Matt Loney's Blog
What it says on the box
Friday 26 January 2007, 10:58 AM
Open sourcing your proprietary app: an act of desperation?
Having dinner with James Reinders, head of Intel's software division this week, conversation turned to the wisdom of releasing proprietary software as open source. Intel is currently grappling with the issue for some its software development tools, which you can currently use for free for non-commerical purposes (incidentally, Intel's compilers are on the on the list).
The question Intel is trying to answer is, is it wise to release them as open source? To try to answer this, Intel compiled a list of proprietary applications that have been released as open source; in almost every case, they surmised, it was an act of desperation, regardless of how the company concerned attempted to rationalise the move. This does not imply that open source applications are inferior to proprietary apps, or that making an application open source is necesarily a bad move; it is intended more an observation.
Of course there are companies that make a very nice living from open source, and those who use it in parallel to more conventional business models (MySQL AB for instance). And there are those who, James reckons, had and have a robust strategy (IBM). But the examples of those companies who have released proprietary software under an open source licence not out of desperation, he reckons, are few and far between.
PreviousComments on this post
Courage, vision, desperation, or just failing to take your anti-psychotic medication one morning; these are all reasons to make Big Decisions. But that's not the point. The point is "What happens next?", and this is something which your article utterly fails to address.
Look at Netscape; it was going under but it released its source code under an Open Source license. It didn't save Netscape, but today we have the Firefox browser and the I.E.-specific websites are in retreat. Netscape, we salute you!
I think it depends on what kind of software you are releasing and how..
Both Java, OpenSolaris, OpenOffice, are now open source successes for the companies involved and weren't an act of desperation - they were fine before they were released, and the companies are still doing just fine - in fact Sun is doing so well they're looking to acquire some businesses.
Also see Ingres and Firebird databases, which have been open sourced and successful.
Then there are software components and libraries, which can be better suited than applications to open sourcing, IBM, Sun and other's have open sourced a great deal of their UNIX libraries for everything from Journalling Filesystems to profiling tools and libraries.
I think that, with all due respect, Mr Reinders hasn't investigated this thoroughly - it's a simple matter of picking where, when and how it would be appropriate to release technology as open source - whether to share the underlying libraries, a basic reference implementation or the full application, whether there is an existing community that would jump on board, or if the product or library would be likely to create or encourage a community, and once it was released how much involvement from the company would balance against autonomy for the project - Apple has stifled darwin development, some projects are just too big and complex and messy for a comunity that is unfamilar with the codebase to dig into (hence Netscape/Mozilla rewrite), other projects have a user-base that isn't interested in customising or doesn't have the technical knowhow to contribute.
Maybe Matt should have asked people who actually have open sourced applications or worked on open source stuff, as this article was a lot less worth reading than I'd hoped from the link I followed.
