The purpose of this paper is to provide analysis on the Linux operating system along with Open Source software in general. Discussion will include benefits of the operating system, some of its downfalls, and the direction where Linux and the Open Source movement is headed. It is targeted specifically for IT managers and CIO’s who are considering Linux Distribution rollouts to replace their existing IT systems. Lastly conclusions and recommendations will be made. This paper was created primarily as a research paper however it includes some personal experiences, and experiences derived from other Linux users.
Introduction
The whispers of a growing number of Linux desktop platforms have reached many an IT managers ear, and as the community continues to grow and blossom it will become a point of hot contention as companies seek to justify switching to Linux. As a recent article at LinuxWorld (Gael Duval, 2003, Subheading 3, Paragraph 1) points out “More and more companies are migrating parts or all of their infrastructure to Linux. They need a system that is more efficient, more flexible, more open, more robust, and more customizable. In these challenging economic times, businesses also need to reduce their total cost of ownership. Linux is certainly the best solution for all of the reasons stated above, and more.” However as with anything new to your IT infrastructure there are many factors to consider before migrating to Linux.
Opening up to Linux and Open Source
So what makes Linux different from every other operating system? The answer is that it essentially is 100% free. Richard Stallman an AI worker for the Massachusetts Institute of Technology, who was offended by the idea that companies could prevent people from freely distributing software and source code, founded this concept of free software. Stallman is famous for his quote “Software is information” and “All information should be free!”. From Stallman evolved the General Public License (GPL), which permits software to be freely distributable, and its source code freely accessible. What this means for IT managers everywhere, is that they can download a completely free fully licensed operating system and distribute it amongst as many computers as they wish. The implications of this license are profound when you factor that a single licensed copy of Windows 2000 Professional can cost anywhere from $170 and up. So the savings even for a small business with 10 computers is easily over $2000.
Many companies have invested in creating their own Linux operating system, which are described as Linux distributions. Although there are many smaller companies and individuals that provide distributions, the main distributions are Redhat, Mandrake, and SuSE. Each of these three vendors provide free downloadable copies of their latest operating systems and purchasable packages for businesses or individuals who wish to acquire technical support or a larger more robust system administration feature set.
Every distribution also comes with a hefty amount of bundled software that you would never find in a Microsoft based operating system. Openoffice.org has recently produced an excellent office program that is freely available, and features a full-blown Microsoft Office compatible word processor, spreadsheet, presentation and drawing suite as well as database access and MYSQL support. Ximian Evolution features basically everything that Microsoft Outlook features, and at no cost. Project management and programming software is also available freely with every distribution. Another great addition is GIMP, which is a graphical manipulation program similar to Adobe Photoshop and offers professional grade graphical editing capabilities. For companies into digital movie editing/production, MainActor and Jahshaka provide professional grade multimedia editing capabilities, and are quite easy to learn. All of these pieces of software combined with many more make Linux a popular choice for anyone looking to decrease their bottom line. Though some of these programs are not quite as intuitive as their windows counterparts they are certainly still more than adequate.
Perhaps one of the biggest areas Linux has powered into is the server environment. Imagine the ability to have your own Web server, DNS server, SQL server, or your own file and print server readily available at your fingertips. For a fully featured server in the windows environment with the same services an IT manager would be looking at spending somewhere between $1000-$2000 depending on which version of Microsoft Server you decide to purchase. Configuring these servers although not trivially simple is no more difficult than configuring them in Windows. In fact in my own experience, configuring Apache in Linux was much faster than trying to get Tomcat to work in a Windows 2000 Server environment. Some distributions such as Redhat provide simple tools to aid the user in configuring these services, thus greatly easing the pain of manually editing certain files.
Arguably the largest benefit of running a Linux system is security. Although there are tens of thousands of viruses for Windows with new ones produced every day, there is only a couple hundred viruses for Linux most of which are not nearly as severe as their Windows counterparts and are produced on a much more infrequent basis. Customized security programs such as Bastille can lock down your system into a very secure state within a couple of minutes, and provide much better security than the default servers built into newer version of Windows. Even though viruses on the Linux platform are very rare, for peace of mind there a number of organizations that continue offer free virus scanning software.
Current Issues with Linux and Open Source
Although there are many selling features of switching to a Linux system, there are equally just as many down sides. Firstly, the Linux community is plagued by perpetual beta releases of every piece of software that reaches a certain milestone. Instead of waiting until a program, or more importantly an Operating System is fully operational and ready for release, a lot of companies have been marketing products that are not yet complete, or contain a large number of bugs. As Tower Group’s Shahrawat states (Jim Middlemiss, 2002, 3rd Last Paragraph) a “Major downside is that the Linux marketplace is the Wild West.” This analogy of comparing the Linux system architecture to that of the old Wild West may certainly be a little extreme, however it definitely does state a point. There are many issues that hamper Linux, and when one jumps on the bandwagon it can certainly get wild at times.
The learning curve from switching from Windows to Linux is certainly quite steep and in fact many users who are not experienced IT professionals give up on a regular basis. The problem is, that documentation for most programs is highly technical and often missing vital pieces of information that the company just leaves up to the end user to figure out. A frustrated editor trying to install a PVR/media jukebox server system from Extremetech (Dave Salvator, 2003, Paragraph 3) writes, “I took my time. I read all the documentation. I diagnosed and fixed multitudinous glitches. But instead of success, I’m sleep-deprived, frustrated, and ready to chuck the box out the window.” Although this may be a somewhat extreme example, the difficulty of installing and understanding how to maintain a Linux based system is not in the least trivial. Even for seasoned professionals, the task of installing certain applications can be tedious due to the poor package management system. A good example of this is an application (A), which requires for package A to be built, packages B and C are also necessary. However where the trick comes in, is that for C to be built you must get packages D, E, and F. Often the necessary packages you need to install are not available for your distribution, or you receive errors while installing them that prevent you from installing any further. The simple fact that installing programs is difficult should concern IT managers looking to run Linux as a desktop environment for users, as training these users on how to utilize the new system will be a long and costly process.
There are a number of frustrations that Linux as a whole has yet to overcome. One issue is the widespread lack of driver support for non-mainstream hardware. Another issue is the technical difficulty level of configuring servers and sharing network resources. Lastly yet another issue is that technical assistance is more or less left up to the user unless they purchase a vendor package.
Upcoming Ethical Issues with Linux
Over the last 20 odd years, Linux has progressively built on its code foundation due to generous donations of code by single programmers, groups of programmers, and companies. If it were not for the community that Linux has built itself, it would not exist in the state it exists today. However a recent court case propagated by SCO against IBM has launched the very core of Linux into the midst of a large ethical debate concerning intellectual property. For the purposes of this paper we will define the term intellectual property as any intangible asset that consists of human knowledge and ideas.
The argument set forth by the SCO Group concerning Linux is that IBM has misappropriated some of SCO’s Unix trade secrets and has built them into Linux. According to SCO’s CEO David McBride, (Stephen Shankland, 2003 A, Paragraph 4) “We’re finding…cases where there is line-by-line code in the Linux kernel that is matching up to our UnixWare code.” McBride said in an interview. In addition he also said “We’re finding code that looks likes it’s been obfuscated to make it look like it wasn’t UnixWare code–but it was.”
IBM of course responded and has claimed that (Stephen Shankland, 2003 B, Paragraph 4) “While IBM has endeavored to support the open-source community and to further the development of Linux, IBM has not engaged in any wrongdoing,” Certainly it appears that IBM does have the upper hand in the case as it is yet to be seen if SCO can actually produce verifiable evidence that IBM has been stealing its intellectual property. IBM has also backlashed with accusations that SCO is trying to interfere with the Open Source community and the further development of Linux. Indeed if SCO can’t prove its claims this would look very suspicious as Linux has been continuing to replace UNIX servers worldwide. Given this fact it is entirely possible that SCO is trying to stall Linux in an effort to slow transition of servers from UNIX to Linux.
Although the lawsuit at this time is pointed squarely at IBM it may not stay that way for much longer. More suits are expected against key Linux providers such as Redhat, SuSE and Mandrake as the case gains momentum. There are a number of implications in this suit and the way they affect business. Firstly there is talk that if SCO wins their lawsuit that there will be a tariff set forth by SCO on all Linux products to account for their so called losses. Several rumors have already spread that SCO is considering an additional fee per purchase of $96 US per CPU , for distributions that SCO claims have violated their intellectual property rights. Secondly with the upcoming suits again IBM and other Linux supporters/vendors, it is likely that we will see Linux development begin to taper a bit as companies will be less willing to invest in an operating system with questionable ethical issues. This will not only slow development in the market, but it may also push a large number of existing Linux networks into converting to other operating systems in order to avoid possible licensing embroilments.
As an IT manager it is important to realize that not only will Linux become more expensive per seat if SCO has their way, but the potential fallout related to this ethical issue could be critical in a time where Linux is needing corporate recognition. Even if SCO does not succeed, the results of negative press and poor PR that Linux is bound to receive during the court case could mean that Linux will not receive as much corporate support as they may have had prior to this lawsuit. In addition to this it will set all sorts of future precedents involving intellectual property. By having intellectual property as a claimable commodity, application designers have the hassle of reinventing the wheel every time they develop new ideas. What this would mean is that any time they wish to perform a function or concept that has been previously coined by another organization, they would have to come up with a new way of doing the same thing. This would not only waste unnecessary time in the development cycle, but it would also mean that finances would be drawn away from creating genuine new concepts, and technological progress would slow to a crawl.
It will be difficult to fathom how much this will cost Linux distributions if SCO wins, and whether or not existing users will be susceptible, but IT managers should carefully watch this issue before investing heavily in Linux infrastructure.
Industry Direction and the Future of Linux
Linux has recently garnered 10% of the market share, and has accelerated into being Microsoft’s number one competitor. With an estimated 18 million users across the world the Operating System is definitely going places. As an operating system it has experienced incredible growth in the last number of years, and in my opinion will continue to do so due to several key players now jumping onboard. Notably IBM, Sun Microsystems, and Novell have all begun development and heavy investment into the operating system. Novell in particular perhaps has the most to bring to the table, as their groupware suites are unrivaled in the windows environment. This will likely push Linux right into the waiting arms of corporate America, which is just beginning to release itself to the concept of Linux and reducing the stranglehold that Microsoft has set forth. “We’re putting our entire ecosystem behind Linux.” said Jack Messman in a recent interview (John Leyden, 2003, Paragraph 4), chairman and chief executive of Novell. “We are not abandoning NetWare, we are adding Linux. It’s all about choice for the customer.” Indeed it will be choice that will be driving consumers to choose Linux as their desktop platform. With support from such large corporations, Linux has acquired a strong financial backing, and will only continue to grow and prosper.
With the upcoming next generation release of Windows Longhorn in 2005 many eyes are fixed on whether Linux will muster the strength to provide the features that Longhorn will offer. One well-known industry analyst, Rob Enderle became famous by his shockingly true 5-year industry forecast on Operating Systems in early 1995. (Rob Enderle, 2003, Page 2 Paragraph 7). In his forecast he accurately predicted the decline of Unix and Netware and the rise of Windows NT. The most controversial aspect of all this was his prediction that IBM would abandon its OS/2 operating system which made his life very difficult. In his most recent article in Internet Week Rob Enderle states “Linux can match Windows 2005 on performance and trust. Like Windows, Linux runs on Intel (processors). And IBM is leading the effort, with Intel’s support, to support Intel’s trusted computing technology on Linux, the same technology that Microsoft is using in Windows 2005.” In the remainder of the article he goes on to state that Linux will continue to gain ground due to its incredible price advantage. He predicted that in particular, 3rd World countries would continue to embrace Linux with open arms favoring it highly over its expensive closed source competitor Microsoft. What this will mean is that as more and more users worldwide switch to Linux, undoubtedly the level of participation in creation of new program, input on bugs, and solutions to problems will increase. This in addition to the many businesses and governments that are finding Linux to be a great alternative simply by its terms of openness will provide an unprecedented level of growth in the Linux market.
Over the next few years I would expect to see growth in the development of Linux applications. The power of the Linux architecture is just beginning to be realized, and commercialization of Linux has just begun. The fact that Linux has been developed by thousands of individual people without any form of governing body will entice all sorts of entrepreneurs to begin building applications, as they will not have to pay royalties to Microsoft. What this will mean for IT managers is a wider range of available applications and proper support for those products. Lotus Notes is already being ported to Linux, and more mainstream applications are bound to follow. The outlook on Linux is good and the time to adopt is rapidly approaching.
Conclusions and Recommendations
Linux has finally become a viable windows alternative. However there are still a few pitfalls that could cause some havoc in a large corporate environment. An IT manager should evaluate these potential downfalls carefully, and weigh them against the costs of continuing to run the existing system. While certainly Linux is free, the cost of teaching new techniques to hardened Windows users is a time consuming and costly process. Although the average time required to do tasks in Linux is similar to that of Windows it does not take into account learner resistance or poor teaching skills, which means that it may take significantly longer to learn than say a new version of Windows which is merely a new look and a slightly different layout.
Prior to considering rolling out Linux in a full corporate environment IT managers should take the time to play around with different distributions and really research which matches their specific environment best, as each distribution has its up and down sides. The key is to choose a distribution that not only compliments your corporation, but also comes with a software suite that suits all your corporate needs. If additional software is needed, make sure this software is easily installable prior to rollout and that there are no huge bugs that will cause interference with daily production.
Linux has entered the corporate environment and it is here to stay. With all the built in free features it offers it certainly provides an edge that Microsoft operating systems do not. Switching should not be a matter of purpose; it should be a matter of choice, for that is what the Open Source movement is all about.
Glossary of Terms
Information Technology (IT) – a field that deals with management, business (especially online businesses, and retrieval of data; IT provides customer and business care.
General Public License (GPL) – The Free Software Foundation (FSF) is a non-profit institution that designed the GNU General Public License (GPL) to promote the publication of free software. The GPL is used by thousands of programmers who want to give others the right to copy and modify the source code of their programs. Millions of people benefit from this.
We use the GPL to allow everyone to use, copy and modify the Internet Junkbuster as they wish. Companies can use it for commercial purposes, but they are not permitted to use it in products that they claim as their property without negotiating a separate agreement with us beforehand.
Linux Distribution – A version of Linux produced either by a certain company or individual. Linux distributions tend to vary wildly in what software they contain, what hardware they support, and what features are included by default.Redat, Mandrake, Ximian, Lindows, and SuSE are considered the most commercialized distributions and are available with full technical support for a small fee.
Rolling Out – A term for distributing a new computerized system or Operating System within a department. Roll outs are usually fairly large scale and can incorporate a sudden switch to a new platform or a slow gradual replacement of old software with new software.
Tomcat – A stripped down version of Apache used primarily as a lightweight web server.
Bibliography
Proffitt, Brian (2003, October 3rd) Open Source needs long range plans. Retrieved October 9th, 2003 from
http://linuxtoday.com/news_story.php3?ltsn=2003-10-03-018-26-OP-BZ-LL
In this article the author discusses the need for Open Source projects to seek to create long range planning both in terms of development and in terms of financial backing. He points out that if Sun were to halt a large amount of it’s funding for major Open Source software many projects would suffer.
Salvator, Dave (2003, September 2nd) We have met the enemy and he is us. Retrieved October 13th, 2003 from
http://www.extremetech.com/article2/0,3973,1236323,00.asp
In this article the author discusses some of the major frustrations many new Linux users face. He goes on to make some interesting analogies on how to fix these issues, and draws some interesting conclusions on the fact that Linux needs to begin hiding some of its more advanced features from users so as not to confuse them.
Leyden, John (2003, September 8th) Novell ‘puts entire ecosystem behind Linux’. Retrieved October 10th, 2003 from
http://www.theregister.co.uk/content/53/32708.htm
An article on what Novell plans to do with its new Linux acquisitions comprising of Ximian and SuSE. It also discusses how Novell views the ongoing issues with SCO and to what extend it believes this suit could affect them.
Rapoza, Jim (2003, September 1st) Linux is not so ‘Free’. Retrieved October 12th, 2003 from
http://www.eweek.com/article2/0,3959,1234349,00.asp
This article discusses some of the aspects of what makes Linux not as free as we’re all led to believe. The term “total cost of ownership” is discussed which brings up several good points about the costs associated with learning Linux. It also coins in on issues some of the big Linux providers such as Redhat have while trying to turn Linux into a profitable market.
Duval, Gael (2003, August 11th) Why migrate to Linux? Retrieved October 8th ,2003 from
http://www.linuxworld.com/story/33886.htm
The author of this article provides a very well researched view on some of the upsides and downsides of switching to Linux. He analyzes the operating system, comparing its usability to Windows and provides some conclusions and recommendations on its use. Overall this article presented a very positive view on Linux.
Forrester Research (2003, April 5th) Linux questions and answers. Retrieved October 11th, 2003 from
http://rss.com.com/2009-1069-998389.html
This is a special FAQ posted by Forrester Research on technical aspects of Linux. The goal of the article is to advise IT managers for CIO’s of some of the benefits of Linux, the costs involved, some disadvantages, and who else is making the same Linux leap in the corporate environment.
Middlemis, Jim (2002, October 7th) Linux garners Wall Street’s attention. Retrieved October 10th, 2003 from
http://www.wallstreetandtech.com/story/mag/WST20021
007S0007 This article discusses the fact that Linux has recently drawn the attention of Wall Street. Indeed most of the key corporate players are now evaluating Linux in their IT departments to varying degrees. It talks about some of the corporations who have turned almost all their IT operations over to Linux, and some more who are on the verge of doing so. It also discusses the role of Linux in the large corporate environment and some of the issues that corporations are having with Linux.
Rosenfield, Eric (2002, May 30th) Switching to Linux. Retrived October 10th, 2003 from
http://www.yankthechain.com/linux.html
This is a editorial on switching to Linux. The author provides some background information on Linux and delves into why he believes Linux is a very viable desktop solution. It discusses many issues from application software to games, and provides a general overview of the pros of Linux.
Newman, Nathan (1999, December 12th) The origins and future of Open Source software.Retrieved October 9th,2003 from
http://www.netaction.org/opensrc/future/
This is a fairly large whitepaper on the origins and future of Open Source software. The paper discusses the history of Open Source, some of its standards, and a breakdown of government issues with Open Source software and a overview of where the movement may be going.
Stephen Shankland (2003 A, May 1st) SCO: Unix Code Copied Into Linux. Retrieved October 16th from
http://news.com.com/2100-1016_3-999371.html
In this article the author outlines SCO’s case against IBM. It discusses some of the areas in which IBM is accused of stealing some of SCO’s source code and including it in Linux, and provides some insight as to why SCO is suing. Concerns are also outlined by members of the Linux community regarding SCO, and counter allegations against SCO for FUD(foster fear, uncertainty and doubt ) slinging are also discussed.
Stephen Shankland (2003 B, May 1st) SCO: IBM Denies Charges of UNIX Theft. Retrieved October 16th from
http://news.com.com/2100-1016_3-999261.html?tag=st_rn
This is a counter article on SCO’s article on the misappropriation some of it code by IBM. IBM denies SCO Group’s allegations that it misappropriated Unix trade secrets, however it does not discuss what IBM’s strategy will be in its upcoming suit.
Rob Enderle (2003, November 17th) How Linux And The Mac Can Compete Against Windows “Longhorn” Retrieved November 18th from
http://www.internetwk.com/breakingNews/showArticle.jhtml?articleID=16100741
The author of this article has a very strong reputation of being able to accurately predict what will be coming down the Operating System road. In this article he summarizes what he thinks the Operating System scene will look like by 2005. He surmises that Linux will continue to grow while Macintosh is in for a bit of a rough ride. Windows Longhorn and Linux will definitely be going head to head more often under his outline.
This paper gives the SCO lawsuit far more credit than it deserves: “IT managers should carefully watch this issue before investing heavily in Linux infrastructure”. Um – why? Assuming they’ve watching the developments up ’til now, I don’t think they have much to worry about. SCO haven’t provided a shred of real evidence to back up their claims since day one, and have contradicted themselves several times. Thankfully it’ll be over soon – they were recently ordered to disclose their evidence (http://yro.slashdot.org/article.pl?sid=03/12/05/2021209) and laughably their legal representatives didn’t even bother to show up for the hearing.
I’m also curious where the market share statistics came from. Specifically, that “Linux has recently garnered 10% of the market share” and “an estimated 18 million users across the world”. Sources, please?
Instead of doing jobs, they will probably arguing why it is or it is not ready for the Desktop or switching between KDE or Gnome sessions – wasting any alleged savings from using linux and then some.
I never knew Linux’ market share was that large though. Overall, I thought it was a balanced view of the state of affairs. Definitely not a FUD article. It calls “a spade a spade” and it’s well written.
I would like to know where he gets 10% market share from.
From what I have seen Linux has around 1/2% market share on desktops and about 1/4 (25%+/-) of the server market.
Where do these people get these numbers?
From what I have seen Linux has around 1/2% market share on desktops and about 1/4 (25%+/-) of the server market.
Source please, where did you get ur numbers?
No, 1%+ market share on desktops
http://www.google.com/press/zeitgeist.html
Developer October 2003 Percent November 2003 Percent Change
Apache 28235972 64.61 30298060 67.41 2.80
Microsoft 10252227 23.46 9449180 21.02 -2.44
SunONE 1528090 3.50 1525202 3.39 -0.11
Zeus 735179 1.68 743611 1.65 -0.03
http://news.netcraft.com/archives/2003/11/index.html
question is how many apache run on linux.
for an enterprise to adopt Linux/open source SW is that their IT infrastructure remains under their control, and does not become subject to the whims of a vendor who is trying to “flush and roll” their customer to meet their revenue expectations. “Free” is as much about “Freedom” as it is about “no cost”. This is perhaps the most underrated advantage of Linux and open source software.
First, who is this guy? How is he qualified at doing an analysis at all. Does he have any experience in research? Almost all references are to oridinary news articles from the Internet, not the best references IMO. Maybe it’s just me but a paper to me is a research paper and they are very different from this.
I seem to have forgotten to include that in my bibliography but the source is from here:
http://www.linuxmednews.com/linuxmednews/1066189575
Really I think 10% is quite a conservative figure given that China is very heavily into Linux and they have a population of something around 2 Billion. This figure was meant to be a worldwide estimation. I don’t believe at this time that there is little else but estimations on Linux market usage. Does everyone who runs a Linux based system report that they are using Linux? This was merely meant to be a stabbing guess.
Hope that clarrifies things a bit. This essay was done as a 4th year term paper for one of my Computer Information Systems courses.
As to why my articles are all internet based. The reasoning behind this is that I found it VERY difficult to get my hands on printed material that covered the topic as it is still a very modern issue. Our university library did not have any information whatsoever on the topic, and I could not find any books on the subject online either.
Linux has more like 75% of the desktop as it gets deployed in China.
Like I said before Linux is good, but it will never be anywhere near what Windows is today. Really come on, Windows on the Server and Windows on the end users pc makes any network easy to work with and it saves companies money in the long run. Why would you not want that? Plus you get good techs for cheap because every techie knows windows. Having a Linux guru would mean companies would have to pay more for a skilled Linux zealot. Not to mention scaring the CIO from firing them cause who would look after the Linux servers? I am glad the days of paying techs $95,000 a year, letting them get away with slack dress code, and doing basically anything they see fit for the IT department are over.
Pleasingly detailed, but you need someone to proof-read it before handing it around (plenty of corporate types will pick up on errors and may develop a negative view).
For example:
SUSE (no longer SuSE)
Red Hat (not Redhat)
OpenOffice.org (not Openoffice.org)
there are / there is confusion
“for peace of mind there a number of organizations that continue offer”
etc. etc. etc.
Otherwise it’s well written, but if you want it to look professional and clean then get someone to go over it!
Overall good and honest article. It’s not going to be easy for the Linux to gain any share of the Enterprise Desktop market. I agree with the author. Lack of good/solid professional applications written for Linux is the problem. I spend a lot of time video / graphics editing. Professional tools I use are not easy to understand. I had to buy books to figure out how to use them. There ARE NO books available MainActor and Jahshaka. I don’t even know if I’ll be able to install these applications in the first place. This is one of the reasons that stop me from using Linux.
I wanted to install Debian. As many times I attempted I failed every time. Try searching for any Debian books on amazon. There is NONE.
Tomcat – A stripped down version of Apache used primarily as a lightweight web server.
Not quite. Tomcat is a servlet container, used to run JSP or Servlet web applications. It does not share any code with Apache HTTPD (which is the Apache you are thinking of), and can hardly be considered lightweight since it needs a JRE to run. Tomcat can be integrated into Apache HTTPD, or run by itself as a webserver. It is usually daisy chained/proxied into Apache because Apache is much better at serving static content.
i think this author is hot. He gives credit where it deserves and i look forward to more of his upcoming articles
I wanted to install Debian. As many times I attempted I failed every time. Try searching for any Debian books on amazon. There is NONE.
Huh? Are you sure your spelled “Debian” right? I tried an Amazon search and got 307 results (admittedly, not all specific to Debian, just had Debian named in their description, but still). Try looking again. I recommend “The Linux Cookbook” and “Linux Administration Handbook” (covers RH and SuSE, too).
Anyway, about the article. It statesv (pg.1):
the main distributions are Redhat, Mandrake, and SuSE. Really? A quick Google for “linux distribution poll” netted me
http://www.desktoplinux.com/cgi-bin/survey/survey.cgi?view=results&…
http://www.osnews.com/story.php?news_id=3640
both of which show Debian at/near the top. Why was Debian ignored? Perhaps if it wasn’t you would have not made this erroneous statement as well (pg.2):
A good example of this is an application (A), which requires for package A to be built, packages B and C are also necessary. However where the trick comes in, is that for C to be built you must get packages D, E, and F.
apt-get install <package-name>
will automatically retreive and install ALL related packages (after, of course, showing you a list and getting your permission). Nothing could be easier. Easier, in fact, than most auto-installing Windows apps I’ve come across.
Cheers
Kudos…very fair and balanced article.
I really like how you used the term “trade secrets.” IBM authored the code in question, and it is GPL licensed (free for anyone to use). Even Darl McBride can’t get the terminology regarding his own company’s lawsuit right.
Actually SCO is asking for $699 per processor. Considering the fact that SCI doesn’t own the rights to the code in question, this is extortion. But that is a story all to itself.
What Americans choose doesn’t amount to a hill of beans, the future is being decided in China right now. You will do what you are told to do by your employers, and you will think what you are told to think, how else did you become a supporter of a monopoly?
I thought the SCO lawsuit was blown out of proportion in importance and quoting anything from Rob Enderle is undoubtedly a questionable practice as the man is a Microsoft stooge. Otherwise, I thought the rest was fairly good.
One and a third billion people, and an economy that grows 8 percent per year. China is going to steam roll the United States, and they will not be using Microsoft.
The truth is that the Unix (and unix-like OSes) is the future !
Even if Linux fail, all investiment made to learn Unix is not wasted. You can use FreeBSD, NetBSD, OpenBSD or any other Unix-like ou commercial Unix (including MacOS X too).
People should not be slaves from a monopoly. Windows is a proprietary and closed operating system, totally different from the true standard of modern operating systems: the Unix.
Unix-based technology is the future.
“there is only a couple hundred viruses for Linux most of which are not nearly as severe as their Windows”
What? Couple hundred? Where does he get “couple hundred” from? I have yet to see a single virus or worm for Linux in over eight years of usage. They simply don’t exist in the wild.
I think this was a well written article. I think the author has some very good points about the role of Linux in the future. And for all you MS people out there…if China, the biggest nation in the world is primarily using Linux (or other non MS software), isn’t that a sign of things to come? Keep them articles comming!
Linux has more like 75% of the desktop as it gets deployed in China.
I don’t think that is right. That Chinese goverment deploying Linux doesn’t mean every Chinese will use Linux on his/her desktop system. As I have heard many times they can get illegal copies of Windows at a very very low price (like few bucks). Why on Earth they want to use Linux instead Windows? They already use Windows programs intensibly. Also for games young people not gonna switch from Windows to Linux. Goverment may use Linux though, not individuals.
Do you have any proof?
it’s obvious the guy hasn’t gained much experience with linux though, at least he’s not researched packaging. sure RPM based distros can get stuck in dependancy hell, but debian and gentoo obviously have much less of these problems if at all. the statements about buggy code is true for sure though, that’s why i like debian stable and older versions of slackware. get past the eye candy and go for stability. I could see linux as a desktop solution, but people fear change, plain and simple. throw a change like from windows to linux and if the people (you’ve essentially interrupted) don’t pick it up and run with it, productivity could be hurt dramatically. That’s the real issue about ‘making the switch’.
Like I said before Linux is good, but it will never be anywhere near what Windows is today.
————–
A lot of people think its a lot better than Windows is today.
Really come on, Windows on the Server and Windows on the end users pc makes any network easy to work with and it saves companies money in the long run.
>>>>>>>>>>
Windows is easy? Hardly. Ever since I set up my first Windows network with Win95, I’ve never been able to get the damn things to work properly. There are always annoying issues with file sharing or printer sharing. I’m about as advanced a PC user as you’ll get, and I still can’t get WinXP to print over the network to our Canon i850. Hell, just transferring files can be a pain. In our college dorm, we’ve taken to using AIM file transfer for everything because Windows file sharing is just such a pain.
Why would you not want that? Plus you get good techs for cheap because every techie knows windows.
———–
A cheap tech is worth exactly what you paid for him.
Having a Linux guru would mean companies would have to pay more for a skilled Linux zealot.
————
No, it would mean you’d have to hire someone with UNIX skills. For the enterprise, there are tons of old guys who still have UNIX skills, and for small businesses, college students with Linux experience are a dime a dozen.
“For companies into digital movie editing/production, MainActor and Jahshaka provide professional grade multimedia editing capabilities, and are quite easy to learn.”
Have you actually used these programs? If so, have you actually used some competing products like Premiere, Final Cut Pro, Shake, or DDC? MainActor and Jahshaka are not professional grade multimedia editing applications by any stretch of the imagination. They are prosumer at best. I seriously doubt your going to find Lucas Arts running MainActor or any other “companies inot digital movie editing production”.
“What this means for IT managers everywhere, is that they can download a completely free fully licensed operating system and distribute it amongst as many computers as they wish.”
Thankyou. I’ve been saying for quite sometime that people around here have this idiotic notion that random downloaded ISOs from some distro website = an IT infrastructure. I was hoping that someone would finally come out and say it since it is what all the linux seem to think in the back of their heads.
“Customized security programs such as Bastille can lock down your system into a very secure state within a couple of minutes, and provide much better security than the default servers built into newer version of Windows.”
Let me get this straight, you’re going to compare using customised security programs that are obviously not installed nor enabled by default in linux, to a default built in server on windows. I don’t even like linux all that much and I can think of better ways to show linux is more secure than that mess of words you’re calling an example. Why not mention something like support for CFS, competent usage of acls, or great patch response time? It would be a hell of a lot more effect than stating the equivalent of “windows is more secure than palm OS because Norton makes anti-virus software for windows”.
What are you talking about? UNIX is as closed as Windows.
SCO owns the code, and OpenGroup owns the trademark/specification.
Sure, there’s lots of clones, but the UNIX IP is not open.
> Linux has more like 75% of the desktop as it gets deployed
> in China.
Not when you speak of 90% piracy rate in China.
Even leave that out I don’t see Linux has 75% market share or install base yet. However, the coming years for JDS from Sun will be very interesting for China.
> “For companies into digital movie editing/production,
> MainActor and Jahshaka provide professional grade
> multimedia editing capabilities, and are quite easy to
> learn.”
>Have you actually used these programs? If so, have you
>actually used some competing products like Premiere, Final
>Cut Pro, Shake, or DDC? MainActor and Jahshaka are not
>professional grade multimedia editing applications by any
>stretch of the imagination. They are prosumer at best. I
>seriously doubt your going to find Lucas Arts running
>MainActor or any other “companies inot digital movie
>editing production”.
Shake is available on Linux, but the price tag of Linux version vs that of Mac OS X version will make people think again: you would rather buy a G5 “for free” rather than buying the double-priced Linux version.
Fairly speaking, there is not a single open source Media app really designed to rival commercial app quality on Linux; But when you have several open source apps work together you might just get what you want from a commercial app at a much lower cost.
In so far, SCO’s lawsuit is overhyped.
The evidence on code showing is lackluster, the only uncertainty to be worried about is SCO-IBM’s contract issue.
Take a look at this article by a real IP lawyer:
The SCO Litigation: Maintaining Walls Around Trade Secrets or Attacking the Knowledge of Those Outside the Walls?
by Douglas L. Rogers
http://www.vssp.com/CM/Articles/Articles1016.asp
Really I think 10% is quite a conservative figure given that China is very heavily into Linux and they have a population of something around 2 Billion. This figure was meant to be a worldwide estimation.
China has a population of 1.2billion and Indian has one of 1.0x billion; Indians “billionth” child was born last year and there was a lot of hoo-ha made about it.
I don’t believe at this time that there is little else but estimations on Linux market usage. Does everyone who runs a Linux based system report that they are using Linux? This was merely meant to be a stabbing guess.
The better question is how many people use Windows; not as a platform to run a telnet session and do 99% of the worth with, how many actually USE Windows. Go into ANZ, all of their systems run Windows, yet, the only application ever open is a dumb terminal session to the mainframe.
Out of the “millions” of installations, I would be very curious to know how many of these systems are used for more than just being a low cost thin client. If the operating system is being used as a thin client platform and all the work is done on some other one, won’t it be there for logical to assume that there are alot more mainframe and UNIX users out there than what the so-called “analysts” say?
Hi
Google search results are not a good estimate of the amount of linux desktops deployed. Its is a result of how many people search for something like “linux desktop”. Present better results next time
In what way is this article a research paper? What has the author researched? I think reading a few news items on Linux hardly qualifies.
Tomcat being a stripped down version of Apache mamde my day. Is obvious that this guy knows what he’s talking about. I wonder if he uses Tomcat as web server only. I hardly doubt that someone will ever use Tomcat just to display static content via coyote.
Poorly researched.
“I am glad the days of paying techs $95,000 a year, letting them get away with slack dress code, and doing basically anything they see fit for the IT department are over.”
Yeah, this is what the business weenies are trying to do but in the end everyone will eventually see that companies need the techies more than they do the overpaid execs. As more engineers rise to higher positions in the company (due to technology becoming more and more integrated into business) you will begin to see programmers/administrators begin to get the compensation and recognition they deserve. Face it, without them most companies today would not run (not for long anyway).
“Over the last 20 odd years, Linux has progressively built on its code foundation due to generous donations of code by single programmers, groups of programmers, and companies.”
Assuming we’re talking Gregorian Calander years that number should be halved as Linux hasn’t been around since 1983, but rather 1991.
While not being surrounded with FUD, this article certainly skips more than a little on the details of installing and supporting linux.
It doesn’t address how difficult linux is *still* to configure for a variety of hardware, and how linux could address single-sign on for example in as simple as fashion as microsoft has.
Remember kids:
Adding value doesn’t mean doing the same thing a different way.
Linux adds value when it doesn’t crash. This means its no better than Windows 2000 or Windows xp. It is better than windows 98 or ME.
Linux adds value when you can easily manage and expand the network. Its not there yet.
– Microsoft Fan
“Google search results are not a good estimate of the amount of linux desktops deployed. Its is a result of how many people search for something like “linux desktop”. Present better results next time.”
Wrong! – the figures come come from the operating system the browser identifies itself as using when a google search is performed.
Linux use may be underestimated by the fact that a proportion of Linux users masquerade their browsers as IE running on Windows (This is exceptionally easy with Konqueror but may be done with the Gecko based browsers as well) in order to get into sites that put up an IE only barrier.
Linux use may be overestimated by the fact that Linux users may search Google more than Windows users who are more likely to be ignorant and use MSN instead.
So it probably balances out.
It doesn’t address how difficult linux is *still* to configure for a variety of hardware, and how linux could address single-sign on for example in as simple as fashion as microsoft has.
Humm, as a microsoft fan you should know better. It’s just as hard in a mircosoft environment to have a end-to-end single signon solution as it is in any other OS. Oh you mean the Active Directory stuff? Yeah, indeed, that lets you sign onto the network, but you seem to forget that infrastructure is designed to support applications, and those in general couldn’t care less abou single signon…Oracle Forms login boxes, anyone?
So, blame the applications first for not honouring anything standard with respect to single signon technology, and then start complaining about how one OS does it “better” than the other. I fact, I dare to contend that the situation in the linux world is much better with a *lot* of services allowing for LDAP backends for their authentication than the Windows world.
A stripped down web server….
You have GOT to be kidding me.
Is this person employed? If so, why so?
-gc
Well I think some of you are clearly overshooting the point of this article. Its meant to be directed at an IT manager. If you were an IT manager wouldn’t you want to know the worst case scenarios for any major switch you were about to implement? As for minor things like Tomcat definitions being wrong and a few slight grammatical errors I still think this is a good article. The problem with the Linux community in general is that it can’t take criticism. This guy is blatently stating that Linux is a good O/S and its getting closer to adoption on a mass scale… and all you guys can do is bitch. I don’t know if you’ve noticed but no other editorial has even bothered including their sources in a proper bibliography, or even bothered to include a glossary. A lot of time and effort went into this and I think this article brought up some very good points. A job overall well done.
I too think the Article is honest. He made good points about the growth of linux in other countries, especially linux’s potential in countries in Asia (China) or the middle east, including the difficulty of installing linux an its related sofwares along with the SCO lawsuit. However, I think the point he made about “linux will never be anywhere what windows is today” can not be determine now. Besides, if the new release of Microsoft Longhorn in 2005 doesn’t meet expections (e.g security), microsoft will experience a backlash by companies looking for alternative sofwares. I have seen many System Desingers switching to LindowsOS as a viable option to microsoft deskstop operating system on their netwok (e.g not for server purposes). Many IT consultants express their satisfaction towards the CLick_N_Run feature lindows offer–which automatically install software packages without any configuration on the user’s part. Check it out for yourself at http://www.Lindows.com ! It is true Lindows is not free, but for $59 dollors you get a lot more than you get on Microsoft operating system alone, remember there is no license fee per seat, just buy one copy per computer and you are ready to go. In addition, you could buy a membership for $14 dollors a month and get access to thousands of opensource softwares at the lindows software warehouse. Some of the powerfull enterprise softwares carry additional charges, but they worth the headache of installing software under linux and don’t have to worry about additianal charges for downloading the same software. Also, if you want to run some microsoft compatible software under linux, just buy Win4linux, it runs many of the microsoft compatible sofwares with no license fee. Works for me! The bottons line is, companies are looking for choices and trying to lower their TCO (e.g Total Cost Of Ownership). As far as ‘the days of IT professianl making $90,000 and walking around with baggy jeans are over’, remember this, companies will rather pay $90,000 per IT personel in other to save $500,000 a year in TCO (.eg Total Cost of Ownership) trust me. Finally, its a matter of time before the opensource movement floods the IT industry. In response to your statemet in your analysis I quoted above, we have to wait and see!!!
No doubt SOME of you guys can’t take criticism. Freaken out about little things like “stripped down version of Tomcat”. I have better things to do with my time than pick out small details like that.
The author and several commentors expressed a wide range of opinions as to just how ready open source software is for everyday commercial use. That doesn’t mean that any of the opinions are wrong. It means that different pieces of software are at wildly different stages of maturity. But there’s a logical reason for that.
The parts of the system that code hackers care about got done first. The actual operating system — kernel, utilities, C compiler, windowing system, and so on — is a cast-iron industrial-strength heavyweight. Almost unnoticed, it’s quietly taken over the lead in kernel technical development. That happened precisely because the source code is accessible to anyone seeking to add capabilities or correct errors.
About 5 years back, those components were stable enough to go to work in corporate network infrastructures. That’s when news started trickling out about Linux and BSD print servers, file servers, web servers, and so on quietly carrying the load, out of sight of upper management.
With the bare operating system proving itself in a business environment, development priorities began shifting toward applications, particularly business applications. Hardware manufacturers, eager to lower their cost structures and selling prices while improving margins, saw the benefit of developing only the features they needed, instead of the whole system. Major data base vendors saw an opportunity to reach a wider market by porting to less expensive hardware and software platforms, without compromising reliability. Other companies saw strategic reasons to throw previously closed applications open to cooperative development — so now we have complete open source office suites that major corporations and entire governments are bringing into mainstream service. Some of these still have noticeable rough edges, particularly in their backwards compatibility with closed-source equivalents. But they meet most definitions of “good enough” now, and will continue to improve. They also present special advantages, such as stable and fully documented file formats, which are of critical importance for permanent public records.
Consumer applications, such as manipulating sound and video streams, operating CD burners and DVD drives, got started much later. Quite a bit of this software is in very early alpha stages, and it doesn’t help any that hardware device manufacturers give little or no help, even to publishing technical specifications. This is one area where tales of frustration, failure, and devices that just don’t work are still common.
Specialty technical software for open source platforms is probably the least mature category. It’s hardly gotten started. Commercial vendors of truly professional software, such as packages for desktop publishing, printed circuit board design, chip design, and mechanical drafting, are still focused on Unix and Microsoft platforms. Open source developers haven’t gotten where they are by waiting around for commercial suppliers to meet their needs, though. Projects are under way in many application areas. Some have achieved at least minimal usability, others haven’t. Probably the most outstanding success of this kind to date is The GIMP, often compared favorably to Adobe Photoshop. Much less far along, but good enough to get actual work done, is Lilypond, which does a beautiful job of music typesetting. In chip design, there are projects that show promise, and some programs that do parts of the project, but nothing that can handle the whole job. A few years out, when more of the installed base is running on open source platforms, the commercial application vendors will have to follow in order to survive. That’s already starting to happen in high-end chip design software. But mostly that’s in the future.
So… is open source software ready to be a workhorse in your business? The right answer is, it depends on what you want to do.
If you want it to run your on-line web sales system and your corporate data base, absolutely.
If you want to run standard office applications such as word processors, spreadsheets, web browsers, and so on, yes with some qualifications. Documents that use particularly obscure features of legacy systems may not render the same or be fully functional. Some web sites with abnormal, browser-specific code do still exist, though they’re getting less common. Other than that, users presented with something like OpenOffice on KDE have sometimes hardly noticed that they weren’t on their old familiar systems. Expensive retraining for end users? Baloney!
If you want to edit video clips and burn them to DVDs, well, probably not. Right now doing that on Linux as a primary business process is for major movie studios, who have systems specialists on staff. Same thing for running video games; there aren’t that many for open source systems, and the kernel infrastructure for that is still in active development.
(By the way, look a little deeper than salary comparisions to get a handle on relative support costs. Unix and Linux sysadmins don’t have their time consumed recovering from crashes. Consequently their users don’t spend their time waiting for systems to come back up. So these support staffs are smaller and handle more machines, they have more time to add new equipment and features, and they tend not to incur overtime.)
Any accurate statement about what’s ready for business use and what isn’t, is necessarily a snapshot. In six months the answers will be different. In a year they’ll be different again. So by all means evaluate what’s out there against your own requirements. Run small-scale trials before committing to large-scale deployments. Check back periodically. And if an application appears to be almost ready, except for one important thing you need, you aren’t stuck waiting around the way you would be with closed-source, vendor-controlled software. You have the option of adding the feature in-house or contracting with a local consultant. You can get it done according to your own needs and schedule. That freedom is what open source software is all about.
“Tomcat – A stripped down version of Apache used primarily as a lightweight web server.”
This kind of comment does a fine job to help evaluate the author’s knowledge. Add his description of dependencies and what he is considering to be major distributions, and this evaluation becomes more precise.
One may hope that his assumptions are based upon something more reliable than his shaky grasp on Apache and Linux.