Re: Store Specs With Trial?

[Peter Brawley]

In database design, feelings often yield mistakes.

Is what you call a spec a document, or some other kind of object? In either case, is it atomic for the tests you are modelling, or are parts of it subject to change? If the former, presumably some other process disassembles it for edits then reassembles it for use. If the latter, then a photo of it will introduce the likelihood of skew, so you probably want to have your test data contain a reference to where the spec exists rather than keep its own obsoletable copy of it.