Web Services

So AudioScrobbler is in the midst of some Planned Downtime (as is ‘sister’ site Last FM) and HaloScan is currently having a few issues since introducing a new server. Blogrolling seems to be the most stable of the bunch now, but then there is a business behind that service.

One thing that everyone who uses these things (and this harks back to the initial BlogShares offering as well) needs to clear on is that, on the whole, they are NOT being run by large, funded corporations. In fact they aren’t being run by small, un-profitable companies.

AudioScrobbler has been asking for donations (although I’m not clear on exactly who is behind them and if there is a business model), and by his own admittance, HaloScan is NOT a business it is run by a full time college student (I often wonder if Jeevan has a life at all as he’s constantly answering questions in the forums).

The obvious downside of these scenario’s is that, whilst they appear to be a ‘product’ that has been researched, designed, planned, and is being maintained, and whilst they appear to be complete and fully functioning, they are not. What they lack, and what a business plan would’ve accounted for is scalability. It’s a common theme and completely understandable.

scenario: You come up with a ‘killer’ idea for a web-based thingymejig, you ponder it a bit, maybe code up a demo to see if the idea works, then you polish it, publish it and announce it to the world. You are working alone and do all this when not working/studying. Naturally you do wonder what would happen if hundreds of thousands of people started using your new thingymejig, and you probably have an idea of what server load will be required to handle all that traffic. In fact you may even have a highly detailed plan of how to scale your new thingymejig but that doesn’t really matter, does it? It’s only you. You aren’t going to spend too much of your own hard-earned cash on this are you? You don’t even know if it will catch on, so you try and attract people to your new service and spend the next year firefighting. /scenario

OK, I’m making several assumptions but you get the idea. None of the people involved in these services wants to let people down. They are, rightly, proud of their creation and want it to grow.

Now, take a ‘funded’ web service. Let’s take something well-known and widely used and backed by a few billion dollars – MSN Messenger.

Ever tried to sign in and not been able to?

My point is this: web services fail sometimes. Ranting and raving about it won’t stop that happening. Granted the ‘small guys’ suffer more failures than the big guys (I don’t ever remember Google giving me a 404) but when you use their services you need to buy into the idea of the service as much as the implementation. Support the little guys, for one day they may be bigger than you.