Early free-software advocates like Richard Stallman raged against the copyright-toting software capitalists, yearning for a brighter day of peace, love, and (GNU) Linux. In 1998, afraid that this quasi-hippie ideal might scare away the business world from embracing free software, Eric Raymond and a few others came up with the term "open source," broadening the tent well beyond free-software radicals.
Today that tent is broad enough to include everyone from Stallman (still fighting the same fight he always has) to Microsoft, with the poignancy of the term "open source" coming to lose some of its fire, even as the power and breadth of its code base increases dramatically.
Has open source come to include so much that it's somewhat meaningless? If so, should we return to the free-software roots that defined its infancy?
Personally, I think the bigger the tent, the better it is for open source. Even a tent that includes a wide array of open-source leeches, an issue taken up by Bill Snyder in InfoWorld.
I like the big tent (and even, increasingly, the "leeches") because the broadening is largely driven by the same motivation: finding ways to feed one's family by writing more open-source software.
That, after all, is really the reason for the term "open source" in the first place: make it more relevant so more people could create sustainable businesses around it. Indeed, as Jack and Suzy Welch recently opined in Businessweek, the hallmark of a great company is how much value it drives to its employees and shareholders:
A company's foremost responsibility is to do well. That may sound politically incorrect, but the reason is inexorable. Winning companies create jobs, pay taxes, and strengthen the economy. Winning companies, in other words, enable social responsibility, not the other way around. And so, right now--as always--companies should be putting profitability first. It's the necessity that makes every other necessity possible.
Red Hat is nice when it writes open-source software. Red Hat is great when it employs thousands of people to do that, people who can care for themselves, their families, and their communities.
Doing good by doing well. This is the principle that drives the open-source start-ups that I know, from MindTouch to Cloudera to Compiere, but it also drives the "proprietary" companies like IBM and Oracle that contribute a great deal of code to open source while also ensuring they pursue business strategies that enable them to be financially competitive.
It's fine for free-software purists to insist that the world abandon all proprietary designs, but until they demonstrate that more money can be made in this way and, hence, more total social good can be achieved, then their arguments rightly ring hollow.
I've been in that camp. I urged the world to "burn the boats." I hold by that counsel, as it was a direction to look forward to new licensing and business models, rather than fixate on the past, largely as Google has now done with its Wave product.
But if you torch your "boat" into the ground without any idea as to how to get your business to $100 million in sales, you're rightly going to get singed (or worse) with your business. Open-source developers have a responsibility to do well so that more open-source software can afford to be written. At present, this may require keeping some bits proprietary. Hopefully in the future it won't.
Until that blissful day, it is foolish to hold to a purist software development philosophy if it minimizes your ability to feed your family and to be widely relevant to the industry.
Follow me on Twitter @mjasay.
I've talked before about the close affinity between open source and SaaS, including that SaaS is built using open source, but it's only been this week that it the full import of open source-begets-SaaS came home to me.
Perhaps it was seeing Zimbra and OpenX both announce hosted offerings of their respective open-source email and ad server offerings.
Or perhaps it was my conversation with Bill Kaiser, a friend and Greylock venture capitalist, wherein it seemed that the most interesting new venture opportunities were in SaaS, and that open source has not resolved one of the fundamental problems with enterprise software: it's too complex, too cumbersome, and too Soviet in its design aesthetic.
This thought was underlined by Terry Barbounis, a friend and Christian Science Monitor CTO, who practically gushed about the positive experiences with SaaS offerings like Jive Software's Clearspace he has been having lately.
Open source is a massive upgrade over the proprietary lock-in of incumbent enterprise software solutions. It returns control to customers and makes it easy for them to try software for free and without obligation, as well as to tailor it to their individual needs.
But open source has not gone far enough - at least, as a movement - in addressing the need for software that is easy to pay for and use.
Intriguingly, somewhere in that need open source can be forgotten, as Dusty Davidson of BrightMix, an OpenX customer, unwittingly reveals in endorsing OpenX's hosted version:
... Read More
I had a fascinating conversation with a Silicon Valley VC this morning. He has successfully invested in a range of open-source and software-as-a-service companies, and made a somewhat startling comment:
It's becoming harder and harder to discern between open-source and SaaS companies. Their revenue looks increasingly similar, as do their business models. It's becoming hard to tell, for example, whether SugarCRM is a SaaS company or an open-source company. In reality, it's both, and it's hard to tell where one begins and the other ends.
He is absolutely right, at least about the blending of SaaS and open source. As I noted in a previous post on SaaS business models, both open source and SaaS have increasingly turned to "freemium" models. Both are all about getting software in the hands of prospective customers as fast as possible, and looking to monetize that adoption downstream rather than upfront. Both provide a lower cost of entry and (often) a long-term cost savings, with a subscription-based revenue model.
Intriguingly, he told me that SaaS vendors have discovered that there is very little cost to supporting a customer, and little risk of churn once the customer is actively using the product. So more SaaS companies are finding ways to give away free versions of their service (which costs them little), monitoring actual usage, and sending in the sales team once adoption starts to increase. Sounds like open source, right?
... Read MoreI loved this presentation by David Heinemeier Hansson of 37Signals. His topic? How to make money as an online software company.
His verdict? Charge for your product, but be careful whom you charge.
Chris Anderson elaborates on this theme:
37Signal's secret is not to target consumers (who don't like to pay) or big companies (that's a crowded space). Instead, they target the "Fortune 5 Million"--small companies with specific needs that are underserved...
It's interesting how closely some of Heinemeier Hansson's ideas map to the commercial open-source world, in which charging for one's value is, as Roberto Galoppini suggests, not always straightforward.
Why? Because proprietary vendors have long conditioned customers to expect to get charged for the wrong things or, at least, to expect to get charged too much for the right things. Larry Augustin suggests, in response to a post of mine, that "one of the things companies using an open-source model need to do is make sure they get paid for up-front costs up-front." Easier said than done.
As Larry summarizes, enterprises have been conditioned to expect their vendors to dump all the risk of a software decision on themselves. They try to "get back" at the vendors by writing punitive terms into license agreements (e.g., acceptance periods that make revenue recognition difficult), requiring the vendor to jump through demos and pilot hoops upfront at the vendor's cost, and more.
Hopefully, as enterprises come to invest more trust in the open-source vendors, some of these practices will fade and the process for selling software services will level out. In the meantime, however, Heinemeier Hansson has it right: you need to charge for your product, and usually whom you charge is much more important than what you charge.
(Credit:
Bungee Labs)
Open source and SaaS dramatically change the way enterprises consume software. But today I heard this taken to the nth degree, with a shift in a company's revenue model, as well. Bungee Labs helps enterprises (of any size, but with a particular focus on the tens of thousands-strong SMB market) build rich Internet applications (RIAs) in the cloud, for the cloud.
It's one thing to build RIAs using desktop development tools like Adobe's Flex. This is a good model and will persist for a long time. Bungee Labs, however, represents a future that I think we're rapidly approaching: a future in which developers write for the web with the web and deploy "to" the web without skipping a beat. It's very cool.
Even better is how Bungee prices this service.
... Read More
- prev
- 1
- next
