I spent the last week in Portland at the O’Reilly Open Source Convention also known as OSCON. We were part of the IBM exhibit talking to people about DB2 Express-C. OSCON is an interesting conference. Basically it is a big lovefest for the open source guys. Imagine about 2500 geeks all assembled in a single place for a week to talk about free software. The best part for me is that OSCON is not fixated on Linux. There were plenty of tutorials and talks on PHP, Python, Django, Ruby on Rails and of course databases. As one would expect there was quite a bit of MySQL presence and PostgreSQL, Ingres and EnterpriseDB all either had sessions or booths on the exhibit floor or both. And then there was the IBM booth (that you could not really tell was IBM) with DB2 Express-C.
I have to admit that talking up a non-open source product at the open source conference was a bit intimidating. In the past, I have on occasion run afoul of the people I call “open source fanatics” and I was a bit afraid that we would be going in to a hive of these and stirring the nest so to speak. In the end, I lived through it and I did not encounter any of the fanatics at the conference or at least none of them threatened me with physical harm. I hope this is a sign that fanatics in the open source are nearing extinction and are evolving in to more rational pragmatic bunch or at least are content not to force their views on the rest of us.
So, what do I mean by the open source fanatics? Have you ever met a person for whom commercial software is evil or the one that would argue that open source is the better way to produce quality software? I am not talking about these people. These are just opinionated folks. I am also not talking about those who thrive on making fun of the large software companies. Hey, making fun of Microsoft and Windows has become a sport. And there was some of that at the conference but it was all in good fun. I should mention that Microsoft was a sponsor and they had good presence with IronRuby. The people I am talking about are the ones that take offense to the fact that someone would dare to even compare a commercial product to an open source project. On several occasions I had people jump on me (figuratively speaking) when I compared DB2 Express-C business model to the very popular open source business model of a free license and a low cost support subscription. When someone considers a mere attempt at comparison to be an assault by the big bad corporation on the defiant and honorable open source project to me that is a clear manifestation of fanaticism. So, at the conference we talked to a lot of people and I made sure to say right upfront that DB2 Express-C is not open source and despite what may have been said in the press we have no plans to open source it. Conditioned by the verbal assaults on the forums, every time I said this I instinctively was raising my arms to guard my face and body. I continued to make the comparisons with the open source databases and I stand by it. DB2 Express-C is free and optional low cost subscription is available. And, unlike other “Express” databases from Microsoft and Oracle DB2 Express-C does not have crippling limitations on the size of the database (4GB user data limit for both Oracle XE and Microsoft SQL Server Express). So from this point of view DB2 Express-C is not much different from MySQL or Ingres. What I thought was great at this conference is that not a single person thought that not having access to DB2 Express-C source code was an issue for them. I guess the number of people who want to tinker with the database engine is not all that big. I think this is in line with what MySQL is seeing. Monty Widenius, co-founder of MySQL, talked about very low code contribution from the MySQL community. The plan is only 8 patches from the community in 2008. On the other hand, I found the feedback to the model of free license and optional support subscription to be universally positive. In general, most people who came by to talk to us were genuinely surprised that IBM had an offering like that and even more that this offering was so function rich and was not crippleware they may have expected. When we talked about the hybrid relational-XML storage engine and data access most people were not aware that there was such technology available. The fact that it was part of the free DB2 Express-C was a huge surprise.
Another thing led me to believe that demise of open source fanatics was close at hand were many comments who saw great value in a large corporation like IBM offering free products. Many commented that it brought legitimacy to the business model that many in the open source community rely on. People also really liked that they could take DB2 Express-C and bundle it in to their solution open source or commercial and have the reputation and clout of IBM be on their side.
Despite the fact that the radicalization of the open source community in my opinion seems to be on the decline, I was still not entirely in agreement with some of the messages from the icons of the open source movement. Every keynote had some sort of a statement on the open source being the engine of innovation. It was almost as if the speaker after speaker were trying to say that without open source there would be no innovation. I know I am going to get some hate mail on this but I will say it none the less. I actually believe that overall open source has not been very innovative. Many of the open source projects are simply implementations of very successful commercial products. Don’t take me wrong, I think it someone taking the time and energy to provide a free alternative to people who could not otherwise be able to purchase a product is great. I think that putting your code out there for everyone to see and critique is gutsy. I just don’t think that this is the definition of innovation. There was one moment that really brought home this contradiction around innovation. Mark Shuttleworth (Ubuntu guy) praised open source community as great innovators and then called on them to build a desktop interface to Linux that would rival that of Mac OS X. I think it would be great to have Mac OS X interface brought to Linux but, hey, give the credit where credit is due. Apple is the real innovator here.
Overall, I was really liked OSCON this year. I thought it was full of great sessions with really funky titles and a great open exchange of ideas. I was surprised and very glad not to be in a middle of some ideological debates on the value of open source or have the dogma of FOSS (Free Open Source Software) shoved down my throat. If anyone has a chance to attend OSCON 2009, I would very much recommend it.
Sorry I didn’t bump into you at OSCon, the guys I did bump into at the IBM wasn’t terribly approachable (in fairness, I was sporting a postgres shirt, so that may have been intimidating). Anyway, couple thoughts:
First, the database market gives you a very weird perception of open source, where open source is not simply access to the code, but also involves open community development. That’s actually a big part of what people think of when they talk about open source, but its gets summarized as access to the code, which is misleading. Yes, MySQL (and Ingres, and many others) gets very little community development, but much of this is because MySQL is not a community developed project, but a corporately developed product. If you want to see community development in progress, you have to look at Postgres, which does have a number of commercial contributors, but everyone approaches it as a community developer (ie. working for Sun gets you no special rights compared to a guy who just rolls in off the street, it’s all about the code)
Second, while I would agree that much of Open Source development is re-inventing existing wheels, you’re own example of Apple shows you are overlooking the contributions that open source made to Apple. Consider the difference between MacOS 9 and OSX… without the open source underpinnings, I don’t think Apple would have ever been able to push the desktop in the direction it has. Another example might be Google, which has clearly done innovative things in the search and systems management areas, but they have also said it would not have been possible without the open source toolset. I think that’s typically what we see from most open source development now days… it doesn’t necessarily innovate itself, but it enables innovation that just couldn’t happen otherwise.
Maybe that just means it depends on what you mean by the “real innovators” are… if it is the specific people who push something that next step, the whole open/closed thing is pretty irrelevant to innovation in general… what really innovators are smart programmers who can dedicate themselves to a task without having to worry about paying their mortgage, once you have that the rest is just details.
Robert, too bad you did not get to talk to us at OSCON. We are not at all intimidated by Ingres shirts. We have had a number of Ingres employees stop by the booth and we do have very good relationships with all of the open source databases. Not sure if you heard but IBM even funded EnterpriseDB to the tune of $10M.
I have to admit that open source databases are quite a bit different than most other open source projects. If one was to look at the most well known open source databases like MySQL, PosgreSQL, Ingres etc. only PostgreSQL, EnterpriseDB etc. really accepts any significant volume of contributions. The others simply publish the source and it is way too difficult to contribute even patches never mind any sizable chunks of function. This was one of persistent complaints at the MySQL conference and believe it or not it came from Yahoo and Google employees.
On the innovation. I completely agree that is tons of innovation in the open source community. I just disagree with the general tone that the icons of the open source are putting out there. I think there is a heck of a lot more innovation done by the startups and commercial developers. Granted, commercial developers may use a bit of the fruits of labour of the open source guys (they’d use a heck of a lot more if it was not for the lawyers and viral licensing issues). And unlike the pre-bubble era, startups use a lot of open source now but they still add a ton of innovative technologies and business models on top. Yes, Apple based Mac OS X on BSD Unix but there is really no comparison between the seed and the fruit (sorry, I know it is cheesy but I just couldn’t resits).
Hi Leon,
I’m glad you made it out to OSCON this year. I was in town, but I was busy with a big roll-out for a client and could not make it out to the conference. I was hoping you’d call me to meet up for a beverage, but you were probably too busy spreading the word about DB2 Express-C.
See you in Warsaw,
Fred
Well, at least open source has had very innovative in terms of customer relationship.
Without the Open Source products success, IBM (and others) would NEVER have given such a product for free…
Patrick, there is no question that Open Source movement has had a huge impact on the way commercial software developers like IBM operate. You are absolutely correct, if there was no MySQL there would not have been DB2 Express-C.
Great article Leon, and I’m glad you touched on the Open Source fanatic culture. You’ve pretty much summed up my own feelings and analysis of Open Source software, so I won’t touch on that since you described it really well.
For me personally, I recommend to my boss the best tool for the job. That may be Open Source, or that may be a commercial, enterprise level product.
In house, 9-5, portal content management system? You want to use MySQL? Yeah, we can take a look at that. Your 9-5 application code and support is optimized for SQL Server? Okay, we can look at that too.
You want to install Open Source on my 24/7 ERP system? I don’t think so, no way, no how, no chance. Wait, okay, I’ll agree to do it, but you have to carry my Blackberry around now and answer any calls that come in. 🙂
Pick the best tool for the job, not by ideology.
Patrick,
A big part of the success of Linux is because of IBM’s tremendous investment into it. At one point we joked that it was IBMinux, because of the large amount of contributions IBM made, not only with functionality and coding, but with a huge marketing push.
If I could have picked a “dream” open source platform, it would have been Postgres on FreeBSD. This is an enterprise level combination, though getting vendor support for it would be a nightmare. Both systems offer a very mature codebase, and rock solid stability. Scalability has slowed down a bit, compared with commercial offerings, but these are still two very well engineered, blue-collar Open Source tools that you can throw everyone and the kitchen sink at and they will still stay up and running.