We all know about the benefits of open source software in lowering total cost of ownership, offering more choice and the increasing quality and functionality of the code. The rise of Linux as an edge server and now migrating toward the data center is clear validation that the open source model has taken root. But there are still significant barriers to overcome before Linux and other open source projects are broadly accepted across enterprises, says ZDNet’s Dan Farber.
Six Barriers to Open Source Adoption
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
I think that much of the article is based on intrinsic factors that can’t be helped and probably won’t be changed. By its very nature Open Source has been and is implemented by folks who volunteer their time when they have it. Thus, a roadmap, a systematic support system and a predictable release schedule cannot be guaranteed. Red Hat was probably very smart when they decided to focus on corporate service and the support of legacy systems. I think a ot of companies in Linux IT service can take these ideas and really make good business out of them. Very compelling article.
If you’re trying to explain to your boss why you should use open and available solutions then yeah, “Open Source” is a great thing to be able to talk about. However, if you’re trying to explain why it is fundimentally wrong to prohibit people from understanding how the everyday things that surround them work and prohibiting those same people from helping each other, “Open Source” is a pitiful term for discussing issues which are primarily about freedom.
The “Licensing caveats” sections shows just how much the author cares for Free Software.
The rest is pretty boring.
Frankly, and unfortunately, many commercial entities and individuals aren’t interested in the ideals of free software. It’s sad, but it’s true. Among them are people who use free software and at the same time claim it sucks neither helping to improve nor even bothered about it.
Overall, I think the articles gives an interesting analysis of the barriers against open source software adoption. I don’t think open source developers need to make the adjustment, however, the ball is in the court of the entities interested in open source.
For example, if there is a potential security vulnerability in the kernel, you can’t insist the kernel developers release a patch only when you are ready to upgrade. You’d simply be ignored. The entities will have to work around the fact that open source software changes at sometimes unbelievable paces, and those entities will need to intelligently improvise means of keeping of with such changes.
In modern open source projects and distributions, updating software packages hardly has any visible effect on the deployment. The open source model is better with regards to updating packages in that reboots are never necessary, except when upgrading the Linux kernel. The Windows mentality of Update/Upgrade == reboot==lost productivity is almost totally absent in many open source projects and distributions.(I’ve been running some servers for months with many updates and without any upgrades or downtimes, well except for one hardware failure in one server machine.)
So if your entity is lucky, and not hell bent on installing the latest kernel, you can keep a whole network of Linux computers running without having to reboot them for months. In my mind, it is better when compared to the alternatives that insist your reboot the whole network every month for the next security vulnerability patches or minor updates.
There are tradeoffs, but it isn’t the responsibility of open source to identify them for potential commercial entities, it is the entities duty to carry studies that evaluate how they can incorporate open source into their operating environment.
The article makes some good points, especially considering that it was written by a “wolf in tie and short sleeves”. I think we should stop looking for enemies among those we disagree with.
The term “open source” isn’t itself that bad when separated from the bad reputation its advocates often attach to it. That is, open source, when used as a term to enterprise-types, should refer only to the open status of the source code, not the nebulous movement surrounding it. You’ll know you’ve gone too far when you start saying the “free…beer…speech” nonsense and your boss goes “HUH?”.
Open source will be adopted like wildfire when it becomes profitable. Not for the programmers or IT staff, but for the company as a whole. An earlier comment correctly noted that this isn’t “open source’s” responsibility to create profit for corporations. True enough. But it is also not the responsibility of corporations to change their value systems solely to be consistent with their choice in software. Its like two people not getting along. It doesn’t mean either is right or wrong, just that their values aren’t the same. Neither can necessarily change the other. Corporations aren’t “evil” for rejecting OSS solutions; nor is OSS of second-rate quality because corporations don’t accept it.
Open source or not, I couldn’t really care less…
Is it the most economical benefitial tool for least money which takes least time with the right results?
Good, that’s the solution we’ll choose.
Open Source? A bonus placed far down the ladder… if it’s free then it’s good for the calculation above since the economic reasons is right. This tends to drive me to choose BSD quite often.
Open source tends to do very well on the bottom levels of the software stack where infrastructural components like Apache and MySQL can get improvements from those not necessarily in the business of software. People that are in the business of software can compete with Microsoft by leveraging open source components with proprietary software that ties everything together for those PHBs that are nervous about the mismash of components that are open source.
As others have pointed out, if PHBs start hearing some RMS-inspired rant about how source code is equal to civil liberties then they’ll be turned off and less likely to listen to reasoned logical persuasive arguments on how open source will result in lower TCO.
Open source currently thrives upon the willingness of contributors to do things for free.
But the moment you start propagating open source as a business model, you run up against walls. People need to get paid. They can’t work 40-hour weeks for free and survive on intellectual gratification alone.
Thus, a price tag needs to be assigned to whatever the collaboration yiels. But how can you charge for a product that’s available to everyone, everywhere?
Even the companies that build their products around open source (Red Hat, for example) have some kind of proprietary added value that enables them to charge a fee.
Open source will never replace the capitalist business model. It can only hope to compliment it.
The key is the value-added proprietary software at the higher end of the software stack. What you can get is open source lower stack level software like Apache and MySQL getting trickle down benefits from the proprietary vendors. You can also get businesses that aren’t in the business of software contributing and benefiting from open source.
I don’t know why you’re saying “You’re all wrong”. I was the one that said “co-habitation of open source and proprietary”.
I don’t understand “Open Source Adoption”!
I adoopted it. Like me there are many people. Probably the title of the article should be changed.
And yes, it will take some time when OSS will step into enterprises just because IT Dept. chiefs preffer to blame somebody(MS,Oracle,Sun etc) when something is not good, instead of working for what they are paid for.
Perhaps someday you will understand that there are other people in the world who have not adopted yet, and that the world does not revolve around you.
Yada, yada, yada. Open source is not ready for the enterprise (Google, Amazon).
Yada, yada, yada, Open source does not scale well (SGI 256 CPUs, 512 by year’s end)
Yada, yada, yada, there is no support (Novell, Red Hat, Mandrake, HP, IBM, SUN and tons of solid consultancy services and thousands of LUGS, depending on the level of support you need).
Yada, yada, yada, there is no way to make money from open source software. It’s about freedom, not cost. Again, it’s about freedom not cost. A supported option with 24/7 live support will cost you.
Yada, yada, yada,business types do not get open source and only value its cost. Nope. Open source is a strategic move towards independence and ownership of your IT future. A few business types that get open source (Bank of America, Goldman Sachs, Credit Suisse First Boston…)
You people can go on deluding yourselves. I have one word for the naysayers. There is still time to learn new things. Develop new skills or you may soon be out of a job. At my Linux User Group, we have more job offers than the 85 users can handle and the paper is full of jobs for linux developers and administrators. Get with the program…
It may be quite difficult to talk about open source as a single entity, and the obstacles open source as a whole may or may not face. We have too many sorts of different server, embedded, desktop, OS sofware etc.
My opinion is that there is more and more demand for basic operating system sofware to be open source. However, a great deal of business software, and also lots of entertainment sofware like games, are going to be proprietary in the future too. If the open source community cannot produce (usually more or less for free) a good enough version of some needed software, someone will make a proprietary version and sell it for a good sum of money. What could change that fact?
As to operating systems, I think more and more people are simply tired of proprietary OS monopolies. And also the open source OS model allows people (like for example Hollywood movie makers) to tweak the system internals a lot more than any proprietary OS allows. And it is possible to look for security bugs and tweak the security levels too. With an open source OS it is possible to make customized like embedded versions of the OS etc. – and without the need to any licensing money. With proprietary OS’s, you’re left with what the proprietary OS maker cares to serve you – and if you want something special, it will costs you a lot, if it is possible at all.
Read the many articles about why Hollywood movie makers have chosen opne source Linux OS (but proprietary movie software on top of it), or why also proprieatry OS manufactures try to open their closed source more at least to some of their customers. Such high tech computing and OS trends may tell a lot about the OS future as a whole too.
Well said. It’s high time to see some more comments like this, because it is much closer to the reality of what is right now.
I personnally feel a lot more confortable with the perennity of a vast number of open source projects than on the willingness of proprietary companies to continue supporting some of their products when they become a cost rather than a source of income. Some roadmaps are not worth the paper they are written on. Remember, a roadmap is not a contract. A commercial company is in no way bound by such fuzzy marketing commitment.
I know the Linux kernel will always be around, same with Apache, Gnome and Kde, ssh, Debian and plenty of others. In the banking industry, there are a lot of software companies who go belly up, leaving their clients without support and without the ability to contract support with anyone else. And very often, the support they get is, well, pretty bad or pretty expensive and sometimes both.
Of course, prudence is the rule if selecting an open source project : it might just stop being updated, it might take a direction you don’t like etc….. But the same applies to proprietary software, and you don’t get the code if things go wrong. Again in the Investment banking sector, some consultants specialise in helping banks select banking software, not only based on feature requirement but also helping them manage the risk of dealing with companies that are very often private and on which information is not readily available.
So I agree with doggedblues : a lot of the objections are rubbish. A lot of very serious financial institutions rely on perl, Apache, Mysql, Php and Linux for what really matters, also because it will be around a lot longer than, say VB
And if any of these objections had been guiding factors for the “business types”, Windows would never have replaced Unix on trading floors (where stability is FUNDAMENTAL) before the emergence of NT4 that was the first somewhat stable realease of windows. Everyone migrated starting from windows 3.1 because it was just cheaper, period. Open source project can’t afford compensating unreliability with marketing effort.
Also, stop giving us the “license cost doesn’t matter”. I happen to have seen the yearly MS bill for a few medium sized companies (for desktops) and it runs in millions of dollars. I don’t imagine that they provide support for that much. That’s outsourced to another company and we have a bunch of guys running around to troubleshoot PCs because they can’t do it remotely (except password resets). That doesn’t come cheap.
A business will adopt open source when the costs to them are worth the change over. If business IT staff finds a bug in the kernel, they can alert the kernel team and patch it himself. That is called security through many eyes.
Linux is on more embedded electronics than anything else. Why??? because it is inexpensive to change to meet your needs.
Security updates are only more numerous when you factor in that you are update not a OS but a whole machine filled with 1000’s of software titles.
Businesses don’t care about zealot’s ideals until they piss off the zealots. Take EV1 they have lost nearly 1000 hosted sites just due to their buying a SCO license. That is a company who pissed off zealots. Cisco is breaking copyright law with regards to GPL products. The zealots are there to make sure that their work stays around with Liberty. As long as the businesses respect that there won’t be probelms.
Open Source? A bonus placed far down the ladder… if it’s free then it’s good for the calculation above since the economic reasons is right. This tends to drive me to choose BSD quite often.
But if you’re just using the tool (as opposed to enhanceing and selling), does it really matter which licence you choose? After all, the only restrictions on code under the GPL are your responsibilites to provide the source. If you’re using it ‘in-house’, how does it matter?
But, overall, I agree with your point. As a manager, your responsibility is to make sure the company is a profitable as possible. If using OSS is the best route to that goal, use it. If closed source can provide a workable, turn-key solution, then, please, use that. The problem I find is, most ‘off the self’ software solutions provide only about 15% of the functionality you need. The other 85% just sits unused, or, worse, serves to confuse the user.
Anyone ever use Harvest? Its a really horrid pieace of closed-source development management software. At least if it where opensource, a company could trim out and extend the functionality that it wanted (or did not want).
I work for a large telecommunications company in the US, and they have a long standing policy against using opensource software (although, they also make the mistake of putting opensource in the same catagory as freeware and shareware. There is a difference. I just wish the pointy-haired bosses would listen why I explained it).
Recently, we needed to revamp our intranet forum. Through some serious arm-twisting (and rule bending), I got them to use phpBB (deployed on a Linux box. They bent the rules because it was determined that a forum was not ‘in the critical path’, and OSS would be okay for that). They weren’t going to let us use it, soley because it was OSS. But, I showed them phpBB in all its glory, and challenged them to find me a better, non-OSS solution, that wouldn’t have cost someone their pention. They couldn’t. Many of the commercial solutions where far inferior (especially the Java ones) to phpBB. Hence, the ‘Best Tool For The Job(tm)’.
Perhaps thats the tact that OSS proponents should take. Don’t push OSS because its ‘open’, push it because it can be ‘Best Tool For The Job(tm)’.
… feel a lot more confortable with the perennity of a vast number of open source projects than on the willingness of proprietary companies to continue supporting some of their products when they become a cost rather than a source of income. Some roadmaps are not worth the paper they are written on. Remember, a roadmap is not a contract.
I know the Linux kernel will always be around, same with Apache, […] In the banking industry, there are a lot of software companies who go belly up, leaving their clients without support […] And very often, the support they get is, well, pretty bad or pretty expensive and sometimes both.
[Sorry for chopping up the 2 paragraphs, I just wanted to get the meat of it]
I think you have misinterpreted the importance and function of roadmaps in companies buying/using software. The whole reason a company would want a roadmap would be to plan future purchases of not only software, but also hardware, and to plan the deployment process for that hardware and software. The company I work for is still developing their deployment plan for Windows XP, and implementing the infrastructure that will support that plan, and this is an OS that came out 3 years ago and was well marked on a roadmap, probably for an earlier release date. Roadmaps slipping and companies folding can be a problem, but the real issue is wanting to have a good idea of when major changes are going to be made to the software, and knowing when and how to plan for those changes.
On the other hand, I think it’s somewhat overstated as a problem, because you can generally get a feel for the development plans of OSS if you keep an eye on the development community for a particular piece of software. Most of the distributions of Linux are built by companies that should be able to give their customers at least some idea of what a roadmap would look like for their distribution, especially since most of the distributions will lag behind the latest releases slightly to maintain a stable branch to put on shelves.
Of course, prudence is the rule if selecting an open source project : it might just stop being updated, it might take a direction you don’t like etc….. But the same applies to proprietary software, and you don’t get the code if things go wrong.
Just a note: try not to make the mistake of equating proprietary software with not getting the source. In my experience, almost all proprietary software is open source if you have enough money to throw at it, and this is even more true if the company is liquidating their assets (because they’re folding).
Again in the Investment banking sector, some consultants specialise in helping banks select banking software, not only based on feature requirement but also helping them manage the risk of dealing with companies that are very often private and on which information is not readily available.
And there’s probably a very wide-open business sector for consultants to recommend long-standing open source software that can be used under a reasonable assumption that the software will be maintained for some time to come, without having to resort to maintaining it yourself.
Also, stop giving us the “license cost doesn’t matter”. I happen to have seen the yearly MS bill for a few medium sized companies (for desktops) and it runs in millions of dollars. I don’t imagine that they provide support for that much. That’s outsourced to another company and we have a bunch of guys running around to troubleshoot PCs because they can’t do it remotely (except password resets). That doesn’t come cheap.
Most of our support and all of our license management is outsourced. I’m sure the bill runs into the millions (since the company I work for is quite large), but the license costs only really matter if they’re paying the license for software that isn’t being used. The really sad part is that all of our PCs came with Windows XP installed, and they were all cleaned and had Windows 98 installed on them (with a couple of 2000 machines), and then recently all of them were cleaned again and had 2000 installed. The XP licenses are still being held, though, because as I said at the beginning, they’re still planning the transition to XP. Somehow, though (terminal services…), they still manage to do remote support just fine. The only reason we do any local support is so that people can get back up to speed on little problems without having to go through the help desk.
“Businesses don’t care about zealot’s ideals until they piss off the zealots. Take EV1 they have lost nearly 1000 hosted sites just due to their buying a SCO license. That is a company who pissed off zealots.”
Bad example. Even non-zealots find SCO’s Linux licensing to be a shakedown attempt with a poor legal basis, and because of the nature of the GPL, it is not at all clear that buying a SCO license would allow one to legally continue to use Linux if SCO by some fluke won its suits. So EV1 spent a lot of money to end up (1) setting a precedent of giving in to a “protection” scam, and (2) put itself in a precarious legal position that could monkey-wrench the customers it was supposedly trying to protect. One does not need to be a zealot to be mad about this.
“Cisco is breaking copyright law with regards to GPL products. The zealots are there to make sure that their work stays around with Liberty.”
Again, bad example. Is is zealotry to protest breach of copyright?
You, I am sorry to say, are wrong. Redhat does not provide a ‘proprietary’ addon to make money. They provide a service that is basicly support.
It is one way to make money from open source and RH seem to be good at it…
“Redhat does not provide a ‘proprietary’ addon to make money.”
Except for 2 pavkages containing artwork and logo’s. Those are trademarked and perhaps copyrighted. With SuSE opening YaST up they’re heading to the same way, perhaps they’re already there.
“In my experience, almost all proprietary software is open source if you have enough money to throw at it, and this is even more true if the company is liquidating their assets (because they’re folding).”
Now someone try to tell me Open Source means the same thing as Free Software.. C’mon.. just try. Just try to tell me which one convays Software Freedom better? Which one is more easy to understand? Because the OSI says that they both mean the same thing. So why does everyone constantly get the meaning of the term wrong?
“Redhat does not provide a ‘proprietary’ addon to make money.”
Redhat is an OS manufacturer, and like I said above there seems to be more and more demand for good open source operating systems. There’s lots of market for others than Redhat too, and those OS companies, at least some of them, may get enough revenue from support and services only. However, other, more specialized software may be totally another story, however.
Does Redhat (or others) have good open source finance sofware for large companies? or does anyone have good open source movie making sofware, like like special effects progs for Hollywood? Or are there many successful open source computer games? In all those fields, and on many others, proprietary software is clearly dominating and I don’t see any change coming in near future.
However, of course sofware like Open Office.org is very good and can often compete with MS Office and other proprietary office suits very well. The biggest problem is not the still big problems in the OOO.org software or some licensing matters etc., but simply what most users are still used to – MS Office… People don’t want to learn new things…
Some people may, of course, also still need some special features that e.g. MS Office have but OOO.org doesn’t, but, for example, I personally don’t need any other basic office programs besides of OpenOffice.org suite.
The same with GIMP, though Photoshop is clearly still better than GIMP, for my personal and web photo manipulation needs GIMP offers me all I need, and for free. And the next 2.0 version of GIMP will be even closer to Photoshop in features.
Open source has lots of potantial in nearly all software fields – including even finance, movie making, gaming etc. software that I mentioned above as examples. But I think that open source has by far the best advantage and most demand in operating system field. What closed source OS can beat OpenBSD in security? Or NetBSD in portability? Or Linux in the speed and wide range of development or in the amount of enthusiastic developers?
“Does Redhat (or others) have good open source finance sofware for large companies? or does anyone have good open source movie making sofware, like like special effects progs for Hollywood?”
“In all those fields, and on many others, proprietary software is clearly dominating and I don’t see any change coming in near future.”
I guess you do not see change coming in the future because Linux use for movie-making by the major Hollywood studios started years ago, e.g.:
http://www.salon.com/tech/feature/2001/11/01/linux_hollywood/
fan.theonering.net/sfx/
If you swim in the river of software and come away with 1% or 10% of the river… over time your have a nice ocean of your own…
If the NHS had been in the drug business at the start in the 50’s 60’s they would of had a lot of knowledge and products on the shelves.. even if its only 1% of uk market…
Think about it walk into tesco / asda and see NHS Paracetamol 40 pence more expensive then own brands (tesco/Asda) most people would buy NHS with thought that they are probably better… and supporting free health care for all…
Linux may of started 10 years late to the software party but they have a model that can catch up… (go back to 2002 or 2000 and see what was on offer) its the fastest development model ever seen fueled by the fastest biggest most open commincations (the internet)…. The treat is it will outpace normal software offerings…
It only needs to get to critical mass where it will go from 5% or 10% and jump almost over a nite (year) to 50% or more..
1% of the whole software industry is a lot of pie….
I make £10,000 profit using proprietary software for the solution…
or make £100,000 profit using Open Source… I am sure I would not mind investing say 50,000 in Open Source..
for the chance to remake the profits… in the next big thing…
“I guess you do not see change coming in the future because Linux use for movie-making by the major Hollywood studios started years ago, e.g.:
http://www.salon.com/tech/feature/2001/11/01/linux_hollywood/
fan.theonering.net/sfx/”
But that was just what I was talking about in my previous message… Yes, Linux is even THE dominating OS in movie special effects making etc. in Hollywood nowadays. But Linux is only the OS (or kernel if some prefer that…;). But what about the specialized high quality movie software itself?
The movie and special effects software used on top of Linux in Hollywood stidios is to a great extent proprietary closed source software. That is why you cannot just apt-get (Debian) or emerge (Gentoo) that same top class movie sofware that those big movie studios use. Often those studios have developed such sofware by themselves using many developers of their own, and the studios are usually not willing to give it away for free to their competitors.
You can edit home movies on Linux quite well nowadays. But even there closed sourse software is still dominating. Though open source home movie editing sofware seems to be getting better quite fast too.
Anyone ever heard of a product called Apache? The most successful (in terms of market share) product on the market.
How you figure open source not accepted or needs to get adopted? IT ALREADY IS!
I guess OSS not the issue, perhaps there’s something else with Linux that bothers people.
>>People need to get paid. They can’t work 40-hour weeks for free and survive on intellectual gratification alone.<<
*Sigh* for the billionth time, OSS has nothing to do with people not getting paid. People get paid at redhat, novl/suse, mandrake, etc. All of those companies are profitable, their employees all get paid.
Free software does not need to be OSS. OSS does not need to be freeware. OSS is *not* public domain. All of that is just msft FUD.
Frankly, ziff-davis has long since impressed me as something for people who can’t get their mind around anything non-msft. The columists as zdnet, are often much less knowledgable than some of the posters you will find here, or slashdot, lwn.net, etc. Also, to me, zdnet seems to have a blantant pro-msft biased.
Free software does not need to be OSS. OSS does not need to be freeware. OSS is *not* public domain. All of that is just msft FUD.
Here’s where you’re wrong… this is Linux Zealots which pump this out, not M$… All Linux Zeals keep saying everything is free and that’s why it’s soooo coooool…
Where are my free airline tickets ??….free automobiles for the people !!!!!….and free housing too !!!
You ‘FREE’ people are a bunch of losers.
As an aside, why does OSNews use the GNU logo for “Open Source” matters?
The FSF is completely against the term, because they feel that Open Source advocates focus overly on pragmatism (the material benefits of code openness) as opposed to liberty.
So – why use their logo? You could just use a resized copy of this (http://opensource.org/trademarks/opensource/web/opensource-110×95.p…) instead, which would be far more appropriate. Then you wouldn’t be wrongly associating the FSF with a philosophy they don’t agree with.
>>All Linux Zeals keep saying everything is free and that’s why it’s soooo coooool…<<
Linux “Zeals” are right about there being free versions of Linux. If they find that cool, then great.
Again: free software does not need to be OSS. OSS does not need to be freeware. OSS is *not* public domain. All of that is just msft FUD.
So please stop saying that OSS won’t work because people need to paid etc. That line of “reasoning” is all wrong.
“Linux is on more embedded electronics than anything else. Why??? because it is inexpensive to change to meet your needs. ”
Wrong, pal.
Linux is almost negligible in embedded electronics.