In software development it seems making any promise or prediction about the release date instantly jinxes the project. I've tried to accurately predict the release of DMX 4 a couple of times but it's a very slippery thing. With the number of third parties involved it gets incrementally more complex to make any sort of prediction. I use ComponentArt for my UI, Telerik for upload, and a partner's project for WebDAV in the new release. And you run into issues you would never have dreamt possible. And the third parties are no less surprised when disaster strikes.
Just an example. I use the upload component on a dynamically loaded panel in a wizard control. You still with me? OK, so this works. The upload control is shown and the upload progress bar shows the user the status of the transfer. It is exactly what I wanted. Almost. I changed the style of the buttons to correspond to the look and feel I was aiming for. A few tests later I noticed the progress bar was not showing up any longer. Great. Now what? After hours upon hours of stepping through recent changes, I finally nailed it down to the styling of the buttons of the wizard. Something completely and utterly unrelated! Needless to say I had quite a few hairs less by this time. One of those 'WTF' moments you encounter in development. Finding and fixing this has taken several days. Days I could otherwise have devoted to finishing the documentation.
Anyway, about the expectations. The past two weeks no issues have come out of beta tests that were real 'bugs'. It was mostly brushing up and making sure all is hooked up. I'm confident I'm inches away from release. The documentation is almost done. When this is complete I will have to do the promotional material (my site, Snowcovered, etc) as well as finializing licensing and pricing decisions. But as far as the technology is concerned I think it's ready for release. A few last test runs (upgrade and regular install) will need to be done once more to see if all packaging is correct, of course.