes. Among the various buttons on the machine (Emergency stop, Start and so on), there was a button labeled "Blue Screen," that the operator pushed when the tell-tale hex dump of a system crash reared its ugly head. Of course, unlike the other buttons on the machine, this one was hard-wired into the inner workings of a computer -- it was the reset button for the controlling PC.
Linux, due to its high software quality, rarely exhibits such show-stopper bugs. In general, it is extremely stable, even under extremely high load conditions.
Show-stopping bugs are not viewed as marketing or sales problems by the Linux community. The tools to track and fix them are clearly open to everyone who has the motivation to learn to use them, and the quality is taken to heart. Bugs themselves can be referred to as such.
At least one proprietary vendor of late has even gone so far as to never refer to a bug as a "bug," but instead, an "issue." The technique brings to mind Orwellian overtones that are another selling point for Free Software: If a user wants to call it a bug, they have every right to do so. The community doesn't attempt to censor speech, or attempt to make light of something so serious as software quality by attempting to control the thoughts about the issue, or bug as it may be.
Security problems are addressed in a similar vein. Security fixes happen in real time, and are not in the hands of one controlling entity. Recently, a disturbing security bug was found in the file serving subsystem of a well know proprietary operating systems vendor. The vendor decided it was too much trouble to fix, even though it was an enormous security problem for end users.
The real problem here, a most disturbing one actually, is that despite the overwhelming outcry of the user base and security experts in the field, the company made a decision that will likely grossly effect their customers in a negative way.
Linux is a truly open system. People that use it have a say in how it is created, developed and maintained. Serious security problems, like the one above, are fixed by the community.
It would appear also that lacking a governing body, or a corporation to back it, that Linux would be at some kind of disadvantage when a security hole in the system was discovered. The irony is that the community response to security holes is often very fast. Recent studies on the subject show that on average the Linux community is at least twice as fast at fixing security problems than a lot of proprietary corporations.
Product Release Cycles
The Linux development community structures product releases by a simple rule: "When it's done, it's done." From a standpoint of stock holder expectations, this could be viewed in a negative light. However, with quality being put first over all monetary motivation, the end result is software that people learn to depend upon. How ironic that software coming from the Free Software movement includes the phrase, in all caps, "ABSOLUTELY NO WARRANTY," yet it is often of the highest quality available for the task.
Contrast that with "End User License Agreements" that pop up during installation that contain similar phrases, but have much higher costs associated with them, monetary, and otherwise. The user typically has very little say in the development of these applications, has paid a higher price to obtain them, cannot share them with a friend, and cannot modify them if they so desire. To top it all off, the user often faces restrictions such as artificial limits on the number of people who are allowed to connect to their machines.
Add to these ailments the fact that software released by single sourced proprietary vendors of late have been shipping with literally thousands of bugs. All, it appears, to appease a stockholder audience, instead of putting quality first. Long delivery times by proprietary vendors are often a mystery or more often accompanied by unbelievable explanations being proffered by company marketing staff.
In total stark contrast, if a user wants to tune into the intricacies of Free Software development, they can simply join a mailing list and track the project of interest for themselves. The ability to see exactly what is going on is a benefit to the consumer of the technology. By knowing that such lines of communication are open, an end user does not have to doubt the "real" reasons for delays -- they are self evident. No amount of money can compensate for this value, literally.
Clearly, Linux holds the high ground here for the person who will invest their time in learning about the product. The community aspects of the software bring to bear a stark contrast to singularly developed proprietary alternatives. Often in these proprietary locked-in marketplaces, the end user is viewed as a competitor to the very company that sold them the product in the first place. The Linux community suffers no similar ailments. New features are added for the simple sake of having new features, and not for the sake of increasing revenue or to motivate people to re-purchase technology they already own.
SoftwareLook before you leap. All in all, if someone is contemplating a switch to Linux, these benefits are in favor of the decision. There are real limitations in regards to software availability, however. Are all of the needed critical applications available for Linux? Check through existing software archives, and see what plans (if any) the vendor of that software has for providing a Linux supported product. Call them and ask for their plans if unsure. If the vendor isn't planning a port of their product to Linux, are there free or commercial alternatives?
As of this writing, for example, there are several non-free and at least two free (covered under the free GPL License), Office suites for Linux. Many games have been brought to Linux from other platforms, and more are on the way. 3D gaming under Linux requires some specialized setup at the moment, but overall the die-hard 3D gamer should be very pleased at the amount of control Linux provides at the hardware level.
Finally, bear in mind that products like VMware are available, which allow an end user to run any operating system, including proprietary products from Microsoft and others, in a window. This is less desirable than having a native version of the software, but can be used as an interim solution in a scenario where the user needs one or two applications that are not available yet as native Linux products.
As this article goes to press some of the less technically savvy PC OEM's are not offering Linux as yet for a desktop choice. Dell, IBM, Compaq, Penguin computing and VA/Linux are all offering hardware solutions centered around Linux technology, so the choices among quality vendors are starting to emerge.
Preexisting hardware is another issue. Linux supports a vast range of PC hardware, and even a lot of Apple hardware as well. The main thing to bear in mind is that a lot of hardware is supported, but if an existing system is targeted for upgrade that is based currently on legacy proprietary operating system software, some homework will be needed to first make sure that all of the hardware is supported.
This is typical of a switch to new technology. OEM Vendors that sell PC's as total solutions have to ensure that the hardware they are shipping to a customer works with the supplied operating system technology. End users attempting to switch that hardware to a different technology typically have to go through the same amount of work, and it can be daunting. This work can make it appear incorrectly that the older software was better at supporting a broad range of hardware.
If this is too much for an end user, then a selection with Linux pre-loaded from one of the vendors above is recommended. Face it, not everyone has time available to spend on the task of researching if their sound card or video card is widely supported under Linux. The balance is that time may likely be saved reinstalling operating systems, or addressing the myriad of privacy, security and stability concerns that proprietary alternatives seem to be presenting at the moment.
Linux offers a lot of positive qualities to the new user looking for a change. It is not an end all solution, and it's certainly not for everyone. But it does do a lot of things very well, and it provides a new choice to a new competitive PC marketplace.
More users are joyfully discovering these new choices on a daily basis, and the number of Linux users rises substantially worldwide, every year. The end result is that a growing demand for more applications will be filled by developers, either new or old.
The Linux community offers the end user products that are of typically higher quality than proprietary alternatives. These products are developed by programmers that focus upon quality and adherence to known standards. The training and support revenue streams for Linux products are disconnected from the development costs. These factors make it a very good long term choice for end users and businesses looking for desktop and server alternatives in today's challenging technological world.
Paul Ferris is the Director of Technology for the Linux and Open Source Channel at internet.com, and has been covering Linux and Open Source news for over 2 years. He is an editor for Linux Today and a contributing author on Linux Planet.
This document is a work in progress, and suggestions for improvement are welcome.
Notes and Hypelinks:
Popular Linux DistributionsCaldera
Related Free Software/Open Source projectsApache Web Server
Samba File Server
OEM's Offering Linux on new PCsCompaq
Support/TrainingLinux Professional Institute
Open Projects IRC Linux Support
Office SuitesApplix Ware
Internet Sites tracking Linux software developmentFreshMeat.net
E-commerce sites offering Linux on CDCheap Bytes
Linux System Labs
ReferencesThe Cathedral and the Bazaar, an essay by Eric Raymond
The Free Software Foundation
0 Talkback[s] (click to add your comment)