Comments on: MySQL getting too big for its corporate britches?

MySQL has a problem: the project has become so important that many have forgotten the company at its heart.

[astamer]

The popularity of open source database software such as MySQL and PostgreSQL notwithstanding, its reputation as a cost-saving strategy did not fully hold up in Aberdeen Group?s recent benchmark study on Protecting the Database (November 2008). In the areas of security, compliance-related audit deficiencies and ongoing operational costs, the performance reported by current users of open source database software was below that of the Industry Average.

The following links provide complimentary access to the research:

· Open Source Database Solutions: When Low Cost / No Cost, Isn?t - http://aberdeen.com/summary/report/sector_insights/5750-SI-open-source-database.asp

· Protecting the Database: When (Most of) the Eggs are in One Basket - http://www.aberdeen.com/summary/report/benchmark/5302-RA-protecting-database-security.asp

Andrew Stamer, Research Associate, Aberdeen Group

[mdcallag]

Matt,
Are you calling for an open or closed source fork? Why can't MySQL monetize the forks and patches done by the community? They have the expertise and the brand. What keeps them from selling support on this and from providing binaries with community patches to MySQL Enterprise customers? Their support is excellent and I can't imagine customers not wanting it for code produced by others.

[Matt Asay]

I'm not really calling for a fork to the core, but rather to more additions beyond the core that Sun can provide to enhance the MySQL experience for paying customers. I worry that the community forks are diverting revenue from Sun, making it harder for the company to invest in MySQL core development. Why can't Sun play this same game, but to the community's betterment?

[jrepenning]

If Sun forks, who would be left on the trunk? If Sun forks, who will deliver their core changes back to the trunk? I think Sun needs get some skin in the actual open-source, communitarian part. Whether they build their proprietary value on top of that, or in a fork, can't be clear until they clearly establish custodianship of the community.

[emacsuser]

A better solution is for Sun to repackage the improved MySQL bits from Drizzle, OurDelta, Perconaand and charge for support and upgrades ..

[Matt Asay]

This would be better if enough people paid for these, but I can tell you from 10 years of experience selling support for open source that it's a very weak business model. No one makes serious money doing that, including Red Hat.

[brian_aker]

Hi!

You say "it's fair for Drizzle ... to enhance and extend the MySQL experience, then it's fair that Sun do" like Drizzle is something that Sun is not involved in. We have seven people internally now who work on Drizzle for their job (which doesn't include the others who work on it in their spare time but are still Sun employees). Sure, Drizzle has a lot of folks who do not work at Sun who contribute, but Sun is still the largest single contributor.

Cheers,
-Brian

[Matt Asay]

Yes, Brian, but that's not being entirely forthright. It was not started as a Sun-blessed fork, and I'm sure that the business executives would have preferred to hear in some other way than through blogs. :-)

[malcolmspence]

the problem is that Sun spent $1BN on MySQL and so now they have to figure out a way to recover that cost, or as they say, monetize the product.

Red Hat has the same problem with JBoss. They all have to fork the code, add value to differentiate themselves and the original idea of open source (lots of users adding their feature and fixes) goes out the window.

Malcolm in St Louis

[arjenlentz]

OurDelta (http://ourdelta.org) has most accurately been described as a "distro for MySQL", similar to how Red Hat and Ubuntu create distros of Linux. After all, we know that a tarball does not make a product.

[Matt Asay]

Agreed on OurDelta. In my effort to be concise, I threw that in as a fork, but it's not. It's a fine distinction, but I do think it's different.

[Travis5650]

Forks are inherently expensive things for an infrastructure software provider to do, open source or otherwise. Especially when you have lots of customers in production. Customers want reliable bits that have been tested and certified by a professional QA entity. Customers pay for support, QA, and a reliable upgrade path forward. Forking would only weaken Sun's ability to ensure this for customers, unless of course they plan to invest wildly in QA infrastructure and practices(but Monty's recent comments didn't exactly describe this....yet). Unless Sun has a darn good strategy for monetizing, spending more for what they already have and potentially destabilizing their ISV and "to be upgraded" customers, would be a bad decision.

[ZUrlocker]

Let me add to Brian's comment and be clear: Drizzle is a Sun project. It involves folks outside of Sun also though its still a bit of an experiment.

--Zack