Should you switch to Plone 3 or not?
A lot of people are still on Plone 2.5, and the question for these is if they should move to Plone 3 or not, and in that case, when. I’ll give my point of view here.
Plone 3 contains loads of new things and features. These may or may not be of use to you. The main new feature for me has been the content rules, which are very cool and handy. But Plone 3 also have refactored some features, making them better, but also often more complex. The portlets are a good example of that. They used to be an ugly hack, where you for some bizarre reason created portlets as macros, and now they are neat and organized, but hard to understand.
There is also many new recommended techniques, like buildout and generic setup. But they are available for Plone 2.5 as well.
Are you mainly using Plone 2.5? Are you happy with that, and know it well? Should you switch to Plone 3, and then when?
Well, my answer to that is: Switch to Plone 3 when you need to. Do you need content rules? Switch. Is there a product you need that only runs on Plone 3? Switch. But as long as you can happily put together sites on Plone 2.5, then stay on Plone 2.5 for the time being. There will surely be a time where the switch is beneficial to everyone, but that time is not now.
However, even if you are on Plone 2.5, learn buildout and generic setup now, and start using them. Not only will that help you when you do witch to Plone 3, it will help you now. And start using the “best practices”. Do the development on the hard disk, not through the ZMI. Put methods that are unique for a content type directly on the class, not in a skin. Version everything with svn and tag releases.
That’s my €0.02 on the issue. Your Mileage May Vary.