New OSS Doomed In Enterprise?

Rob wrote to mention a Computer Business Review Online article which posits that immature open source software is doomed in an enterprise environment nowadays. From the article: "Open-source startups and relative newcomers must target a new breed of CIOs, which Graf dubs chief process innovation officers. Rather than old-school CIOs who focus on a company's data management, these guys design processes with the company's network. "If you want to become strategic to the company, you need to deal with business processors. 'The key question for open source is, Which open source technologies are mature enough to survive the consolidation that's coming?' Graf said. 'Linux? Definitely. Eclipse? Definitely. Mozilla? Most likely.'"

Comments

OSS will almost always be doomed in Enterprise.

[Anonymous Coward]

Its all about accountability. Even if Microsoft may not have the best product, when it fails, the suits are able to hold Microsoft accountable. A little harder to do that with Debian, or any OSS without corporate backing. Nobody ever got fired for buying IBM.

Re:OSS will almost always be doomed in Enterprise.

[inter alias]

Read the MS EULA.. They can't be held accountable for anything.

What they offer is paid support. 3rd party paid support is often availible for OSS, but some exec's probably feel it need to come directly from the maker.

Re:OSS will almost always be doomed in Enterprise.

[haus]

Please name one example of Microsoft every being held accountable for their software failing to work as promised?

Re:OSS will almost always be doomed in Enterprise.

[AKAImBatman]

As already pointed out, MS will do very little.

That's not entirely true. Microsoft will sell you an entire army of tech support drones and Microsoft Certified Support Providers. That way your boss can go back to his boss and say, "Microsoft is working on it." To which your boss's boss will say, "I'm glad we paid for Microsoft! Just imagine how difficult it would be to get support if we paid for Linux!" Thus your boss's neck is saved from the chopping block by simply passing the buck.

If your boss decided to keep things internal, he'd have to tell his boss, "We're working on the problem right now and hope to have it fixed soon. We could purchase support from company XYZ to speed up the process." To which your boss's boss will say, "If we're supporting it internally, why did it break in the first place and when is it going to be fixed? Is that third party the vendor? Then how do they know anything about anything?" If he gives the answer, "See, this open source stuff...," he'll hear the words, "You're fired!" before he finishes the sentence.

Of course, your boss's boss may be smarter than that. But many managers won't take the that risk with their own necks.

Re:OSS will almost always be doomed in Enterprise.

[typical]

The point is not that Microsoft can be held accountable -- it's that the purchaser cannot be held responsible.

Corporations are highly risk-adverse in culture.

Re:OSS will almost always be doomed in Enterprise.

[ScuzzMonkey]

Whether they can be held accountable is up for the courts to decide.

Provided you have sufficiently deep pockets to fight it out with the legal juggernaut that is Microsoft's counsel in that venue for a decade or more, sure. But Microsoft is actually pretty good about keeping the few major corporate entities which might do so appeased for their particular needs, so this is unlikely to happen, and for all practical purposes for the vast majority of users, the grandparent is correct.

Besides, at the heart of the argument, they have pretty clearly signalled that they don't intend to accept responsibility just by including that language in the EULA, so that should give pause to anyone who thinks that's an important factor in purchasing decisions.

Keyword = immature

[CodeShark]

Enterprise environments won't use -- or at least should not use immature software from any source -- open or proprietary. In fact, most major enterprises have a very slow and conservative adoption software adoption process to prevent a single application from breaking other existing applications. Software requiring interfaces between systems are even more rigorously tested -- and this is all a good thing.

My thought is that the problem is that few enterprise businesses are assisting in developing the maturity of applications that would enable more widespread use. Every large enterprise has small projects that would benefit from open source tools, etc. out there, but if the enterprise isn't willing to spend the developer resources, then it essentially locks the door to the acceptance of more mature open source tools that are validated "in-house", thus facilitating greater acceptance throughout.

Re:Keyword = immature

[Creepy]

Actually, I'd go one step further - Enterprise environments won't use any software unless they get enough customer demand for it.

The company I work for is no exception - our ex-VP (due to aquisition, now an upper manager) claimed we would "Never, ever support Linux." The reason for that claim was that Linux had no paid support if there were problems. Since that time, we started supporting two Linux platforms that have paid support (Red Hat and Novell [SuSE]).

    That's not the end of it - some of our UNIX customers (specifically AIX and IRIX) started to put pressure on our management to support the "unsupported" Mozilla because AOL was not supporting them well with Netscape. Shortly thereafter, we started supporting Moz. Eclipse support came more as a need - we needed an extensible IDE tool that we could ship for XML editing.

SAP certainly didn't jump out on a limb naming those 3 tools because I know several other enterprise tier companies that use them besides ours.

SAP says that immature open source software...

[s0l3d4d]

relatively immature open source software has little chance of surviving in the enterprise, said an SAP AG executive during a speech at the Open Source Business Conference in San Francisco.

D'oh? News value? 1) immature software has never had good survival rates in the enterprise environment and 2) SAP probably wants to sell SAP software, so even if there was an open source, MATURE application, that would be enterprise strength, to be used where SAP is used, I don't somehow think that SAP would suggest anyone to use that.

Re:SAP says that immature open source software...

[electroniceric]

2) SAP probably wants to sell SAP software

You can take that to the bank. Why is it that this article only quotes one company's VP of marketing, and yet we're debating the premise that there's some wave of consolidation coming, and open source is not yet mature enough to be part of it - and oh, by the way, SAP is trying to sell some new consolidation platform. This is only news in that it's coverage of what line a company is pushing. It'd sure be nice to see some background reporting that establishes whether this claim is even reasonable, but the reporter probably isn't making enough money in this story for that.

Immature is always doomed

[tsa]

Of course immature OSS is doomed in the Enterprise. Who wants to use any immature software where correct functioning is important? Software being open source or closed source has nothing to do with that. At the end of the day performance is the only thing that counts.

Re:Immature is always doomed

[dasil003]

Of course immature OSS is doomed in the Enterprise. Who wants to use any immature software where correct functioning is important? Software being open source or closed source has nothing to do with that. At the end of the day performance is the only thing that counts.

Also implicit in the article is that OSS needs to succeed in the Enterprise. But that is not true at all, open source software is not beholden to investors seeking profit. A project can languish in obscurity for years before it matures and suddenly becomes useful to huge numbers of people. Or a better project can make it obsolete overnight. But whatever happens the code is out there. There's no point in trying to ascribe business life cycles to open source projects.

Businesses are like animals struggling to survive, open source projects are like whole eco systems that fluctuate widely based on climate, but don't sink or swim by single events.

OSS must gain the enterprise in order to thrive

[ficken]

I agree - OSS must gain ground in the enterprise for it to thrive (more). Once good OSS has established itself as reliable and accountable, then the software will gain the respect it deserves. In order for this to happen, the software must be mature enough to withstand 'the beatings' of the suits (think maintenance costs, feature requests, etc). The only immature OSS products we use at work are in-house stuff. It is just a matter of time before the (other) big players in OSS come to the top - not just Linux, MySQL (and more).

Enterprises are not the sweet spot.

[bubulubugoth]

Mostly of them, will use High level, IBM, SUN and Microsoft support, they will pay for it. IBM OpenSource strategy is start using OSS products, grow in demand, and switch to their high class product...

Example, start with bluecode+geronimo and later switch to Websphere+Db2

But, Enterprise, are only a niche market, very well payed indeed, but there are just a few. The other market, medium sized enterprises, small, and micro, those are the sweetspot for Linux, becose they CAN work with "inmmature" sotfware becose theire also "inmmature" bussiness...

That is the figth OSS is winning, with mysql, postgress, apache, php, samba..

Fortune 100 have the money to pay for another Fortune 100 for its IT integration... but again... there is only 100... the other hundreds and thousands of bussiness, those are who need linux to lower costs, add more technology to their process...

Further more... the lack of applications for linux, is a normal step in the madurity of a market.

Rigth now, there are may software houses, developing, specific solutions, and in a few years, will become mainstream solutions. There you have compiere, OpenOffice, they still need work, a lot, but its getting done.

Out of the box solutions for linux are needed to the mainstream, and may are building them...

Microsoft, Oracle, IBM, all must what their ISV which are making crossplataform or linux plataform applications...

Doomed I tell you, Doooomed

[MarkusQ]

This is a crock. As others have pointed out already, immature software is unlikely to be used in an enterprise environment (unless it was developed in house) regardless of the license. But wait, there's more. I happen to have a number of immature open source projects of my own at the moment, and I don't give a flying fig if they "make it" in an enterprise environment.

Why? Because unlike Microsoft, I don't expect any revenue from them and thus won't be disappointed if I don't get any. I wrote them because I needed them and open sourced them because I wanted a few more eyeballs on them. But even if no one else ever even downloads them, I'm not <voice='spooky'> Dooomed </voice> because I'm not selling them in the first place. For the vast majority of open source projects, saying that they won't make it in "the Enterprise" is about as relevant as saying that cows will never use the iPod.

--MarkusQ

Re:Doomed I tell you, Doooomed

[corellon13]

Although I agree with your assessment that the success of OSS is not solely dependant on its success in an enterprise environment, I think you underestimate the importance of success in that environment.

In a perfect world everyone would develop software and approach it the way you do (which I applaud). However, the fact remains that enterprises have a HUGE influence on what is successful and what isn't. Once an enterprise becomes invested in a product, it is in their best interest to support the wide use of that product. As this product becomes more of a standard and widely accepted by everyone (even outside of the enterprise environment), they are basically developing future talent at no cost.

Using your Microsoft reference, Microsoft has been able to successfully create a Windows competency throughout the computer world. This created a kind of standard and demand that really all started in the enterprise environment and spread like a virus to homes and everywhere computers could be found.

All this to say let's not under estimate the value of getting OSS accepted in the enterprise. This can be done with price, if not maturity, as most enterprises are always looking to cut costs and you can't beat free!

Re:Doomed I tell you, Doooomed

[brunes69]

In a perfect world everyone would develop software and approach it the way you do (which I applaud). However, the fact remains that enterprises have a HUGE influence on what is successful and what isn't.

You're not getting it. What is "successful for an open source project? If an authour opens his code, and one other person finds it useful (either as it is, or in another project), then that project iss a success. Basically, it is having a userbase of two - the authour, plus one other person who finds it useful.

Anything else, more users, more support, financial gains - that's just icing on the cake. It does not define "success" in an open source project. This is what business people can not seem to grasp - the vast majority of people involved in open source software are not looking to recieve any kind of financial gain, or any kind of market penetration. They are just doing it for themselves, and for other people.

Re:Doomed I tell you, Doooomed

[Anonymous Coward]

Agreed. Completely irrelevant. All code is written and is "immature" at some point, even the code created in the enterprise. In addition, all code of any complexity is immature when viewed from some perspective. I think it is ridiculus and arbitrary to say that it is time to draw a line in time and say that any code not yet "mature" cannot pass. Why is that some feel it necessary to set limits on things they obviously do not understand?

I am the software architect for a major entertainment company whose products many of you probably consume. I use what works, my code, others code, opens source, whatever. I read the code, understand it, own it, integrate it. If something need fixed, I fix it. Done.

I pity the enterprises that have some clueless tyrant trying to tell their peeps what to use and what not to. *sigh*

Uh, huh

[typical]

Large companies buy software from "stable organizations" not because they're worried about the quality of the software, but because it's safe. Nobody ever got fired for buying Microsoft or IBM (or, increasingly, Linux or Eclipse). You're part of a crowd -- safety in numbers.

A purchaser at a corporation might get *fired* (cutting his salary to zero) because he bought something that turns out to not be what the company wants, but he isn't going to get that much of a reward (say, doubling his salary) if he manages to save the company the cost of the purchase by finding a free alternative.

As a result, it's in everyone's best interest to keep their head down, run with the herd, and make maximally ass-covering decisions.

If I'm trying to solve an engineering problem, I'm more than happy to use all kinds of high-quality packages that aren't backed by a large company. But that's because I'm trying to solve an engineering problem.

A purchaser isn't trying to solve an engineering problem. A purchaser is trying to solve the problem of how to maximize his job safety and income. And today's corporate reward structure heavily penalizes risk-taking.

If you want to produce solutions more in line with actually solving the original engineering problem, you go work at a startup or other small company where people don't have any problem with risk-taking.

If you go to work at a large company, you're going to be working with a large collection of highly risk-adverse people. That may be perfectly reasonable for them -- if one is middle-aged and has a wife, kids, and a house, stability matters a hell of a lot to you. If that doesn't fit with your mindset, though, you might want to try out those smaller companies.

SAP says OSS has issues?

[HangingChad]

I just love it when CEO's from companies producing large, expensive, proprietary data systems start throwing rocks at OSS because it's immature. When I hear that my internal babelfish translates it like so, "We're getting our ass kicked by some open source product so we have to frame the discussion in a way our customers can justify spending thousands more than is really necesary."

Well, it may be immature but it's not bloated or overly complex. And it doesn't cost a fortune to implement or require expensive hardware or expensive training for how to customize their proprietary business objects or require any of the monstrous administrative overhead systems like Siebel demand.

Maybe you should focus on making your product a value proposition instead of trying to run down open source. If you did more of that then maybe your crapass product wouldn't be getting the snot kicked out of it. Funny how big government and big business start thinking they have a right to exist instead of earning their living like everyone else.

This is a no-brainer

[br00tus]

As the article says, *immature* open source projects are done, the key word being immature. Certainly - any company large enough to have a real CIO position would have been hesitant to put immature technology into operation in the past, with not that much changing up to the present.

The article writer considers Linux as a mature technology. I worked at a Fortune 100 financial company not long ago where engineering was testing Red Hat Linux, but had none of it in production. Whether you call it data management or business processes, a critical machine could have literally billions of dollars worth of trades processed over it in a day, and engineering placed a very high value on stability. Most critical machines ran versions of Solaris which were more than one version out of date, even if they were new machines with new applications - IT management didn't want any surprises, they didn't want to be the ones to find a bug in the latest version of Solaris, or even the previous one. And if you have, say, hundreds of Sun Enterprise 4500s all over the world, you might tend to see bugs that shops with dozens, or a handful of E4500s might not. Wanting maturity is not a new thing.

I would agree that consolidation and focus on business process is the new fad among CIOs. So perhaps the days where in large companies an immature open source project would make its way in by 10%-20% of the environment are gone while this fad lasts. But there are plenty of smaller companies who do not have the budget, and are willing to use it. My friend works for a company with $1 billion in revenue, which is one company in a corporation which has over $3 billion in revenue - the revenue is just shy of putting the corporation in the Fortune 500. Despite all of that, the IT department uses a ton of open source, and only uses propietary technology when necessary. They've even been using immature open source software when mature, good propietary solutions exist for some things, simply due to budget. Despite a lot of things, at the end of the day, free as in beer looks very, very attractive to a lot of companies over even a slightly better competitor that costs tens of thousands of dollars. Even for companies almost in the Fortune 500.

Re:Other things that have to be considered.

[Theatetus]

ne is that when I'm installing F/OSS software, almost always, there's bunch of dependencies that I don't know about when I first install it. Yes, yes, there's the 'README' that has some of the dependencies. But it almost never mentions the dependencies of the dependencies.

Do you live in a strange world without package management? apt? yum? portage? ports? I know software on Windows has horrible issues with library versioning and sharing, but that's no excuse to drag that same problem into a real operating system.

A lot of F/OSS projects are started and never quite finished or they're not suported any longer - so, it'll be up to the users/consumers to maintain it.

A lot of proprietary projects are started and never quite finished or they're not supported any longer. In this case the users/customers don't even have the option of paying someone else to maintain it. It's just gone. My last company had about 3500 POP users on an old POP server called Post.Office. It hasn't been supported since 2000. When the company that made it went under and support disappeared, they couldn't even find somebody who could migrate the user base to another server. If the POP server had been open source, they could have sunk in a few grand to get a contractor to do the migration and recouped the cost pretty quickly.

The GNU license. If I ask, let's say, if I compile some proprietary code with gcc, does that mean my code is now under GNU? I'll get basically two different answers: yes or no

Ummm... no you don't. Are you trolling or just stupid? The GPL is quite clear that the output of a program is not a derivative work and so need not be distributed under the terms of the GPL.

At the very least, there's going to legal costs associated with settling that. If I end up going with this start up, I will have to get legal advice

Check out the MS Office EULA some time. The license forbids you from using Office to create materials critical of Microsoft (no, I'm not kidding). Did your lawyer check that out? Why not? If you can skip on that legal question, why would you hire a lawyer to find out what's explicitly said in the GPL, that the output of a program is not a derivative work?

Re:Maven 2

[xero314]

If you are looking to have to rewrite most of the available plugins and fight with configurations for weeks on end please consider Maven2.

Someone at my company recently convinced the rest of the team (excluding myself) that we should begin to use Maven2 for out build process. Maven2 is neither complete not bug free. Please do yourself a favore and stick with ant, where you have more fine control of your file structure and many more options for tasks already available (and I will also add that the exisiting tasks work).

Of course this is probably why OSS will find it hard to get into enterprises space. To often OSS is released before it is ready for production use and a bunch of fanboys (I have never felt the need to use that term before) push it like it's the silver bullet for whatever problem.

Nice Microsoft Troll

[WebCowboy]

Its all about accountability. Even if Microsoft may not have the best product, when it fails, the suits are able to hold Microsoft accountable.

Masterful troll, and as anyone who has had experience dealing with an uncooperatinve Microsoft-based solution knows it is a statement so blatantly full of crap it is hilarious. MS has a whole departmnent of legal people whose sole job it is to make sure Microsoft holds as little accountability as legally possible.

A little harder to do that with Debian, or any OSS without corporate backing.

Which is why Debian is not the favourite of corporate customers. IBM, Novel and Red Hat DO back their open source offeringe VERY well though. In fact, an open source solution from one of them is probably a much better bet in terms of accountability than Microsoft, becasue they are "solutions providers". Microsoft is a fairly immature player in that game--their business model is built around selling little boxes stuffed with shiny discs full of data and a bundle of useless paper certificates and "getting started" manuals. MS "innovation" with respect to their business model has been insignificant--it has amounted mostly to replacing physical boxes with certificates, product keys and activations. Since they have such a product-oriented mindset, the best you can hope for accountability-wise is a few hundred dollars in refunds for a scratched disc or botched install of Office as per some canned EULA. A "solutions provider", on the other hand, negotiates a far more comprehensive contract with explicit terms and conditions their business customers can rely on for accountability. The business never gets EVERYTHING they want, but they get a much better deal than Microsoft can normally offer.

Nobody ever got fired for buying IBM.

Absolutely true. That is why there is a lot of open source out there that has been deployed and managed by IBM. Often their "free software" solution costs more money than going with Microsoft, however IBM has the reputation of being more reliable, mature and accountable than Microsoft.

More SAP Bullshit

[Master of Transhuman]

'"The mess that companies have with their IT today is unimaginable, and the larger they get the more mess they have," Graf said. Some SAP customers have as many as 3,000 systems, for example. "They would be happy with just 1,000," he said.'

The above is the only part he got right.

The rest of it is mere justification for SAP's position in the ERP marketplace - and a response to the fear that ERP is being blamed for most of the mess he describes.

Sooner or later CIOs will realize that building apps from OSS tools is far cheaper and more effective than being saddled with a dinosaur like SAP for the next twenty years.

Take anything SAP says about OSS with about the same barrel of salt you can take from anything George Bush says about Iraq.