Like a lot of people, I spend quite a lot of my working day for one reason or another in Excel. And I've learned not to trust it. Not the software - it seems pretty reliable (if a little inflexible at times.)

I don't trust my own work - its too easy to make a mistake (mistyping a number or formula, putting the variables in an equation the wrong way around etc.) I double (or triple) check everything. But I have a reasonable idea how enthusiastic/bored I was when I was doing a particular piece of work, and how likely I was to have made a simple error at the time.

For other people's work, I'm less trusting. If someone has figures where I would expect to see a formula, I'll try to put together the formula to check that the figures are right. I'll check that percentages add up to 100% - basic checks that I probably wouldn't do on my own work.

But print… well, I trust that a bit more. Because the numbers on a page are what they are. Its quite literally black and white. Partly, I think this is because its easier to assume that, for example, the cells that should be formulas are actually formulas (I'm an optimist…) But also because there is a finality to print — if I'm saving a working file to a network drive or emailling someone a work in progress, I'm not going to double check it in the same way as if I'm printing a copy off.

So I was pretty surprised to see this story about Xerox photocopiers 'randomly' altering numbers that they were scanning or photocopying. I had assumed that a copy was just that — it had never occurred to me that in a digital age, those massive copiers would be running compression/decompression algorithms.

So now I don't know what to trust any more…

Comment