The latest from the MySQL community

I had been wondering this lately, and so have been asking people: who are IBM's software customers? My company sells into a wide range of Global 2000 companies, but we almost never bump into IBM databases or application servers (or hardware, for that matter). I can count the number of times on two hands, yet we often run into Oracle, Microsoft, BEA Weblogic, even Sybase. Rarely IBM.

IBM is doing a ton of revenue in software sales. But to whom does it sell its software?

...

Is your MySQL server doing an extra index scan on queries that need to check a key for matches or NULL? It’s easy for this to happen accidentally, but it’s also easy to fix, especially in MySQL 5.0 and up. Here’s how. If you read the manual page for EXPLAIN, you’ll see the ref_or_null “join type” (I think “access type” is really a better term) mentioned. If you see this in EXPLAIN, it means MySQL has to search the index for matches, then search again for NULLs.

A good three months ago, I had the pleasure of being the referee for seven DB egos. Now, the match has appeared as a Dev Zone article:

One of the most popular keynotes of the MySQL Conference & Expo 2007 was called “The Clash of the DB Egos”. It was a fight amongst seven database luminaries, all playing an important role either within MySQL AB or as providers of Storage Engines that work closely with MySQL. This article attempts at giving a picture of what the fight was about, through reciting the egos and the questions posed to them by the referee.

This is a small trick and it is, conceptually not new. It is common practice to add version numbers to resource files of a website. These are image-, CSS- and Javascript files. The reason to do so is to make sure users do get new versions of the files when you change them instead of when their browser decides to flush its cache. And you certainly don't want to tell your visitors to do a shift-reload for each small update you do. So you could do something like

    <script type="text/javascript" charset="utf-8" src="prototype1.50.js"></script>

and update the version number whenever you change things. This is a maintenance nightmare, though, because you'll have multiple versions of the same file lingering around next to each other and all the beauty of version control (You do use it, right?) is lost.

What you can do instead, is adding a GET-parameter to the filename, like this

I got this:

in the mail yesterday .. Thnx MySQL !

I went to the keynotes, took some pictures, which I will post to flickr.

While walking to my first session of the day, I ran into one of the OLPC people. Carrying around those bright plastic devices is a great calling card. We chatted a bit, and he told me what evil underhanded stuff Intel, Microsoft, and the Gates Foundation has just pulled last month that seriously threatens to destroy the OLPC.

Now I'm in the "Managing Technical Debt" by Andy Lester. Good stuff, careful notes. But mostly stuff already know.

The exhibition floor. Someone has a Penguin Robot! $85. Runs Python, controlled by Python. Can be used as a VoIP phone. It reads email and rss feeds.

Someone is showing off their 3D printer in the lobby. It can actually print it's own parts. The guy says that the version 1.1 will be completely self-fabricated. It will need a couple of bucks of cheap electric chips, and then clever …

Recently I just started working on a new Open Source website (Crash at Mine) as an alternative to CouchSurfing.com (you may have seen my blog post about that yesterday - I was trying to get it to *not* go on PlanetMySQL).

It's still early days in development, but I wanted to make sure that both:

• Performance won't just tip off when the data set becomes to large to just fit in memory.
• Contributors have a way of working with a semi-realistic dataset, without having to have access to real data (honest truth: real data doesn't exist yet).

So, I googled away, and quickly got the top 100 female and male first names. Then I googled for popular last names, and got 1000.

200 x 1,000 = 200,000 dummy users.

…

Over the last several months using Zimbra Desktop, I?ve become very dependent on having my Zimbra email available offline, and now we're happy to announce in the latest version Zimbra calendar is available offline too.

For those less familiar with Zimbra Desktop, it?s among the next generation of messaging clients- you get all the benefits of a rich AJAX email application in the browser (fast search, mash-ups, client-like UI, etc) offline and none of the downside associated with traditional clients (platform dependence, stagnant innovation, storage woes). Since its inception, Zimbra Desktop has been a big hit.

Starting today you can download the lasted version …

Interesting enough this year people at OSCON do not show to much interest in the MySQL, and Databases for that reason. Our talk comparing performance of MySQL Storage Engines had probably 20-30 people, Monty's talk on MySQL Source Code had about same number, and as Monty told me he never had so few people on his talk. This is from massive 2600 people attending the conference.

The other Database related talks I visited had similar size of audience - talks on migration To and From MySQL, Josh Berkus talk on Performance Optimization (this one if about full stack), Developing using SQL Lite. On other hand front end related talks such as PHP performance optimization or High Performance JavaScript were attended extremely well.

MySQL BOF supported by bear and pizza had about same number of attendees. With a lot of familiar faces.

MySQL booth was not to active ether - in many cases I could see lonely Kelly sitting on it. But This …

Crap!

     37 if (args->arg_count == 2)
     38 {
     39         strncpy(message, USAGE_ID3_PARSE, MYSQL_ERRMSG_SIZE);
     40         return 1;
     41 }

The problem with using code in talks?

You find bugs!

And what happens when you find bugs?

You want to fix them.

And fixing bugs is more fun then doing slides...