IBM and SUN, Corporate Open Source Shows Its Ugly Side Again

I’m starting to see more and more chatter about a report from MarketWatch about IBM’s apparent move to purchase Sun Microsystems; a lot of it centered around people’s $favorite_open_source_project and how the ramifications of the deal might affect things in the long term. Personally I see a lot of downsides for open source as a whole if this deal goes through, based on the contrast of Sun’s tight control of many of todays most popular open source offerings, and IBM’s more general hands off approach. Open Office: What would IBM do with OOo? It seems to me they would be happy to push back on the MS Office beast, but I can’t imagine them wanting to support this given what I’m sure they feel is a better solution with Lotus Notes. Add to this the rising tide of Google’s office related offerings, and Open Office looks like a really hard sell. The question is, could Open Office survive on it’s own? I honeslty don’t know enough about how the project works, but in the past I’ve not seen much room given to OOo to spread it’s wings inside of Sun, so I’m very skeptical they can surive on thier own. It would be a huge test case for the “you can fork open source projects to save them from evil companies” crowd. Java: IBM does have a history of Java support, so I expect this would continue on, but I know a lot of people are uneasy at the thought of IBM in control of Java. Sun has been moving Java towards a more open community the last few years, so this will be where we find out if they have moved it enough. Personally this seems like the best fit among open source technologies moving to IBM. (Open) Solaris: IMHO IBM has pretty much thrown in the towel on operating systems, putting a lot of effort into Linux the last few years. While, I’ll never forget watching that little man running on my screen and falling over (splat!), outside of SMIT, I think AIX is pretty forgettable. On the other hand, imagine dtrace/zfs running on IBM hardware, that could be pretty sweet. With Solaris, IBM would have the means to compete with Linux at the OS level, the question is do they want to? MySQL: As I see it, there are two ways this plays out inside of IBM, and neither of them have much upside for MySQL. I could see IBM repositioning its database tools with DB2 as the standard for big iron, Informix as the mid-level database for small business, and a new database as the low-end offering, but honestly I think they would choose Drizzle for that. Its simple feature set and focus on federation make it good for entry level applications and keep it from butting heads too much with Informix in the OLTP space. The other side looks a lot worse; I know a lot of people think Informix is dying a slow death inside IBM, with all the love going to DB2. MySQL could slide in somewhere behind that and wither just as much, if not more. That’s a scary thought, but on the other hand that might actually help the folks at OurDelta and XtraDB. Is this all FUD? Probably… for one we’re still quite a way from an actual IBM purchase, and there are a lot of different ways this could play out. On the other hand it is kind of justified; if I’ve bet my business on one of the above technologies, I need to have a back-up plan for the vendor becoming my enemy. Corporate Open Source, by its very nature, is prone to these issues, and it isn’t just a theoretical problem… do you think CentOS was born out of love for RedHat’s business practices? (And yes, this is why some people use Debian over Ubuntu) Of course this is also why I have always tried to stress the importance of community based open source projects like Postgres. Consider this; when Pervasive stopped supporting Postgres, we got 8.2, with warm standby, online index builds, GIN indexes, Dtrace probes, and improved performance. Over the last year, Sun bought MySQL for a Billion dollars, pushing PostgreSQL farther out of the picture within that company, but both 8.3, with it’s heap-only-tuples, asynchronus commits, XML support, and built in Full-text search, and 8.4, with Common Table Expressions, SQL2008 Windowing Functions and partial-match full text capabilities, were solid releases. The ability to transition between corporate supporters and still turn out solid releases is a huge plus for any open source project, and for folks who rely on those projects; and that’s the power of community based open source.