'Free' may be losing its allure

There's a range of great chatter on the Web recently that has direct implications for open-source business models, all of it related to the value a vendor derives from "free" as in cost.

First, Plenty of Fish writes that it's on the lookout for a model to replace its famous $10 million-model built on Google Adwords:

The problem with free is that every time you double the size of your database the cost of maintaining the site grows sixfold. I really underestimated how much resources it would take, I have one database table now that exceeds 3 billion records. The bigger you get as a free site, the less money you make per visit and the more it costs to service a visit...There is really no money in being free and we have to start experimenting with other models now or we won't be able to compete in 3 or 4 years.

That's the Web world, where the cost of distribution (and support? Who thinks about support in this sort of business??) is free. What about open source?

Well, The 451 Group's Jay Lyman writes an excellent analysis of a rising threat to open-source software: free-but-proprietary software.

An emerging and ongoing trend that should be a concern to open source software vendors is the use of free versions, SaaS, easy deployment and credit card payment for easier access and low-friction sales by non-open source competitors. Proprietary vendors are finding that they can match, or at least come close to matching, the availability, cost and convenience that have typically been associated with open source software. If the open source players are not careful, they may be on the losing end of the same kind of disruption that characterized their own success.

Oracle and others have tried this before, to little avail. Source really does matter, even to companies that aren't interested in modifying or viewing it.

But that's a short-term advantage. Microsoft and others are figuring out ways to make code visible, if not modifiable, to customers. Open source can't stand still. We need to innovate, particularly in making our software easier to deploy and use. Stasis equals extinction.

[williamlouth]

"Oracle and others have tried this before, to little avail. Source really does matter, even to companies that aren't interested in modifying or viewing it."

Matt could you be discounting the fanboy factor here and misinterpreting this as "source .... does matter" when it is free versus free with little technical analysis as is the case for many things these days.

[Matt Asay]

No, source does matter. I'm constantly surprised by how high the percentage of our customers is that ask for source, modify source, etc. If not them, it's the SIs that want access to source as they custom-fit applications to the customer's needs. Source matters.

[tristanbob]

The source DOES matter. It provides some benefits to consumers (no vendor lock-in, ability to self-maintain code if producer stops) but the most benefit is realized by the people who create the software.

As I have written before, open source is the most efficient software development model because the entire world is sharing a software library. With open source there is no need to spend resources creating generic functionality (web-server, application server, thousands of helpful libraries and minor applications). This enables a vendor to put all their resources on creating new unique value which corresponds to a very low cost to develop the software. This is something that proprietary vendors cannot benefit from with a closed "free" model.

http://useopensource.blogspot.com/2008/11/evolution-of-open-source-software.html

[john.mark]

Yes, source does matter, if only because of the potential it represents. I may not have the expertise to work on my car, but the fact that it is open to anyone with knowledge means I can shop around and find the best mechanic for the right price. Think how high that price would be if only one mechanic had the access to my car's mechanics - the price of maintenance would go through the roof.

Similarly, source matters - and not just access to the source, but the right to modify, redistribute and fork.

[williamlouth]

The reality is:

1) Customers (especially developers largely targeted by Alfresco) rarely look at the source code and might not even have it when the sources are provided as a separate download.
2) Customers rarely change the software if they did then commercial OS would not be a viable business model
3) Open source offers no protection in light of the above from vendor lock-in especially when the software is proprietary in design - not based on industry standards.
4) There is not sufficient evidence to back up claims of the "most efficient software development model" especially when most commercial open source is maintained solely by the developers working in the commercial entity. Great designers create great software. That passion can be found in the commercial world as well as the open source world. The reason it appears to be much more prevalent in the open source world is because they are more vocal (spreading the gospel), its currently hip in the news, and because large commercial software companies generally stifle innovation and creativity in those few passionate designers and developers (who tend to like working in smaller teams at smaller companies). Smaller commercial software companies can easily out-perform (commercial or not) open source companies especially when the open source company is swamped with noise from the community. Its the key people with the vision and ability to interpret requests and forecast trends that make good software not necessarily the community or customer base.

[odubtaig]

Rubbish.

1) It does not matter how infrequently the need arises, that it arises at all is what matters. By that logic we shouldn't bother fire suppression systems in server rooms or offsite backup because customers rarely have fires.
2) You need to look, seriously, into what the actual business model is. As a follow on from another poster, I pretty much don't touch my car's engine but if my mechanic starts messing me about I can always go to another one. If Alfresco ever stop delivering value for money their customers can quite easily go elsewhere while continuing to use exactly the same system. Let's see you do that with SharePoint.
3) Whether software is based on standards or not, that you're comparing deconstructing and charting easily readable code to disassembling and reverse engineering a compiled program (which is easily 5,000 times harder) is a ridiculous start to a ridiculous argument.
4) Unless you define the cost of producing all the F/OSS available today against the cost of the cheapest equivalently functional closed software available as 'insufficient evidence' your last claim is either supremely ignorant or deliberately deceitful. It's not even so long ago that Matt here posted a link to a study costing how much RHEL would have cost to develop as a closed-source system (the word 'billion' was featured). That wasn't even taking into account how much it would have cost to separately produce Novell SLED, Mandriva, Ubuntu, Debian, all of which share a majority of their code with RHEL.

I do hope you never get on a jury.

[williamlouth]

Rubbish.

Your responses seems to narrow on one point of putting out a fire in an emergency by reversing engineering code - whether it is source form or not it is near impossible for most organizations to react quickly enough hence subscriptions and support contracts.The only people doing this in the wild are the ones contributing back and Matt has already shown that to be insubstantial at least outside of the few remarkable open source initiatives.

Just in case you forgot my main argument in your blind rage - "open source does ** not necessarily ** imply great software". There is 1000 or more crappy open source projects for every one gem. Nothing new. Nothing proven.

[odubtaig]

You've obviously not properly read my response Mr Blind Rage. I'll simplify. Read, understand, then reply.

1) Just because it's not needed often is not a good enough reason not to have it. Fortunately you're probably not in charge of car or airoplane safety.
2) When code is open, organisations are not locked to one vendor. They can pay someone else to maintain exactly the same software which would be impossible with closed source. This requires your vendor to deliver value for money. It's called commoditisation. To repeat: I'm not talking about them modifying the code themselves, I'm talking about the ability to change supplier for exactly the same product.
3) Anyone with a reasonable amount of experience can understand human readable code and, with effort, offer a support alternative if your vendor is not delivering value. Try doing that with a dissassembler.
4) There is also a ton of extremely bad closed software. What's your point? Mine was that there have been professional cost analyses but if you're going to start bringing every piece of bad software into the mix I'll start with WinME.