Miner's approach seems way too exhaustive for small teams that are quickly iterating on a product. It's a bit old school. I used to do specs like this and they were extremely time-consuming and were never used as intended by the people who were supposed to reference them. That was when I was consulting though, so my time was their money. Now back on the product side, it's a time suck I can't afford.
There is a magic balance of documentation and design that needs to be available for posterity...and LayerVault's timeline is a good step towards this — may be allowing some way to indicate what was "shipped" in the timeline of a design? My current solution is a more visual/high level spec that explains the goals of each screen and has a semi-final design next to it, all in a Google Doc. I try to update it as we make decisions during product development so we at least have some history of decisions made on the ground.
Miner's approach seems way too exhaustive for small teams that are quickly iterating on a product. It's a bit old school. I used to do specs like this and they were extremely time-consuming and were never used as intended by the people who were supposed to reference them. That was when I was consulting though, so my time was their money. Now back on the product side, it's a time suck I can't afford.
There is a magic balance of documentation and design that needs to be available for posterity...and LayerVault's timeline is a good step towards this — may be allowing some way to indicate what was "shipped" in the timeline of a design? My current solution is a more visual/high level spec that explains the goals of each screen and has a semi-final design next to it, all in a Google Doc. I try to update it as we make decisions during product development so we at least have some history of decisions made on the ground.
But that, also, isn't perfect.