Industry


Ads by TechWords

See your link here


How many billions is open-source software worth?

Open-source software is big business. For example, most of what Oracle is getting for it $7.4-billion purchase of Sun is open-source software. Thanks to a Linux Foundation study, we know that creating the Fedora 9 Linux distribution would have cost $11.5-billionin conventional software costs. So, given all that, what do you think OSS (open-source software) as a whole is worth? How's about $387-billion?

That's the number that Black Duck Software came up with. Black Duck isn't an open-source ISV (independent software vendor). The Boston area company started as an IP (intellectual property) risk management and mitigation company, but has since grown into an open-source legal management firm.

Since Black Duck was founded in 2002, the company has been tracking all known open source on the Internet According to their research, there are over 200,000 open-source projects representing over 4.9 billion lines of code. To create that code from scratch, Black Duck estimates that "reproducing this OSS would cost $387 billion and would take 2.1 million people-years of development."

The company isn't pulling that number out of a hat. Black Duck's methodology looks iron-clad to me. The company used "Barry Boehm's widely accepted COnstructive COst MOdel (COCOMO), an algorithmic Software Cost Estimation Model that relates software development effort for a program, in person-years, to source lines of code (SLOC)." I've used COCOMO on consulting jobs, and it's a darn good tool.

So, that's neat, but what does it really mean? Black Duck said "We estimate that 10% of US-based development, representing $22 billion, is redundant and could be offset using OSS, much of which can be reinvested for true innovation." 10%! That little!? I don't believe it. Every software project I've ever analyzed has been stuffed full of what I call "reinventing the wheel" code.

You know what I mean, programmers writing lines of code that solve the same old problem, which have already been solved over and over again. I see this all the time, and not just in proprietary projects, I also see it in OSS as well. If there's one suggestion I could make to any ISV or in-house programming group it's to dump once and for all the idea that productivity is tied to any variation of lines of code per day. True productivity is making programs that do the job they're meant do on or under budget. That often means using already existing, already tested and working open-source code.

According to Black Duck, "With 4.9 billion lines of 'shovel ready code' available to developers OSS is a stimulus resource that can help development organizations around the world increase innovation, stretch budgets and spur growth." 

By tapping into OSS to companies can shift "scarce resources into projects that represent innovation and competitive differentiation."

In other words, using OSS isn't about being anti-Microsoft or believing in some sort of open-source ethical rightfulness, using OSS is simply a smart business move. Indeed, at $387-billion in value, OSS is twice as valuable as Microsoft's current net worth of $183.5-billion. Not bad for 'free software' is it?

What People Are Saying

Cost to Create OSS

To be clear, we at Black Duck didn’t estimate the ‘value’ of OSS which as a number of people have rightly commented can only be determined in a market sense, i.e., what are people willing to pay for it. See the analysis at http://www.blackducksoftware.com/development-cost-of-open-source . There’s no practical way to come up with a market value for 200,000 projects. What we did do was estimate that it would cost $387B to create the 4.9 billion lines of code in those projects. We’re happy to leave it up to developers to assess the value of these projects, for example, in accelerating their development projects.

Much of it worthless

Ptui! Remove from the analysis the high percentage of that OSS that's worthless, duplicative, not stable, poorly written, not updated, etc, and the number comes waaaay down. Yes, there's much that valuable either directly or for inclusion in other products, but let's be realistic.

There's lots of crappy

There's lots of crappy proprietary software too - but someone took out a huge bank loan to pay people to write the stuff, so it wouldn't make sense to only value a few big projects which most people use.

Take off the Open Source blinders

...and THINK!

You're seriously quoting Open Source value based on lines of code... when the entire *rationale* for Open Source is *code reuse*!

Are you really so ignorant? They're already reusing that code!

...and look, if the wheel occasionally gets reinvented, it's simply because the particular Open Source wonk involved has decided that her wheel is somehow BETTER than all previous wheels... "No really, it's ROUNDER I tell ya...!"

So... sorry, no, you don't get credit for that pointless duplication either.

Meanwhile, a closed-source company is enforcing code reuse and standardized development(...well, OK: except for the Lotus division of IBM, I'll grant you...) and not fooling themselves into ignoring the efficiencies thereof by artificially inflating its value.

When and if Microsoft does a COCOMO II on their development, you can but your a$$ they don't include the .NET framework, or MFC, COM, OLE, or you-name-it's development time in every calculation...

"her wheel" Most opensource

"her wheel"
Most opensource programmers are Men.
Some wheels are rounder than others.

This is a fairly complex question, so - - -

Biggest problem with the article is that SVJN mixes differing valuation models.

If one looks at the value of the labor put into OSS, then, yes there is a huge value of labor and ingenuity.

Then there's the value to Oracle. That's a whole different method of valuation. In business, if I buy something and hope to make money off of it, someone has to buy it from me and be willing to pay more for it than I did. Basic profit motive here.

In this case, Open Office, Java, and other Open Source products created and managed by SUN have a zero dollar value in and of themselves because of licensing. However, if Oracle can build service model businesses around these products, the products form the foundation of the service business.

While any company can base their service business off of these products, Oracle will have an advantage as they know the products better than anyone else.

There's a big dofferemce betweem the value of labor invested to create Open Source Software than the value of that software to a company like Oracle.

Only time will tell if Oracle can turn the Open Source software into a gold mine or whether the software will be a lead wieght around their nece. I wish Oracle luck, it's a tough market out there!

If they could transform this

If they could transform this open-source software into a big business - this deserves a lot of respect. I think so!

In the end a product is only

In the end a product is only worth what people will pay for it. If they pay nothing than you have its true value.

On an economic basis if the amount spent on SW is elastic then we can assume if you began to charge for this free SW the demand would drop quickly. Also if we would assume a fix maximum value for the purchase of SW then in the end the best you could hope for is the replacement of paid SW by previously free SW which would leave no economic gain.

So open source is really just free SW worth nothing. Much like Web2.0 biz that have customers for a service but who do not pay for the service.

Profit vs Revenue

This wrong-headed analysis is no different than claiming that a restaurant like McDonald's cannot possibly make a profit if they give away free parking. It ignores the fact that McDonald's isn't in the parking-lot business, it's in the fast-food business. Providing a parking lot for their customers is just part of their cost of doing business.

Similarly, open-source value is in reducing cost and business risk. Profit == Revenue minus Cost, and when you focus on the price tag of open source, you're ignoring Profit and instead obsessing over Revenue.

Actually, McDonald's is in the Real Estate business

Ray Croc was famous for explaining that McDonald's wa priamary focused on real estate. Selling hambergers paid for the real estate.

That is also why they do not make the "best" hamberger, they make the hamberger most people want at the price they want to pay.

FOSS is the hamberger here...