As a guy who dabbled with databases last century, I was cheered to see young enthusiastic people today realising that they were all kind of re-inventing the wheel although the wheel will never be square.
At the Saïd Business Centre in Oxford, the All your base (belong to us) conference enthusiasts were extolling the virtues of data base management systems. We learned quite a lot.
First up was an MIT scientist, Neha Narula – a former Google employee – who kicked off her keynote by comparing databases to 70s spaghetti western movies starring Lee van Cleef, Clint Eastwood and oh we can’t remember who the third geezer was.
Narula said there were heaps of problems facing RDBMs engineers. One of those is that large datasets, like those of Facebook, Twitter and Reddit, needed low latency – that is they had to be fast – and also needed complex queries to be applied.
The subject of her keynote was caching, and she pointed out that open source projects lags behind database science – which as we old timers know and she confirmed – were set out in papers way back then in the 20th century.
She also said that developers faced a new challenge – and that is fat clients. By that, she means smartphones and tablets because applications have to work on servers, browsers and clients. People have tried to resolve the problem by writing code for every layer, she said.
Developers, and there were many of them in the Nelson Mandela lecture theatre in the Saïd Business Centre, needed to look back to the 1970s and 1980s to find that smart people had worked out a lot of this stuff decades ago. However, she pointed out, and she confessed she was somewhat jet lagged, the really bad thing about modern databases is code complexity, with people having to hack around to make everything work properly.
The conference is part of the Digital Oxford week, in an early attempt to make the other place – that is to say Cambridge – realise that this town is a force for technology for the future.
The conference is in its second year and there's more to come in TechEye coverage, later.
↧
Databases: They’re good, they’re bad and they're ugly
↧