Today at a BBC historical exhibit in one of our buildings, I read a "strictly confidential" letter. It was from a manager at the BBC replying to someone that yes, they would be interested in discussing whether an "electronic brain" could help reduce BBC clerical staff.

The letter was dated 22 November, 1949.

Many folks wonder what exactly I do at the BBC and frankly, this is a tough question. The short answer is "metadata", but that doesn't tell anyone anything. It's much better to show it. So here are the TV listings for BBC One. Pretty boring, eh?

Here are TV listings for BBC Three (requires Flash), powered by the data my team creates. We didn't create the front end, but the richness of our metadata makes a lot of things like this possible. Our data is actually used quite a bit internally, but this is the first public exposure of our data.

Update: Looks like they've taken the Flash version down for a bit, so you can't see what I was referring to right now.

Of course, if you're British, then you probably have heard of iPlayer. That also uses our data, but doesn't expose it.

For our next trick, we'd like to publicly expose all BBC programme data going back to 1936. It's a lot of work and importing the data into our system is tricky. I understand that there's a team of 40 people assigned to cleaning up the data enough so we can import it. We may not be able to make it public for quite a while, but we're looking into that (some of the geekier amongst you may have heard of RDF ontologies). Today I've been creating mocked up data to give us an idea of the impact of the data on our system. We don't even know how much data we'll be importing, but one rough number we've heard is 40 to 50 million broadcasts.

I won't be typing those in by hand.

Recently, on a post to a London mailing list, my friend Andy pointed out that someone on a BBC blog had posted a very detailed error message they received. Not having lived in this country for long, the only BBC bloggers I've read have been tech people and I was thus surprised that someone techy would post something which could compromise our security.

I suppose I could have dashed off a nasty email demanding to know what the hell he thought he was doing and didn't he realize he was telling the world that this software was vulnerable to algorithmic complexity attacks?

Fortunately, though frequently an idiot, I'm not frequently an ass, so I merely sent the blogger a polite email asking if he meant for that entry to be public. The poor guy seemed surprised by the question so I felt it was important to point out some of the security implications. As it turns out, he was not technically inclined, so that was a bad assumption on my part, but he was gracious about it and pulled the offending bit from the blog post.

I kept this to myself as I didn't want to get the guy in trouble, but later in the week, I'm sitting in a security meeting and I debated mentioning this. I didn't want some manager going down and giving this guy a thrashing, but this information leak was not only unfortunate, it revealed some rather dodgy details of our gory internals (of a system which has since been upgraded, thank goodness). Prudence won the day and I mentioned the post. One of our bosses asked me to forward the email and so I did, with the following caveat:

Here's the information about the blog post I mentioned. I hadn't said anything before because the individual in question was apparently genuinely surprised that there was a security issue here and I didn't want to get him in trouble.

My boss didn't comment on this, but it was only this weekend that I discovered that the blogger, Eddie Mair, is is one of the most well respected radio presenters in this country.

Hoo boy. Way to make Americans look like either complete morons or arrogant bastards :)

There's a tool we've delayed releasing for several weeks. I've been tasked with making sure it gets released.

Me: You mean we've delayed this tool for several weeks because we've not been able to add a feature they didn't ask for?

Him: Yes, but ...

Me: And we could add this feature they didn't ask for, but we didn't use the software which could support this unrequested feature because it fails to support another feature they didn't ask for?

Him: Well, yeah.

*brain melt*

It's really fun working for the BBC, but the occasional lack of focus can be frustrating at times :)

I've recently entered the following curious command on my computer:

PS1='\[\033[01;32m\]api_cleanup\[\033[00m\] $ '

This makes my command prompt look like this:

 api_cleanup $                 

So why would I do that? Well, the prompt on my computer usually shows the current directory name and that name, when I'm developing new code, is the name of the current branch (copy of our code base) that I'm working on. When I asked my colleague Richard what branch name I should give our current work, he said "name it anything you want".

Heh.

So when he comes back from out meeting, he'll assume that the current branch is named api_cleanup. Had I not changed the prompt, when he came back and started pair programming with me, he would have seen this:

 richard_and_curtis_mud_wrestling_extravaganza $                 

It's going to be fun when it's time to check in the code. I'm such a geek.

It's officially announced: a beta of the new BBC Homepage is online. Curiously, the Brits pronounce "beta" as beat-uh. Drives me up a wall.

They also pronounce "router" as root-er. In some parts of the US, "rooter" is a slang term for "pig", but it's also a slang term for "sex" (more common in Australia, I believe). I've difficulty keeping a straight face whenever someone refers to a root-er, but then, I've been known to pinch a fag from time to time.

Since I work for the BBC, I've had some people ask me about the "Perl On Rails" project and I've not said much. Now I have.

"If you'd like to speak to X, press 1. If you'd like to speak to Y, press 2. If you'd like to speak to ..."

All things considered, that's a rather confusing thing to hear in an elevator. Seems the BBC is trying to work out some issues with their phone system.

My first three days have been fantastic. The code is reasonable, there are plenty of tests and there appears to be real project management. Despite a few bureaucratic hurdles, things have been relatively smooth. I see plenty of areas where I can bring my testing background to bear and plenty of other area where the other programmers will be able to teach me new things. I'm quite happy so far :)

Interestingly, the blogging policy appears to be extremely liberal. I can't talk about money, rip on other employees or give away obviously secret information, but other than that, I just need to be "professional" when I talk about them and that's pretty much it.

And that's why I'm allowed to say that the BBC, after spending a lot of time and effort, has finally stopped using Ruby on Rails on one project. After a couple of years, aggressive caching, several high-end servers and a lot of effort at customization, page loads for one application were still taking nine seconds. That's far too long for the Web. Since the BBC has a tendency to write huge applications with tons of data, it appears that they could not get Rails up to the task. They might be using it in the future, but for the time being, it's just not scalable enough.

Oh, and their coffee is crap, but hey, they're British.

In a few minutes I'll head off to work at the BBC. I'm to show up at 10 AM today. I never show up that late for work, so I don't know what to do with all of my free time beforehand.

Enough people have expressed curiosity about the BBC that I've asked them what their blogging policy is. One person mentioned it should be fine and another is going to double-check higher up. Even the brief conversations I've had with them have been absolutely fascinating. The project I'm to be working on sounds fascinating, looks like a lot of hard work, and I was thoroughly mystified when it was outlined to me a few days ago. I finally realized the problem: I couldn't identify the goal since I couldn't see the profit. The project itself is important in terms of providing more information to the public. It's just not valuable in the "profit" sense. After well over two decades of working in profit-driven industries, this is going to be a very interesting change of pace.