Real World Computing
Hacking the world
At the other end of the spectrum is the Nabaztag WiFi Bunny. To those of you who aren't familiar with the Nabaztag, it's a gadget that's both interesting and irritating. It's a conical plastic device with two large ears a few inches tall. It contains several coloured lights, a couple of electric motors that make the ears swivel, a speaker and a microphone, all controlled via a microprocessor that has Wi-Fi capability.
After a bit of persuasion and configuration, you end up with a networked rabbit that can receive messages from around the web via its home server. Its main ability is to play back text as speech, so it can read aloud RSS feeds and the like - it can also play MP3s, so it can relay podcasts to you or, more interestingly, MP3s that you upload to the Nabaztag website. This combination of technologies can prove useful - like reminding you of the time and reading out interesting headlines - or else incredibly irritating, when it tells you the time late and reads out those headlines in horrible pseudo text-to-speech English.
However, where it really comes into its own is by offering an API via which anyone can send messages to your Nabaztag. I do need to elaborate slightly on that "anyone", because they'll need to know the address and the authorisation code for your Nabaztag, but if they do then they can make it talk in a collection of different voices, flash its lights and make its ears go round.
At this point, you'll either be with the majority who are thinking, "What use is a remote-controlled talking rabbit?", or an equally large and overlapping group that thinks, "Why am I still reading this, these guys have lost their marbles", or in a tiny elite group who think "A remote-controlled talking rabbit could be really useful".
I'm a member of the latter category, and I suppose I need to justify myself in a hurry using a simple example. In several past columns, Ian and I have discussed the Nagios system-monitoring software, an open-source product currently staggering into its next major revision, version 3, which has become the de facto standard for systems monitoring in the open-source community and widely used in the mainstream, too. Its combination of ease of use, versatility and extendibility means Nagios can be used in all sorts of environments: we use it ourselves at Wide Area, and I know of many other organisations that also use it.
Now, when Nagios finds a problem on your network it has to communicate that fact to a human being to get it fixed, and the problem with problems is that they're often created by people in the first place, and typically the person who created the problem doesn't realise they've done it. As such, when a problem arises they tend to not think of that problem as theirs. A typical case is the "server has run out of disk space" event, where the person who was creating a huge file on the server doesn't make the association between their action and the event, while other people tend to ignore the problem because they know they weren't working on that server at the time.
Now there are various ways in which Nagios can inform people that there's a problem: it can send email, but that's potentially slow and can be easily ignored by the recipient; it can be hooked up to an existing SMS or texting system, but that may be expensive and may have to be broadcast to everyone in the office; instant messaging is a better solution but, again, can be easily ignored. Enter the Nabaztag. Suddenly, whenever a problem occurs, the rabbit can inform everyone by speaking the problem aloud to the whole office. It's very hard to ignore a talking rabbit, particularly now that everyone else in the office heard what it said and realises it was you that caused the problem...
