Looking Back at Open Source in 2005

bhmit1 writes "BusinessWeek is reporting on the open source progress in 2005. Their conclusion: "in 2005, the software movement finally gained traction in Corporate America and saw a new influx of VC cash." Has the shift in corporate america really occurred or are activities like the profitability of Red Hat signalling that the CEO's are still holding on to the old way of business?"

Comments

huh

[dajobi]

Has the shift in corporate america really occurred or are activities like the profitability of Red Hat signalling that the CEO's are still holding on to the old way of business?

Old way of business = profit. New way of business = ???

Seriously that doesn't make sense. Surely Red Hat's profitability indicates that they have a handle on the new way of doing business.

Re:huh

[bhmit1]

You're thinking about the business model of Red Hat itself, which is a good one. I was referring to the business model of the large corporations who seem to be saying "open source is great, where do we buy it from?" Support is a great argument to pay someone else, but that choice frequently backfires. Support organizations make their money by hiring cheap labor and postpone resolving the problem until the customer frequently does it themselves, or blames someone else for the problem. I've seen it happen so often, that I don't understand why managers think support provides a valuable benefit for the cost.

So the better business model for larger organizations would have a stronger IT organization that has enough capacity to understand the applications they are implementing and provide support internally.

For smaller organizations, I think they are better served by getting a local resource that they can call for problems and that performs a checkup a few times a year just like that organization would do with legal and accounting services.

And for the record, I don't think we've crossed the threshold yet, but it's interesting to see what the business types are watching.

Be happy, not excited

[SilverspurG]

Industry estimates show some $400 million was invested in open-source startups in 2005

All $400 million? That's enough to buy Bill Gates' lunch, maybe, if he's eating light that day.

The market is still dubious about open source and for good reason. The big players, the ones pushing around 400 billion dollars, still control the legal avenues and we've seen that the legal avenues are being used in many ways to hedge out the OSS players. If $400 million in VC was invested in OSS startups then it really was venture capital in the truest sense of the term.

I'm happy to see OSS getting a foot in the door but I'm not going to break out the champagne and glasses until we see some real reform on both the business and political fronts.

Support is King

[JumpingBull]

IBM learned long ago that ongoing support generates a constant revenue stream.

That lesson was not lost on Novell, Redhat and I believe Ubuntu is following the same path

I think that we will see the application services and support companies running up the revenue stream. However, it takes talented people to seed this activity; one with a proven track record. I have been told that a VC looks briefly at the business plan, just to see it is thorough, knowing full well that as soon as the business opens it's doors, that plan will change as the prime movers identify the hot market needs.

So the quality of the people in the enterprise, and their successes is what gathers the most attention from the VC. It is the people that will make or break the business.

Companies want someone to yell at!

[ErichTheRed]

Has the shift in corporate america really occurred or are activities like the profitability of Red Hat signalling that the CEO's are still holding on to the old way of business?

The reason why open source vendors who act more like "real" companies do well is because corporate IT absolutely demands that they have someone to complain to when everything goes to hell. Imagine you're the CIO of a 25,000 person company who depends on its IT systems to make money. I think you'd be foolish to trust that the crew of experts you hired is going to stick around, and be able to solve any problem that comes up. Sooner or later, something high-profile will die. Who do you call??

Companies like Red Hat enforce standards in an open source world that really doesn't have very many. They sell RHEL with the promise that you'll get tech support as long as you use their packages and software. That's a compelling argument. One thing I've been impressed with is commercial vendors' ability to call in massive amounts of help when a real emergency occurs. Red Hat, Novell, etc. are capable of that. Even if you have a support agreement with the makers of fooPackage, which happens to be the crucial link in your business process, can they guarantee that they'll work with you as long as it takes to solve a problem. Worse yet, let's say it's a multi-level problem between fooPackage, barPackage and blahPackage. Now you've got "dualing vendors" on your hands all saying "it's not our problem." Not that that doesn't happen in the commercial world, but a commercial OS vendor (Sun, Microsoft, IBM, etc.) is helpful in mediating those fights.

The Red Hats and the Novells are going to be the ones who finally get a Linux desktop on the market. That's because they'll pick one office suite, tweak the hell out of it, and make it a standard akin to MS Office. Companies want to know that their training dollars aren't going to be wasted. Most users learn one software package to do their jobs, memorize the commands, and will not readily learn anything new. That's what the Linux desktop is up against.

Re:Companies want someone to yell at!

[richlv]

we just recently had a "situation" with a big, well known commercial vendor and their "enterprise" system.

the application supports only ancient operating system. and an unpatched version of it. nobody knows wether there will be a new version that would support something else.

then, without a warning, a new beta version is released. nobody knows wether this version will support a newer version of chosen operating system. support is silent.

public download is available. file, sized 1.5gb and containing "multiplatform" in the name. when finally downloaded, turns out, it contains only a version for a single platform. support does not respond to questions about versions/platforms supported (it is in beta already, remember).

and this is for a bloody PAYING CUSTOMER.

oh, installation of the software takes some 7 manual steps, each including a lot of obscure prompts and chances to screw something up. from 8 installations at the education lab _none_ is able to finish on the supported platform, there are no error messages. almost each install stalls at a unique point. the best was a finished installation that was unaccessible for unknown reason. of course, software is closed source, so good luck figuring out what is wrong.

screw big vendors. we have had similar experience with most of them - and problems are either solved inhouse, or we find ways to avoid them.

I think you'd be foolish to trust that the crew of experts you hired is going to stick around, and be able to solve any problem that comes up.

well, from my experience that is the only thing you can trust (of course, by designing systems both from technical and personnel viewpoint very carefuly). unless you can make or break their business, big vendors don't give a shit about your problems even if you are a paying customer.

now, it is somewhat different with all these linux vendors, i assume - you get a support (and, at least at this state, they are interested in solving problems fast and nice) and if the support is unable/unwilling to help you in required time, you can tap into internal resources or look for help elsewhere. i believe that should be a requirement for any serious information system.

Re:Really occurred?

[Anonymous Coward]

Office suites are a bad idea. MS Office is proof of that; it's the best product in its category, but it still really really sucks.

I hate comparisons like this. They are so useless. It's like saying the Babylonians were the best mathematicians in their day, but they were really, really stupid. Or the Germans were the best physicists in their day, but they were really, really bad at physics.

You can gain no insight whatsoever with such a statement.

Re:It's about time

[Bert64]

Training costs are a one-off, and exist because people are trained by repetition, and can't deal with something slightly different..
Training costs will exist wether you convert to opensource or not, when you upgrade your proprietary software there will also be retraining costs associated.

Support costs may increase temporarily whenever anything new is introduced, and will settle down once people get used to it. This goes for proprietary software and even hardware too.

External support costs are likely to decrease, since companies will have to compete with each other to provide OSS support, whereas a proprietary vendor pretty much has a monopoly on support of their own products..

Conversion costs will also occur in any case under oss or proprietary software, when the upgrade cycle occurs every few years..

Future conversion/upgrade costs are likely to be much lower, since opensource software typically adheres to standards and is easy to replace with other standards compliant software.

Open source doesn't force you to upgrade, if you have an external support-provider who is willing to continue supporting a 10 year old version of linux you've no reason to replace it unless you WANT to. On the other hand, support vendors for proprietary software can't provide you any fixes without the original vendor's help.

To give some insight, i provided a few NAT/Email boxes to a few local businesses in 1997.. These boxes run redhat 4.x and typically sat on a pstn dialup when first installed.. These companies pay every month, and i patch the systems against any security flaws if necessary, and update them to handle new types of network connection (dsl etc).. I also add/remove users etc, if requested..
These boxes just provide a nat gateway, and email services so users behind the gateway can read their mail.. Aside from a couple of hardware failures (no real issue since everything is backed up) nothing has gone wrong with these machines, and they're still patched up and secure.