This blog is about OurDelta, my visit to PGDay (and resulting quest to search for a good nntp reader for OSX) and my long term search for a good set of bluetooth stereo headsets. So lets start with OurDelta. The other day Arjen pokes me about OurDelta. The idea is to offer a place for distribution of all those tasty MySQL patches that float around the web (like from Mark, the Google guys etc.), that simply do not fit in MySQL's research schedule. Obviously this is awesome. There are packages for all sorts of distros (I am sure Windows will come one of these days too), which takes away some of the scaryness for people not comfortable with building things themselves. Moreover you know that there are other people that are using the same binaries and I guess one of the key things that OurDelta could build is a better way to communicate about success and failure when using some of these patches.

Which brings me to testing of MySQL in general. Actually I think this is the number one challenge that the guys at MySQL have to meet. They have to make sure that there are a bazillion tests for MySQL. They need this so that storage engine authors have a chance to not make the world at large suffer from unwanted engine interoperability issues. Its the only way for them to get their quality issues in order (and having to delay 5.1 for so long tells me they are having quality issues). Its also great for Drizzle guys to be able to know where they are diverging from MySQL. It also opens up the chance for such patch projects to ensure that things are running smoothly.

Why should Sun/MySQL care about these patch projects? Well thats because things have changed now that they are at Sun. They can now finally leverage the community not only for distribution and 3rd party application development, they can finally make use of the community as a development method too. Of course only if they can make sure that they can accept patches in a realistic timeframe. In this sense OurDelta should be seen as godsend and a friendly kick in the butt.

In other news, I went to Prato for the PGDay held there. David asked me if I would be there on Tuesday and after finding out that every developer at Liip as a personal budget to spend as he wishes for educational purposes I decided to take the night train to Prato, attend the second day of the conference and take the night train back again. It was nice to meet up with the PostgreSQL crowd.

Also reminded me that now after over 2 years of the wiki it might be time to review the content. Seeing where things went well, where things didn't go so well (like stale/outdated content, redundant content, content that should have gone somewhere else etc.). It also reminded me that I have lost touch with the PostgreSQL community since I moved to OSX for one of the lamest reasons. Ever since I dropped Thunderbird for Mail.App I did not have a good NNTP reader, so I stopped reading the hackers list. After a bit of searching I found Pineapple News, which seems to do the trick. Once PHP 5.3 discussions have come down to human manageable levels again, I might now take on the project of reviewing the PostgreSQL wiki (might also become relevant for wiki.php.net in lets say 6 months, so its good to have done it before).

I might also look into collecting requirements for an expanded storage format for EXPLAIN ANALYZE. I think now that PostgreSQL is getting WITH RECURSIVE, it becomes interesting to be able to return the data in some SQL manageable form instead of just a text blob. There has been talk about exporting to XML (there was already a patch for this, but it was deemed too simplistic). But I think its easy enough to get to XML from proper relational data, so why jump directly to XML?

Finally I would also like to mention that I am in the market for a new set of stereo bluetooth headsets. I have owned the Plantronics Pulsar 590's for the past few years. I was quite happy with them, aside from the blinking blue light and the silver look. But since they got stolen at a recent frisbee tournament I am not pondering if I should get them again or go for something else (less blinking, black and BT 2.1 support). Nokia just came out with the BH-504, but its hard to get any information on them (let alone a price). Sony and Philipps also have a few available. In brief my requirements, which the Pulsar all met (aside from BT 2.1 and non obnoxious looks):

• 7h+ of playing time
• BT 2.0 or better 2.1 support
• Sturdy enough (ideally foldable) to survive my backpack
• At least half open (I do want to know what is going on around me)
• Comfortable wear, but should not fall off when I make a quick sprint for the bus
• Chargeable via USB
• Should also support a classic wired connection for in flight entertainment
• Non obnoxious looks

Well thats because things have changed now that they are at Sun. They can now finally leverage the community not only for distribution and 3rd party application development, they can finally make use of the community as a development method too. Of course only if they can make sure that they can accept patches in a realistic timeframe. In this sense OurDelta should be seen as godsend and a friendly kick in the butt.

Actually nothing has changed now that they're at Sun. MySQL can make use of community development as much as they always have. It's always been a scheduling thing and has nothing to do with being at Sun or not.

Where did you get the impression it did?

At any rate, it should be a kick in the butt for Sun/MySQL -- I can submit a patch now and I might see it in 7.1, 3 years or more from now....

It seemed to me like there was this paranoia at MySQL AB about outside code somehow diminish MySQL AB's position as the prime source for MySQL related consulting. Or at least that was the only thing I can think of that explains why MySQL AB seemed so disinterested in letting people help them.

The funny thing is, I am sure if they would have opened up earlier (not that they are there yet, but it seems like they are finally serious about getting there), they would have better QA measures in place today. Then again, maybe they would not have made such big bucks for their VCs.