The conference is now over and I am glad it is. It was an intense four days of information exchange and meeting interesting people. The feeling I now have is twofold: One end says: "Woohooo! That was great, got to do that again."; The other is: "But not anytime soon!". Thanks to Jay Pipes for putting a lot work into making this happen.

And special thanks, of course, to Jeremy Cole and Eric Bergen of Proven Scaling who spent some serious coin to have J.R., Carlos and me attend. You rock!

I couldn't find an official announcement, but there will be a MySQL Conference in Tokyo later this year. Now, what talks could I submit for that? :-)

Rasmus, in both his session and his keynote, painted a very good picture of the state of security of current web applications. Namely, there is none. Even if you make your website as secure as possible, an attacker is able to use your users and their out-of-date (or even current!) software to steal data and brake into systems.

"The web is broken you can all go home now."

This has nothing to do with a certain database vendor. Paul Tuckfield of (first) PayPal and (now) YouTube fame described a nifty method to speed up MySQL Replication in his keynote at the MySQL Conference.

The pitch is simple: He implemented an oracle algorithm. In more detail: In a MySQL setup, several layers of caches exist. In front of the database sits the query cache that saves the MySQL from doing actual work (query analysis, execution, disk I/O). The next two levels sit between MySQL and a the actual disk. The first is the filesystem cache that the operating system provides. The second is the block-level cache that the RAID controller provides. Considering the last, from a MySQL point of view, data has been written, when it hits the RAID controller.

All these caches serve one purpose: Reduce execution of complex or slow operations such as physically reading data from a harddrive or writing to it. MySQL Replication, in a …

First day's over — *phew*

I got to meet a lot of nice people, some of which I haven't seen in years (Heya Lukas and Damien S. :-) and it was great. There was a lot to talk about during the breaks and at dinner so it practically didn't get boring for a solid 12 hours today. I expect 14 to 16 of that tomorrow, maybe more.

The first tutorial I went to was Jeremy's and Eric's MySQL Scaling and High Availability. There are already two decent summaries up, so I won't repeat what written there. I just want to emphasize …

There are a lot of The Conference starts tomorrow posts up right now, so here's mine.

I got down to Sunnyvale this morning leaving downtown San Francisco for five days. I stay at the Pacific Inn in Sunnyvale which is a convenient ten minute ride from the Santa Clara Convention Center, where the conference starts tomorrow.

I met a few *cough* guys from MySQL already, which is great, it's been years since I've seen some them. I look forward to meet all the rest you!

Also, make sure to follow my twitting through the conference.

And I finally got to meet Damien, woohoo! ;-)

In the spirit of April 1st — only the other way around — I'd like to disperse a spoof I've been working on for quite a while now. Last autumn I began to tell people about a new, revolutionary database called "CouchDb".

I explained CouchDb’s advanced design and state-of-the-art feature set, leaving everyone with the impression that a lot of technical problems that exist within traditional database systems have been taken care of. For example, CouchDb’s on- and offline replication feature is the basis for a super-easy to implement load balancing and …

During my work for Journalizer I created a plugin for TinyMCE. What does it do? Well, it solves a problem, but which one? We have list of TinyMCE-controlled textareas and it is so common that you use the tab-key to jump to the next field that we wanted that as well. Unfortunately, this doesn't work in TinyMCE.

Why is that? When using TinyMCE, you declare a HTML form element as "managed" by TinyMCE. What it does is substituting the actual HTML element with an iframe that can be edited like rich text. I.e. you can make things bold and italic and all that just with a click on button and you immediately see the results — nice. But because the actual textarea gets replaced by the iframe, common "tabbing" through your form elements no longer works because …

If you use Cocoa Bindings and have an array of NSStrings where one entry controls a check box in your user interface, you need to provide a Value Transformer in Interface Builder.

I couldn't find one, so I wrote one. The core are two methods:

- (id)transformedValue:(id)value
{
  if(value 
  && [[NSString stringWithString:value] 
  isEqualToString:@"1"]) {
    return [NSNumber numberWithBool:YES];
  } else {
    return [NSNumber numberWithBool:NO];
  }
}

- (id)reverseTransformedValue:(id)value
{
  if([value boolValue] == YES) {
    return [NSString stringWithString:@"1"];
  } else {
    return [NSString stringWithString:@"0"];
  }
}

They convert an NSString to a BOOL and vice versa. This is not exactly true, as these functions …

[Update 28/1/2007: New Version and new website.]

Amit Singh ported FUSE to MacOS X. This is just cool. Thanks!

I created a little Cocoa GUI around sshfs to manage and mount your remote drives via ssh conveniently from a program and not from the commandline. This is early alpha software and may break things. It is not heavily tested and could sleep with your wife while beating your kids, so BEWARE! Additionally, the software it interfaces with is in its early stages as well and it includes a custom kernel extension (written by Amit, so I trust it). You have been warned. It's so new that it doesn't even have an icon. But sssssh!

See this announcement for …