You may or may not have heard of it, SQLite has them, but what is a manifestation type (I'll call them mtypes from now on)?

You can think of an mtype as a untyped variable with a twist. That is you can assign anything you want to it without having to first declare its type OR size. But first lets look at a typed variable example using the C programming language. In C if you want to store a string you will need to declare a fixed char array or create a char pointer to some malloc'd memory cast as char:

// fixed length array
char str[28];

or

// malloc'd fixed length array
char *str;
str = (char)malloc(sizeof(char)*28);

You need to know ahead of time how much space you will need to store a string. If your string was longer than 28 characters you would lose everything after character 28, safely off course by being sure not to write more than 28 bytes ;) …

Matthew Aslett is reporting on MySQL's mostly unnoticed licensing change. As he notes, it's not really a change, so much as putting a stake in the ground to keep MySQL on GPLv2 for the future. As Kaj (VP of Community for MySQL) notes on his blog:

MySQL has today refined its licensing scheme from “GPLv2 or later” to “GPLv2 only“, in order to make it an option, not an obligation for the company to move to GPLv3.

Specifically, this means that copyright notice in the MySQL source code files will change from referring to “either version 2 of the License, or (at your option) any later version” to “version 2” only, in the MySQL 5.0 and MySQL 5.1 code bases.

This is not a once-and-for-all decision, but rather gives MySQL breathing room to wait …

One of the unique elements of MySQL is the ability to use a different storage engine to store your data. You can even mix and match storage engines within the same database.

read more

This is a weird world. I’ve learnt that the Falcon tree is now public thanks to following Brian’s blog (which thankfully I follow, not just thru Planet MySQL, which is also unreachable), and reading Slashdot’s RSS feed. I was sure there was going to be email, and today, I’ve managed to suck mail from the MySQL’s mail server thru some convoluted means and find the announcement :-)

The Taiwan quake has made Streamyx’s quality of service extremely low. More importantly, it’s not allowing me to write or get mail (something I haven’t done in a while). I sincerely hope that no one’s doing silly BitTorrents or podcasts (I think downloading MP3 files fail immediately, or I just can’t even connect to PodShow and get my Daily Source Code), and that some magic routing starts working.

…

Digium Unveils New Software Appliance AsteriskNOW, Digium (Press Release)

MySQL is on track with new storage engine, Computerworld, Jeremy Kirk (Article)

Asterisk an under-appreciated Open Source Success Story, O’Reilly Radar, Tim O’Reilly (Blog)

Mozilla’s 2005 revenue: $52.9 million, News.com Open Source and Standards Blog, Stephen Shankland (Blog)

A hypothetical New …

This week’s podcast is an interview with MySQL’s Stewart Smith, software engineer for MySQL Cluster. Straight from the warm southern hemisphere, Stewart talks about Cluster. Because we gabbed on and on for 19 minutes, we’ve stripped the format down a bit, having the feature as pretty much the only segment.

You can download all the oursql podcasts at:
http://technocation.org/podcasts/oursql/

Direct play this edition at:

http://tinyurl.com/yd2ccc

Feature

Main cluster website:
http://www.mysql.com/cluster

Cluster documentation (in the manual):
…

There's an interesting article called "Open Source 2007" that appears in Educause Review magazine.  The irony is that the article was published in 2004 and was somewhat avant garde at the time.  While it's hard to make predictions (especially about the future!), it's clear that most of the ideas on open source have come to be accepted.  Brad Wheeler, CIO of Indiana University and an open source advocate with the research and statistics to back up his views, makes a compelling case for why open source will has already become mainstream. Adoption of open source software --both infrastructure and applications-- has accelerated faster than many have predicted, though perhaps it has become more accepted in commercial markets and among high-tech companies than in academia. 

…

The folks at del.icio.us must have made it harder to get on to the del.icio.us/popular page over the past few months.

When I first was listed on there you simply needed 10 bookmarks within the past 24 hours. I was looking at the article I published yesterday MySQL Optimization Hints and noticed that it has 28 bookmarks over the past 24 hours, and has not yet been listed on the popular page.

It appears that they still require only 10 recent bookmarks, but they have shrunk the window for what they call recent. It must be down to 10 bookmarks within 12 or 6 hours. Does anyone know the specifics?

Let me present some credentials, I've been a longtime Redhat user. I switched to using Redhat somewhere around the 6.0 releases, before that I used Slackware. Sitting around for my use is OSX, one Solaris machine, and one box running Trixbox (which is CentOS underneath).

Everything that I do in a production fashion is running Fedora Core.

Have I had problems? Not really. I keep fairly current so I've not needed the Legacy Project. In the one case I do need it, for an RH 8.0 box, I consider the existence of the 8.0 box to be a problem (one I am actively trying to solve).

Why do I need it? Because I picked some software to use on the box that won't run on a modern FC setup. I bet on the wrong horse and I only stick with it because its a lot of effort to get off that horse. I'm working on it and I hope by the end of the month I can kill the 8.0 box, but I am not holding my breath.

While I …

There was several changes in InnoDB to fix scalabilty problems, so I ran benchmark to check
new results and also compare overall performance of InnoDB in 5.0 and 5.1 before and after fixes.

Problems in InnoDB that were fixed:

• Thread trashing issues with count of theads 100+. In this case
  performance of InnoDB degraded dramatically. The problem was in the mutex impelementation
  and was fixed in 5.1.12 (more info about InnoDB mutexes)
• Scalabilty issue, the well know bug 15815, that was fixed in 5.0.30 and 5.1.14.

So I took for tests:

• 5.0.27, the release with both problems
• 5.0.32-bk (snapshot from 26 Dec 2006) - with 15815 bug fix
• 5.1.12 with fix of …