1. I’m building a base theme for which all my other child themes are going to based off of. For a client who isn’t going to be changing the theme, because they are up to spec, or they will only change it at their own risk or you are building a theme for a client based off of a theme that someone already built and customizing it.

2. I’m using a theme and I only need to make a few changes, but I don’t want to lose them when that theme updates, whether it is on the WordPress.org repository or it is a theme that has upgrade support built-in.

3. I’m making a lot of changes to a theme or I want to make changes to a child theme that is mostly what I want, but not everything.

I really think people should be able to think outside of the box. Child themes were really meant for the first two purposes. There is no point becoming crazy with this, because then you get to the point where it takes 2 minutes for WordPress to load because you have 20 themes built off of each other.

Think! Build. The structure of child themes should be small changes, if you are replacing most or all of the files, then you don’t have a child theme, you have a parent theme and you should move the rest of the files to your theme.

That is really supposed to be the point. WTF? are you building a child theme that adds functionality or replaces most of the theme? What are you trying to achieve? I’m all about torturing myself, because that is just the way I am, but I had no idea others were into that as well. How about when I cut myself, you cut yourself at the same time?

I guess I’m not exactly sure what the scope is. I’m not that interested in the mentioned themes. I could not use Carrington themes as a good example, because the last time I looked, it didn’t really play well with the whole child theme paradigm. Hell, they might have changed it, since the last time I looked. It would certainly raise the minimum WordPress requirements.

Hell, I tried to explain to the guy that his method for checking whether the file is being accessed is flawed and nearing paranoid (Why are you protecting a file that doesn’t even execute PHP?) Hell, I would have just checked for ABSPATH constant and exited out, but I l33t developers do shit the ass-backwards way. What do I know, I’ve only been developing PHP web applications for 9 years with 5 year emphasis on professional projects (which is often why I only say I have 5 year experience, I don’t count the 4 years I was paying my dues).

Well, it appears I went off on a tangent and I apologize. The point is, is that if you do major changes, then you should just be using the theme as a base and creating a whole new theme off of it. If you are only doing minor changes and tweaks, then it should be in a child theme.

If you build your child on top of Parent X, and then Parent X bails or fails you, the child can in no way can survive on it’s own (can it..?).

Depending on which theme framework you use, your options will vary. In the case of WP Framework, yes your child theme can and will survive. WP Framework contains all the standard functionality a theme should have out of the box. So if you choose to develop a theme based on it, you’re already guaranteed a long lasting WordPress theme from the get-go. If a scenario comes up where development for WP Framework stopped (as an example, this won’t ever happen), then all that means is you won’t be getting any additional features or compatibility updates if WordPress updates the themes API or deprecates current functions like get_header(); etc. (very slim chance of happening). But your theme will still work as should. It might not contain the new x feature that was introduced in WordPress 3.x, but it’ll work.

…Nor can it slot into a different Parent without modification.

As far as porting your theme to another framework (or parent theme), that’s certainly possible. The conversion might not be as straight forward as you’d imagine since all frameworks are different, but it’s certainly possible.

I’m certainly seeing it from a non-framework developer’s view. But I’m seeing more frameworks start up, and it’s a certainty that 1 will (eventually) close down (due to other work commitments or different project startups. NOT due to their workmanship).

I can see all the reasons for using a framework, but I can also see reasons not to.

1. What @Justin said. Basically if WordPress updates it’s theme API in 3.x and the framework (or any theme) uses a deprecated function that’s been deleted from WordPress, it’ll throw up a fatal error.
2. Depends on whether someone works on a theme framework in the Google Summer of Code program

2) I hope not. I’d rather see a lightweight, standards-compliant theme that’s easy to learn from.

I have 2 far more general questions ….

(This is the polar opposite of @Justin Tadlock’s question) ….
1) What happens when the theme framework ISN’T updated?
If I build off a framework, and then that framework is no longer updated, am I screwed..? In 2yrs time say WordPress is at version 3.something, there’s been considerable changes, and the framework needs to be brought up-to-date, where does that leave the user. Does framework leave you too vulnerable to longevity of that framework..?

2) Does anyone know/think that WordPress themselves will incorporate a design-agnostic, base framework..?

If, however, the thematic powered blog wasn’t dependent on thematic (since it would be bundled within as I described), she could easily customize it without any worries of having to upgrade her child theme.

Every WordPress theme is either a parent theme or a child theme, but not every parent theme is a theme framework. I guess it’s important to look at a theme framework as the backbone and not the xHTML.

Users might worry that they need to upgrade their child themes, which isn’t something they should be worried about doing. Child themes I release are starting points for users to build something custom with. Nearly all users are customizing their child theme to some extent. The thing is — they don’t have to update their child theme if I make a couple of CSS changes in an updated version.

Actually, I can see of several uses for that now that I think about it. But it’s about a million miles beside the point, really.