With Lion, autosaving is the norm and is very much encouraged, and those programs that don’t do it at the moment will no doubt start moving over to that model over the next year or two.
That’s by-the-by, though. Scrivener was designed to have autosave from the beginning - as its designer, it’s something I wanted. And over the past five years, the overwhelming feedback from our tens of thousands of users has been that they love the security of autosave and how it works. I understand that your proposal would still incorporate an auto-save of sorts, but changing the saving mechanism when we have so many users who rely on, and like, the current implementation would not be a good move for us at all. I can count on one hand (including this thread) how many users have disliked autosave over the years.
This is me saying that it won’t change and that I have no intention of altering the philosophy of the program. And most users do like it this way, and while that is not a reason not to “improve” it, I disagree that your proposal would be an improvement, either for myself or most of our users. Which isn’t to dismiss your proposal - it certainly has its merits and I can understand why you might like something like that, but I wouldn’t want to “lead you on” and give you the impression that this is something that might be implemented. I can certainly understand that a “Revert” feature would be useful - and in all fairness this is something that most auto-saving programs do have, which Scrivener does not. Ioa, however (who knows what he’s talking about), is right that the saving mechanism is very closely tied to everything else. Reverting to an earlier version of a single text is one thing; reverting the whole binder, when there may be texts created since which are now orphaned along with many other potential issues, is a problem on a whole different scale.
I do appreciate the feedback, though, and the thought you have put into it.
All the best,