This morning I was talking to a guy who was leading a research project. They had a database which, he felt, was abnormal (it was a flat file database—how is that “denormalized”). Anyway, the data were being entered by one person at one terminal, and the person couldn’t keep up. So they wanted to set up a web-based interface to do this.
Now that’s where I think the group think started. Why web?
So they went to the computer gurus who told them that then would need some kind of complex SQL relational database that then would translate the data back into the form they were currently working with it in.
The problem had been originally presented to me as a SAS conversion problem, but actually the conversion ran the other way. This conversion was probably unworkable. Very expensive. They were looking at hiring all kinds of grad students to work on it. At least they realized it was not going to be workable.
So now they were thinking of going back to some kind of SAS solution, since they had the knowledge to do that on their own. What’s the big deal, I wondered. It’s a very complex SAS program. What does it do? He went into a long spiel. Oh, I said, so it’s running some queries? Yup.
Why the web? I asked. We wanted to have several people doing it at a time, he said. Were any of these people out there in the world? No.
No one, it seems, had ever questioned why the web. They had bought into the problem as presented.
Why not do it on the local network? Well, he said, those machines didn’t have enough power to handle it. Why? I don’t know what he was thinking, but it seemed that he had this idea that their software was complex and had tentacles everywhere, and not only that, it was inextricably integrated with the data.
Those machines don’t need power. They are just data entry ports. The server can hold your program and run it.
So all you need to do is set up a few data entry machines, hook them to your server, write a little software to make sure they could handle multiple data entry and be on their merry ways.
Very interesting proposal, he said. Let me run it by my tech guy. Funny. I asked my sister (a big deal web techy) about this, and she never came up with this.
Ah doctors. The cult of expertise. All these people with different hammers thinking everything looks like a nail.
I’d love to know who originally framed the problem… or rather, the solution. I’m not sure anyone even understands the problem since they were trying to jump through all kinds of hoops.
Now maybe there’s something I don’t understand about the software or the problem, but this just seemed too easy to me. And then I thought about group think. And how it is possible that this could come about as a consequence of noone questioning the ideas of the person before them because they were experts running multi-million dollar research projects.
It blows me away. It can’t be that simple, can it? How could all these terribly educated people invent this huge deal around it? Was it a make-work conspiracy? Was I collapsing that house of cards? Would everyone hate me for taking away jobs? Or at least, jobs in their particular department?
I must be wrong. The alternative is too scary. Or am I being influenced by group think?