I've left room for the possibility of edits (this first release is marked inside as "Edition 1.0 (Nov 2012)", with the intention that typos will get point and date updates as needed).
However I opted to avoid the Smashwords meat-grinder and create my own epub (it is going to be automatically converted to mobi, but I've tested that process here using the Amazon Kindle Previewer, and it appears to come out fine). Since I chose to use BookBaby for distribution it means that updates are going to cost me (hopefully not too much).
Print on demand is more problematic. It definitely costs, and it's not as though I can afford to send those people that have already purchased a copy an updated edition. I did consider the option I've read about here (delaying print until e-books have been out for a while and the worst errors found), but e-readers are not so wide-spread among the people that I know here in Australia, so print is likely to be important for my word-of-mouth sales (as opposed to word-of-internet sales). Hence POD was a higher priority to me that perhaps for some others here.
I would also add that the idea that "I can fix it later" is not really one that I think self-publishers should adopt as an easy way out - it won't do much for our reputation. Yes there are probably going to be some errors, and yes it's great we can fix them later, but it is still best if the first editions are of as high quality as we can manage.
(I'm not saying that the previous posts were suggesting otherwise, but I am saying that we need to be careful how we state such things. Quality is one of those things always brought up in support of traditional publishing, it doesn't help if we are seen to be adopting a "let the (paying) readers fix it for us" approach to publishing.)
And, of course, having opened my mouth about my quality expectations, I am now feeling that much more pressure about whether I have done well enough by my own standards.
