Today, FSF released the third GPLv3 draft on http://gplv3.fsf.org/gpl3-dd3-guide . It’s great that the entire, complex process now has completed its next milestone, and I can fully understand why the FSF has seen it necessary to proceed at a pace somewhat slower than originally intended.

My thoughts go back well over a year to the Free Software Foundation’s first conference on GPLv3, at the end of January 2006. It was a great meeting, and as a Committee B member and representative of MySQL, I was asked by China Martens of IDG for my input. My reply “I think there was a good balance between different hair and beard lengths” ended up on China’s compilation “2006: The IT year in quotes” as “a colourful description of the cross-section of attendees”.

I wish I had something equivalently colourful to describe the third draft of GPLv3. “A good balance between FSF’s interest to protect software freedom in the name of the end user, and the interests of the industry to develop sustainable business models upon Free Software“? I hope I will feel able to use that for the final GPLv3 one day.

For now, MySQL continues to be actively engaged in the Free Software Foundation’s GPLv3 drafting process, the next step of which is the upcoming 60 days that should end up in a final draft. Our flagship product, the MySQL database server, remains licensed under GPLv2.

Until the new version of GPL is finalised, we won’t be in a position to determine whether GPLv3 is an appropriate license for MySQL products. As I said at the end of last year in my blog, until we get clear and strong indications for the general acceptance of GPLv3 over GPLv2, we feel comfortable with a specific GPLv2 reference in our license. So while we support and care for the GPLv3 process, don’t expect us to be amongst the first GPLv3 movers.

By the way, don’t miss Eben Moglen’s keynote at the MySQL Conference & Expo 23-26 April 2007 in Santa Clara, California! Eben is not only a brilliant mind and shaper of the future, but one of the best public speakers I have ever heard.

Having taken a look at the 35 MySQL Summer of Code applications, I can now say that we have several very strong applications, such as:

• IPv6: Enabling MySQL to connect over IPv6 internet
• MySQL Based Atom Store (storing Google GDATA in MySQL)
• Test case development / Load testing
• Test suite development
• MySQL Auditing Software
• The “anti-profiler” - a performance analysis tool for database engines

For several of the above, we have multiple good applications. The applicants come from recognised universities across the world, not just the US and Europe. Some have extensive experience from professional software development. Sadly, we will end up having to reject several good proposals.

In addition to the challenging task of picking the best applications, we also have to prepare them for Google’s final approval. We expect to get it done by Friday, and Google to have approved (a subset of) our recommendations by the end of next week.

Today, the time closed to apply for student positions at MySQL Summer of Code. We are very fortunate and thankful to have got 35 applicants fulfilling the formal requirements of an application fully entered and coupled with a task announced on http://forge.mysql.com/wiki/SummerOfCode — thank you!

Our next task is to narrow this group down to the number of projects which can actually start. This will be a much smaller number. Even in the theoretical case that Google would allow us to have 35 projects (and they won’t), we don’t have the internal mentoring resources to guide that many Summer of Code projects.

Colin Charles has already been quite busy replying to the incoming flood of applications, insisting on them being more specific than “I’m in, just give me any task!”. Still, some applications are a bit too general for us to be able to pick them. For the others, we will group them into projects that

1. affect the MySQL Server code base
2. affect the code base of a MySQL Connector
3. affect the code base of a MySQL GUI tool
4. expand or improve the test suites of MySQL Server
5. introduce new code bases for MySQL AB itself
6. involve the code bases of entities in MySQL’s ecosystem

where the last category is quite an interesting one, where both Sheeri Kritzer and Paul McCullagh have signed on as mentors and have cool projects described on the Forge Summer of Code page.

Expect us to be done next week, in time for the Google Summer of Code deadline for picking projects. In the meantime, we will be quite busy internally pairing up the tasks with the most suitable mentors.

Through Kristian Köhntopp’s blog, I ran into a funny posting about “The real difference between MySQL and PostgreSQL” on Andreas Scherbaum’s mostly PostgreSQL related blog. I’ve taken the liberty below post a cropped version of the central picture of that blog entry:

The picture shows Susanne Ebrecht and Lenz Grimmer at the Free and Open Source Software Developers’ European Meeting FOSDEM 2007 in Brussels in February. Susanne hosted the PostgreSQL booth at FOSDEM, and Lenz represented MySQL. Drawing a seemingly obvious conclusion from the picture, one of Andreas’s blog commenters noted that “MySQL is from Mars, PostgreSQL is from Venus.”

Since last month, things have developed. Susanne has gone through a long interviewing process with MySQL AB, where I was one of the MySQLers who had the pleasure of interviewing her. Susanne has now joined MySQL, working for Georg Richter on the Connectors team.

Her joining provoked a number of knee-jerk reactions within MySQL AB. Some (but very few, I am happy to note) recent recruits got run over by their killer instincts and proclaimed some kind of crusade victory. More experienced long-time MySQLers (quite a lot, I am happy to note) instead turned their killer instincts towards these outbursts, explaining the basics of MySQL / PostgreSQL coexistence in the Open Source world.

So let me publish part of my internal message on Susanne’s joining. I’ve censored bits and pieces that would identify whom I was addressing and what that person was saying:

Like you, I’m very happy that Susanne joined us, but her task is not to “kick out Postgres” [..]”.

I envision Susanne as someone building bridges between the MySQL community and the PostgreSQL community, rather than as a “secret agent” swapping sides. I understand [… some named prominent PostgreSQL community members …] don’t see her as a black sheep, nor would that be for the benefit of either PostgreSQL or MySQL.

So to make my Community perspective clear: MySQLers shouldn’t in any way think of or publicly portray Susanne’s joining as “a prominent PostgreSQL person has now changed loyalties”.

On the other hand, I can see several good things coming out of Susanne joining us:

• we can learn to understand Postgres better
• we can understand Postgres contributors better
• we can grow our network towards Postgres and its contributors
• we can publicly underline non-hostility between MySQL and Postgres

Susanne: Welcome to MySQL!

I hope the PostgreSQL community sees similar / reciprocal advantages of building bridges towards us at MySQL.

On Thursday/Friday this week, I visited Linbit in Vienna. They are the creators of DRBD. Quoting Wikipedia,

DRBD is an acronym for Distributed Replicated Block Device. It is a Linux kernel module, that, working together with some scripts, offer a distributed storage system, frequently used on high availability clusters. DRBD works as a kind of network RAID.

This means DRBD can give high availability to MySQL users. Through configuring DRBD to be used on your system, you can have synchronous replication between two different servers, giving a MySQL database a failover server to redirect to instantaneously, should the main server running MySQL fail.

For those interested in more detail on how to combine DRBD and MySQL, let me mention that Kristian Köhntopp of MySQL has written a great blog article on “Quick tour of DRBD“.

I was impressed when listening to DRBD’s main author DI Philipp Reisner describing the technical workings and business opportunities of DRBD. In many respects, he reminds me of our very own Monty years ago.

I also learnt plenty of things from Florian Haas, Senior Software Engineer with Linbit. Among other things, he taught me that r is a vowel (in many of Austria’s neighbouring countries), meaning that you can pronounce DRBD without spelling out the letters. Sounds like “Good day!” in Slovenian.

On a more serious note, I think the prospects for DRBD look fascinating. Or in other words, remembering my recent insight on vowels: DRBD rcks!

In his well-written blog entry Open Development: Diversity matters, Gianugo Rabellino quickly replied to my blog entry from yesterday on Defining “Participatory Open Source”. He sees plenty of common ground in our reasoning, but defends the existence of requirements for neutrality in the definition. I agree with nearly all of his reasoning for why neutrality is important for the development of a community of contributors, but I draw partially different conclusions.

The main reason why my conclusions are different from Gianugo’s is the starting point for my reasoning: There is no inherent conflict of interest between participatory open source and pursuing for-profit business goals, so we must not create artificial obstacles for business wanting to develop Open Source software building upon an Architecture of Participation. Now, I can hear Gianugo objecting “But my entire blog entry and reasoning about why we need an impartial body and non-discrimination was about documenting the very existence of that inherent confict of interest”. Well, read on, and let me start by underlining where I agree with Gianugo.

First, I agree with Gianugo in that he has identified the right passage in my previous blog where our reasoning differs:

The interesting middle ground is whether it’s OK to say “no” or “maybe later” to technically sound contributions that don’t fit with the business interest of the owner of the main code base. I would argue that such “discrimination” is OK, as the “Participatory Open Source” requirements would otherwise impose so severe purity limitations, as not to be interesting for companies with a commercial agenda, like MySQL AB.

He argues that this documents that MySQL doesn’t have a neutral point of view when accepting code contributions, and I agree this is coherent with the rest of his reasoning.

Second, I agree with Gianugo in that turning down a contribution for no good reason is an excellent way of demotivating a contributor. If I know my contribution is technically highly merited, usable for many, and thus expect it to be included as part of the body code, I will get very upset if it is rejected for apparently no good reason. I won’t contribute again, and I will share my anger not just with my friends, but with the general public. Others will likely also be discouraged from contributing. This undermines the Architecture of Participation, and any good community vibrations will asymptotically approach zero degrees Kelvin.

While this seems as total agreement between Gianugo and Kaj, how can we then arrive at different conclusions?

Let me describe a scenario where I think it should be legitimate for MySQL to reject a technically valid contribution on commercial grounds, and why I don’t see it as an inhibitor for creating a vibrant community. I can even make the case fairly generic.

Basic assumption: MySQL has created a for-money-only, commercial offering, which constitutes a compelling reason for some customers to buy something from MySQL AB. This offering is of interest to MySQL users and differentiates our free offering from our business offering. It is the “something more” you get for entering a business relationship with MySQL, where you spend money to save time, instead of spending time to save money. The fact that our revenues are greater than zero is proof that there must be some offering from MySQL that our customers consider meaningful on top of what we offer at no cost to our community. Not all of these revenues are services billed by the hour of MySQL employee time spent. Some services are even based upon the delivery of software, such as the MySQL Network Monitoring and Advisory Service. Let’s use the term “the little difference” for what MySQL offers the paying customers beyond what we offer the community of users who use MySQL at no cost.

Now, should you have a contribution that renders the little difference useless, don’t expect MySQL to be happy about adopting it.

If Gianugo now says, “See, Kaj, that’s my point. If there is not a neutral body governing contributions, then progress is hampered by self-interest. So there cannot be true Open Development, and there won’t be a vibrant community.“, then I disagree at least with the second part of his reasoning, and my ambition is for MySQL AB through its deeds to prove beyond doubt that one can create an Architecture of Participation around Open Source software, while pursuing for-profit interest, even preparing for an IPO. I MySQL can fulfil the ten requirements I listed yesterday

1. software released under an Open Source license;
2. well-defined public guidelines for how to participate in the development process;
3. a publicly available detailed list of bugs;
4. a public roadmap with detailed descriptions of planned tasks;
5. code reviews accessible for the public;
6. public code acceptance criteria;
7. active public instant messaging between the developers;
8. well-maintained public system documentation;
9. a publicly documented mentoring system for grooming new developers;
10. a worldwide geographic spread of the developers;

then I think we will have created a vibrant community of contributing developers.

So what happened with the upset contributor mentioned earlier, whose well-merited contribution was turned down because it undermines our business offering, and who told all of his friends about his negative experience, and who blogged for everyone to see about why he won’t ever contribute again, and loudly proclaims nobody else should, either? Well, I think that won’t happen, and I’ve actually already communicated why, in the little sentence “Now, should you have a contribution that renders the little difference useless, don’t expect MySQL to be happy about adopting it.”

While most of the interests of the community of MySQL users are very well aligned with those of MySQL AB the company, there is a small subset which is not. Expect us to happily adopt nearly any technically merited contributions that make MySQL more powerful, easy and efficient to use. Don’t expect us to adopt contributions that eliminate or fully undermine the compelling reason to buy our commercial offering.

The virtual Gianugo sitting opposite me now objects: “So why would I as a contributor know, or even be interested in, what you consider to be your business interest, and what will undermine it?“. I hear him, I understand his question, I think it is a valid one, and I even have an answer. You don’t know our business interest. You would merely need to accept that there is a legitimate interest for MySQL to earn money on something, and the rest is taken care of by the contribution process itself. That contribution process has feedback loops between contributors and MySQL to give answers to contributors to the question “Is MySQL interested in adopting my contribution?“, before so much time is spent by the contributor that the answer “I’m sorry, no” will render him or her furious. Remember, the virtual Gianugo sitting opposite me accepts that there are technical reasons why we might see us forced to answer “I’m sorry, we cannot enter your contribution into the code base at this point in time.” — the code might not conform to our coding standards, it might be in conflict with another feature we’re just inserting, it might make MySQL too slow, it might not work with Replication, etc. etc. And if I, as a contributor, have spent huge amounts of time on a contribution, expecting it to be accepted into the codebase, I will indeed be furious, no matter what the reason for non-acceptance is. If I have an idea for a contribution that is deemed unacceptable by MySQL AB for commercial reasons, but it’s turned down by MySQL while I haven’t still spent loads of time on it in the belief that MySQL will adopt it, I will be disappointed, but I’m not going to lead a rebellion.

So this is why the yet-to-be-published acceptance criteria for code patches (item 6 in my ten requirements) will also include commercial aspects. This is why the “well-defined public guidelines for how to participate in the development process” (item 2 in my ten requirements) will state that there must be frequent enough feedback loops between the contributor and the non-neutral body managing MySQL, i.e. MySQL AB, to ensure that a contribution being developed is on track towards acecptance in the code, on both technical and business grounds.

My firm belief is thus that the Architecture of Participation is compatible with commercial interests. Non-neutral owners of Open Source code bases who strive for profit can, legitimately and without a bad conscience, exclude certain contributions on commercial grounds. And they can still establish vibrant contributing communities, by not setting false expectations. I think there is a sufficient number of developers who are willing to contribute code to such projects, because of the many common interests, including the expansion of the world of Free Software.

The Open Source thinker and BSD advocate Gianugo Rabellino just made me aware of a year-old blog entry by Susan Wu in Planet Apache on “Should OSI redefine the label Open Source?”. As I read the entry, it proposes a term “Open Development” and lists these “first stab” requirements:

1. an Open Source license, of course;
2. a non-discriminatory access to the developer’s community;
3. a well-defined and stated process for people to get involved;
4. a neutral and self-elected governing body;
5. (more difficult, could mean having a preferential lane) a neutral party such as a foundation owning the code.

I think Gianugo and Susan are pursuing a worthy cause in defining a software development model that doesn’t just release software based on an Open Source license, but also builds upon an architecture of participation. However, for reasons I don’t understand, they have included requirements that practically exclude companies like MySQL AB from ever fulfilling them. I am referring to requirements 4 and 5, and partly also 2. I argue those requirements are unnecessary: Open Development is possible even in commercial, for-profit companies.

In order not to steal or redefine Gianugo’s term, let me thus propose another term “Participatory Open Source“, with the following “first stab” requirements:

1. software released under an Open Source license;
2. well-defined public guidelines for how to participate in the development process;
3. a publicly available detailed list of bugs;
4. a public roadmap with detailed descriptions of planned tasks;
5. code reviews accessible for the public;
6. public code acceptance criteria;
7. active public instant messaging between the developers;
8. well-maintained public system documentation;
9. a publicly documented mentoring system for grooming new developers;
10. a worldwide geographic spread of the developers;

In some respects, this is similar to Gianugo’s requirements. My 1 = his 1. My 2 = his 3. With his 2 and 3, he may also refer to some of my items 3 to 10, that are on a more detailed and practical level, attempting at describing the baseline components of an Architecture of Participation in software development.

In other respects, there are clear distinctions between Gianugo’s set of five, and my set of ten. Most of all, my set of ten would work for a company with a commercial agenda (like MySQL AB), whereas Gianugo’s items 4 and 5 would not. I think it is absolutely possible to combine a commercial agenda with an Architecture of Participation when developing software, and not just in producing or distributing content (like, say, Google or Amazon). That is why I have excluded any requirements like Gianugo’s 4 and 5 from my list.

On a specific, detailed level, I also see a need for clarification in Gianugo’s item 2 “a non-discriminatory access to the developer’s community”.

• First, what is “non-discriminatory”? If non-discriminatory means “code contributions shall be accepted regardless of the gender, race, country of origin, or age of the contributor”, it’s probably OK (whereby certain countries of origin might incur exposure to legal or other risk). If non-discriminatory means “even technically inferior contributions must be accepted”, it’s obviously not OK. The interesting middle ground is whether it’s OK to say “no” or “maybe later” to technically sound contributions that don’t fit with the business interest of the owner of the main code base. I would argue that such “discrimination” is OK, as the “Participatory Open Source” requirements would otherwise impose so severe purity limitations, as not to be interesting for companies with a commercial agenda, like MySQL AB.
• Second, what does “access” mean? The right to participate? Sure. The same rights as the central governing body? No way.
• Finally, what is “the developer’s community”? OK, this may only need a phrasing clarification, as I suspect Gianugo does refer to the community of people developing the software itself (in our case, those who code MySQL using C or C++), as opposed to the community of people developing code using the software (in our case, those who use MySQL when coding in PHP, Perl, Java, Ruby, Delphi, VB, Python, etc. etc.).

At any rate, I fully agree with Gianugo that there is more to Open Source than an Open Source software license: The active participation of the community in developing the software.

Oh, and finally: I don’t think MySQL AB is yet fulfilling all ten points. But if I can have my way, we will when the year 2007 is over.

From time to time, MySQL employees are approached with proposals from universities and other research institutions about MySQL participating in research programmes financed by governmental initiatives in the US and in the EU. This is much appreciated. Often, there is a mutual interest, where

• academics need an industrial partner to anchor the results of the research into a real-world application and even to fulfil the requirements of getting a grant, and
• the MySQL user community needs new features in MySQL Server, or in the Connectors enabling the use of MySQL from various development environments.

Left: MySQL University refers to the highest form of education when it comes to learning how to develop MySQL Server (as opposed to using it). It’s opening up for the MySQL community to attend, at no cost.

So what do I suggest you to do, if you’re drafting a Framework Programme proposal (these are collaborative research projects funded by the European Commission) or equivalent other type of application involving governmental funding, and you think MySQL (whether AB, Inc., GmbH or KK) would be a possible industry partner?

• First, think about how your research can end up as a benefit for the MySQL user community. Who needs it? To fulfil what requirement? Why is your project going to end up in deliverables that are usable by the MySQL ecosystem?
• Second, think about what MySQL can do to help. We have limited resources for developing MySQL Server further, and we deploy them carefully. However, we absolutely want to help, if your research proposal provides good answers to the primary questions above. We are more likely to be able to help, the more concrete the contribution of the research partners is.
• Third, email me or anyone else in the MySQL Community Team, at firstname@mysql.com. Or if you happen to know someone at MySQL, or you know of a MySQL employee close to your geographic location, approach that person — who in turn will approach the Community Team at MySQL. We will then act as your contact persons and primary interface towards our development organisation, who in turn are the domain experts in Data Warehousing, High Availability, Geographical Data, Clustering, Replication, or whatever the specific domain of your research is about.

While none of us in the Community Team has yet acquired a PhD, we’ll connect you with those MySQLers who have — once we’ve determined that your research proposal has a common goal with what MySQL is trying to accomplish.

I’m just back from SNAP Innovation, the developers of the Primebase XT (PBXT) storage engine, here in Hamburg / Altona. What a pleasure to visit a company that, despite its closed-source roots, has understood and embraced what Open Source is all about and how it can open up business possibilities!

Left: Paul McCullagh

PBXT is a pluggable storage engine that fits in with MySQL 5.1. It is fully transactional, and supports MVCC. That makes it a good fit for many purposes, but the area where it is an especially good fit is for the industry that SNAP has been serving for the past decades: the publishing industry. No matter whether print or web, they have fairly similar needs, where BLOBs are a common denominator.

While MyISAM has supported BLOBs since the 1990s, and while the inventor of the concept of BLOB is Jim Starkey developing the Falcon engine, I can see PBXT satisfying needs beyond the mere moving the storage of pictures, films, MP3 files and other such data off the file system, into the database. Although PBXT isn’t there yet, Paul McCullagh (the PBXT author and SNAP CTO), Ulrich Zimmer (their CEO) and Volker Oboda (their director of Mktg & Sales) presented an interesting vision of PBXT as the engine for scalable BLOB streaming. While most access to the PBXT based data will be through MySQL, SNAP seeks to develop a dedicated API (over a separate protocol) accessing the streaming aspects of the BLOBs more efficiently than can be made through a SQL constrained interface.

I’m happy to note that Paul McCullagh will present at the MySQL Conference & Expo 23-26 April 2007, with a session PrimeBase XT: Design and Implementation of a Transactional Storage Engine (Wednesday 10:45-11:45 in Ballroom F), and on the panel in Lightning Rounds: State of the Partner Engines (Thursday 10:45-11:45, Ballroom H). I’m sure Paul will take advantage of the opportunity to integrate with not just MySQLers but also top PHP & Ruby developers, since I think there are some joint interests between PBXT and several development environments.

We talked about various aspects of better integrating PBXT into MySQL and our community activities. Let’s see if Paul manages to put up a task for MySQL’s Summer of Code page, for benchmarking of the PBXT engine.

Read more about PBXT in Robin Schumacher’s article “A look at the PBXT Storage Engine“. If you want to follow the ongoing development of PBXT, consider reading Paul’s blog (which is also aggregated on Planet MySQL) and take PBXT for a test drive yourself by downloading it from http://www.primebase.com/xt/.

It seems to me that SNAP intends to build up a sizable community of users of the PBXT storage engine under MySQL, and that they’re in it for the long run, aiming for their scalable BLOB streaming backend.

Our friendly Human Resources department started a new initiative this week: We have a Global Sports Day for all our >360 employees in >29 countries. That means that us MySQLers are to get off the keyboard, move our bodies, and document our activities for the rest of the company to see. We are reimbursed for expenses, and the most creative activity / picture / story will get an additional reward.

Lenz and I happened to have planned a Community meeting for today already quite a while ago, so we decided to combine this with the Sports Day. Our local version in Hamburg took us 5,43 km through Niendorfer Gehege in 0:31:35. For myself, this means that I have run a total of 161 km so far this year, for about 15 hours, at an average speed of 10,3 km (I’m sure you’re interested). For Lenz, this was the first time he moved himself this year out of “free” will on a bike.

Lenz getting ready on bike …	… and Kaj on foot.

Kaj running …

… and Lenz biking.

Getting dark in the Niendorfer Gehege.

Kaj and Lenz getting closer to the goal.

The reason I know the exact distance of 5,43 km is that I bought a Garmin Forerunner 305 earlier this year. I fear this gadget won’t be within the budget of the Sports Day reimbursement, but it’s great fun for a running geek. Hey, if it motivates me to run more, it’s worth it! Perhaps I should read up on the notes from MySQL Camp last November about how the Motion Based guys (a Garmin subsidiary) are using MySQL for spatial data.