Free Sofware developers -- in their daily focus -- are commited to making the software GOOD. Free Software companies -- according to their own moral imparative -- are commited to making the software USEFUL.
That the two objectives don't wholly overlap is a source of frisson between Free Software individuals and the Free Software's corporate sponsors. Maybe this is healthy in itself, or an opportunity waiting to be leveraged.
Perhpas it's time for the Free Software community as a whole to think about the role and value that Free Software companies bring to the ecosystem.
In a BBC News thought-piece, Bill Gates recently made a revealing comment ...
"Software innovation, like almost every other kind of innovation, requires the ability to collaborate and share ideas with other people, and to sit down and talk with customers and get their feedback and understand their needs."
This is a veiled criticism of Free Software. He's talking about his software business -- where he often conflates software technology assembly with the organization of people in his business dedicated to solving customers' problems and making alternatives look unattractive.
The quote is difficult to fault as a business statement, since they have proven adept at doing the whole package. But it doesn't account for the superior code produced by both isolated and collaborative work across the Free Software ecosystem.
Mr Gates visualizes the Free Software developer -- the Hobbyist, for whom his distain is on record -- as an anti-social being (see in section, "Non-Assertion of Patents Pledge," the definition of " Non-Compensated Individual Hobbyist Developer"), having long hair, bad breath and working alone in a garret. He can't be entirely wrong because he was a software hobbyist, himself, back in the 1970s -- closely fitting that description.
Free Software doesn't actually suffer from a lack of collaboration on the code. It suffers in the market-place -- that bazaar of products -- from an almost comprehensive lack of collaboration with business.
A few years ago, I thought it was an embarrassment that the bellwether GNU/Linux company, Red Hat, had only passed the $100-million total revenue mark. Given the size of the enterprise and consumer software markets, that number was just ... well, embarrassing.
Just this week, we read with a mixture of amusement, glee and ennui that Novell, having got its reporting methods checked off by the SEC, reported 4th Quarter revenue from Open Platform Solutions (which includes Linux) at $23 million, up 69% versus the year-ago quarter.
Even though Novell's Linux invoicings (something different than accounting revenues) were up 108% in the period, suggesting an annualized Linux billing-rate well beyond $100 million, it is fairly depressing that the company with the very best enterprise GNU/Linux desktop, plus some good identity, messaging and deployment/management products makes as much money in one Quarter as Microsoft makes in a few minutes.
Evidently, GNU/Linux's enterprise penetration today is so minimal that it is hard sometimes to see why Microsoft bothers to oppose it. (Or is it attributable to the monopoly's effectiveness in opposition?) This means that GNU/Linux does not have a depth or breadth of conversations under weigh with corporate IT departments -- the innovative, feedback kind of conversations that Mr Gates rightly prizes.
How is this going to change if not slowly? Progress seems glacial today in light of the Free Software community's apparent and general disinterest in business and willingness to abuse sincere, if controversial, efforts to compete. The small amount of business in the commercial Free Software space makes it hard to attract bright people. The space seems incompatible with innovation. (I'm not convinced it's merely due to monopoly conditions.)
Experience working and solving enterprise customers' problems breeds experience. Solutions as well as sales, deployment & integration knowledge improve in a cumulative way.
I wrote last year in the Financial Times about the more prominent GNU/Linux migration case studies in Europe: the Gendarmes, Munich and PSA Peugeot Citroen. PSA is a Novell customer and the migration of 40% of that car company's desktops to SUSE Linux Enterprise Desktop 10 is a pivotal case for Novell and for the GNU/Linux desktop in general.
Christophe Therry, Novell general manager for France, explained: "Peugeot was impressed with the translucent 3D desktop, with the user-interface functionality and was looking for a way to facilitate the path through the user-adoption curve."
Peugeot is a Lotus Notes and SAP house; these enterprise applications must integrate seamlessly. Peugeot and Novell convinced IBM to port Lotus Notes to Linux (which IBM had no prior intention of doing), and this was accomplished for Novell SuSE Linux Enterprise Desktop 10 in just two months. (IBM and Novell are now cross-marketing Notes 8 on Novell's SLED 10 Linux desktop.) Novell was able to feed back Peugeot's requirements into SLED 10, for which Mr Therry credits Peugeot's culture.
Working together they improved SLED 10's wireless security making it possible, for example, for a laptop's connection to move from one domain server to another while maintaining security.
They improved Linux's ability to integrate securely in a Windows environment; and added coding improvements to the Firefox browser on Linux which make it possible for the internal websites to conform to the exotic requirements of Microsoft's Internet Explorer browser when such pages are viewed on any platform.
Free Software developers -- perhaps those operating from their garrets
-- may ask, "Why do we need to commercialize Linux? It's fine as it is.
It solves all my problems!" It's a fair question, for which
we need a fair answer. Even Linus Torvalds has commented about
virtualization, for example, 'I don't care ... I'm just not that
interested in it.' We are glad they are so focused on the problems that
are important to them; it has made the software GOOD.
Yet PSA Peugeot Citroen represents the kind of enterprise engagement that is not only healthy for GNU/Linux but essential to making it USEFUL. Through the engagement, key customer-recommended features get added or adjusted within the software, and the collaboration with the end user helps Novell address problems of integration or of interoperability that are outside the GNU/Linux code-base (and therefore not typically identified as problems). Thanks to the quality and the efficiency of developer collaboration around the GNU/Linux code, the inside work on GNU/Linux is so advanced today that it is largely these outside problems which present the more significant obstacles to adoption.
Mr Gates is right about software innovation, but he's the wrong person to say it. Enterprise engagement by commercial Free Software companies is critical if GNU/Linux is to be useful as well as good. And we need more of it.