Tuesday, 29 June 2010

Software: no compensation necessary

I've just been reading "The World Is Flat" by Thomas L Friedman. It's an interesting book with lots good to say[1] about how the world is rapidly flattening - ie everywhere is becoming reachable from anywhere.

Along the way (and only incidentally to the main point) the author brings up the concept of Open Source Software.

He discusses the concept in a fairly positive light. He seems quite fascinated by the phenomenon - though doesn't seem to really get why people would do it. Though he makes a valiant effort to see things from the POV of the average OSS contributor, it's still fairly clear that he thinks it's a bit of a quaint, idealistic example of altruism (recognising the "pay it forward" aspect), rather than an effective and useful business technique. This is exemplified by the fact that he interviews a Microsoft Manager to ask whether or not it is sustainable as a business concept. Not bothering with an Open Source advocate, just an employee of one of the biggest proprietary systems out there.

One point stands out from what this employee says about OSS. He states that money is the main motivator for building software. From this the author concludes that any "really good software" will more likely be built by big software houses (eg Microsoft) than by OSS initiatives... after all, they have a lot of money, that can buy a lot of development time. He therefore concludes that The big software houses will always build the best software - simply because they can sustainably pay the most developers.

It's an interesting theory, and the most logical for those who don't really understand the motivation of the average OSS contributor. The reasoning behind this conclusion is of itself quite interesting. They claim that software is developed because businesses need stuff to be done - and thus they pay developers to do it. Nothing wrong with that... but they seem to mistake the effect for the cause and assume that the motivator for software to be built - is the money.

This is not the case.

Software is susceptible to the old phrase of "Necessity is the mother of invention" - ie the cause of software to be built is the business's need for it. The money is a secondary affector - ie the need causes the business to find money to make it happen. The originating need will be present regardless of whether the business has money to pay for developers... after all, even if the business has no money, the need will still be there for the software.

If the business happens to be made up of developers, then they are quite capable of creating the software, regardless of payment... and that is how you get startups: built and run by developers who see the need but don't necessarily have money to pay other developers - they just need the software to get their own business running, and are ready and willing to work in their own spare time to make it happen (nothing wrong with Sweat Equity).

The Open Source world calls this motivation "scratching your own itch" and it is recognised as the primary motivator for Open Source contribution[2]. Even for the really big projects.

Linux, for example, was Linus Torvalds' project to build himself an OS that didn't suck. Since he released it, thousands of other individuals have contributed to it for the same reason. It didn't do what they needed, so rather than bitching about it, they put their money where their mouth was and wrote the fix themselves... contributing it back to the OSS community... for no compensation except the "pay it forward" principle.

Which brings me to the other book I've just finished reading: "Drive" by Daniel Pink. One of the principle findings of this book is that fiscal compensation actually gets in the way of intrinsic motivation[3]. You can incentivise a developer to build something, but then they dissociate their ownership of the end product to the person doing the incentivising... care factor drops radically and you end up with... well... Microsoft products. Corporate and cold, buggy and with no real way to guarantee that the problems will ever get fixed.

This difference in motivation easily accounts for the fact that MS products are made primarily for creating more money for Microsoft (for which purpose they perform admirably), but only a passing interest in actually being fit for use: ie only so much as they can get away with before too many people stop purchasing them.

Now, there's nothing inherently wrong with that. More power to them for staying alive so long, and rather profitably - by keeping so many people satisfied for so long with what they're prepared to shell out... but there's a reason that I haven't used Microsoft products for over a decade now...

By comparison, the OSS world is driven purely by user requirements. If software is not fit for purpose, it simply won't be used... but more importantly, because there's nobody paying for it - if it doesn't do what it's supposed to, it will never be born

Conversely, if it is built, you can guarantee that it meets the requirements of at least *one* real customer (unlike many products built purely with potential revenue as a motivator).

OSS lives and dies on it's own merits... no compensation necessary.


[1] Though I'd take a lot of it with a pinch of salt as well. This guy is clearly of the "yay capitalism, it solves all the world's ills" school of thought... though is slightly tempered by realising that humans are human, and some of them actually need supporting by those who are better-off... it's an interesting act of comparison to read this side-by-side with "The Shock Doctrine" by Naomi Klein - just as equally biased in the opposite direction, but also an fascinating read. I think if I ever put these two books together on a shelf, they'd annihilate one another with a very loud bang.

[2] If you're interested in this topic, "The Cathedral & the Bazaar" by Eric S. Raymond is an excellent book that explains the difference between what can be built from the ground up by need-motivated developers rather than the top-down software built by big software houses spending money to pay for what they want.

[3] which can be paraphrased as that innate curiosity and enjoyment of a challenge that occurs when you take on a new puzzle


Daniel Azuma said...

A very interesting and I think very worthwhile analysis of OSS is The Success Of Open Source by Steven Weber. He effectively shows that Open Source has very little to do with altruism or "pay it forward". Indeed, it is a fully rational expression of a competition-based economy---that is, compensation does matter, but not in the way you would traditionally expect. He does this by arguing that, in the information age, what is valuable has changed. Software itself no longer has as much intrinsic value as we traditionally thought, because it is so ubiquitous and easy to replicate. Therefore, the traditional model of paying programmers---exchanging software for money---does not work as well as the Microsofts of the world think it does. Instead, what is now more valuable is attention, expertise, reputation, and community---that is, human capital. These are the scarce resources of an information economy, the things that are really worth money. And this is what OSS takes seriously. It basically functions as a profitable exchange: trade source code for the more valuable human capital it can attract. This is why, for example, Linux is so powerful: because of its mindshare, community, and the wealth of technical expertise at its disposal. Weber covers the economics, and a number of other related political and social questions such as how OSS projects manage to defy Brooks' Law and coordinate the efforts of hundreds of developers. It's a good read.

Taryn said...

Hi Daniel,
wow great review. That definitely goes on the reading list.

And certainly sounds like it has some interesting things to say on the topic. Reputation has become one of the most valuable currencies of the modern IT industry, and community is also essential. It is one reason why having a social media marketing strategy is now becoming crucial. Building your tribe (as Seth Godin would put it) is almost mandatory t do well in the new economy.

For one thing - the unified community easily swayed our original decision to choose Ruby-on-Rails over the then-competitors. You can know for certain that if a community exists, that there will be some dynamism involved in the project - an active customer base, but (even more important) an active contributor-base.

Microsoft is not entirely unaware of this effect, however, which is why they put so much effort into MSDN. I also don't think that the older ways will simply die out - they also have their place, but are simply part of the new toolkit necessary to thrive.

It'll be interesting to see how the market grows and changes from here.