> >I am still chuckling about how this thread has morphed from >someone trying >to learn the nuts and bolts of exposure more or less being >told, by some >experienced photographers who know this stuff well enough that >they have >forgotten how well they know it, that it doesn't matter, into >a thread about >4x4 vehicles where more or less the same mindset is at play. >I suppose it's the teacher in me that finds it rather sad. >
That's not unusual - we like abstractions to simplify the world for us. 4x4 for wheels, computer-aided matrix exposure meters for cameras they are all abstractions designed to simplify the task and lessen the skills required to do something. But what a lot (I'd even say most) people fail to realize that there's something called the "Law of Leaky Abstractions" (no, I didn't made it up <g>) - at some point the simplifications stop working and you're faced with raw complexity of real world. Like with 4x4 people often fail to realize (and marketing carefully fails to remind them) that when road gets wet and slippery enough and you're cornering fast enough (protected from the slippery road by 4x4 drive abstraction) you're about to experience a unique feeling of all four wheels loosing road contact at the same instance (instead of just two, as happens with 4x2). The same goes for auto-exposure - the first abstractions (CW metering) were simple enough to leak when facing backlit scene. Today's are more complex, but still the will fail under certain circumstances. I'm not against abstractions. They are fine - they let us do things we would not be able to do without them (as a side effect they do produce world where the ratio between the people doing things and people who actually know how to do things is rapidly falling towards zero) and is quite alright to use them. What is happening and was being expressed in this thread, though, is that they work that well that people start to believe they'll always work. Being a simplified model of the real world they inherently can't - otherwise they'd become as complex as the real world they model and we'd need new simpler model to be able use our initial model (or something like that <g>) Leon PS: Another example: people are putting so much faith in the might of spelling checkers that protect them from the incredibly stupid rules of English spelling that I've often seen word "asses" being used instead of the intended "assess".