I think a bundle will be more convenient don’t you

> The license in each source file (i prefer just a license.txt file)

We need to copy license summary on each field for GPL to ensure that someone that find a file alone cannot say:
“Oh, I didn’t know that file was protected by GPL !”
I prefer the DSSL: Demerden Sie Sich Licence

> The billion tagged values (i have lost hours here)

ArchGenXML (and Genesis) want to make to more things from poor data. the best way to play with tagged values should be reverse-engineering on code that fill them automatically.

> The command line

You like buildout, don’t you? :)
The advantage of the command line is to not be englued with an editor or an IDE. But like buildout do, it should be better to generate only a skel from command line instead of reinvent the wheel.

> ArgoUML

Use Poseidon or Apollo for Eclipse ;^)

> All the hacks done every where to make the code compatible with two versions of Plone

Just wait for Plone 4 0:^)

> You can’t modify a line of code generated without lost it if you re-generate your code

Yes, this one is very important and it is link to tagged values and reverse engineering.

On workflows: I’m mainly agree with you. Keep in mind that DCWorkflow won’t live for ever and making simple task would make evolution easier.

On your template: You should add a doc folder to centralized documentation ;)

Please, before to write an advanced example show us how to install and configure this in eclipse and in a PyDev project.

I will be yours in Paris sprint.