On Fri, Jun 27, 2003 at 10:48:30PM +0200, Paul de Vrieze wrote:
> - We need to have herds.

Agree. :)

>   There must be rather strong arguments against it to let go the choice of XML

XML is annoying but it is extensible and gets the job done. Thus I recommend
it.

XML encapsulates data in a structured way. It is not a good language to
encapsulate functional or procedural code. It does not have OOP features.
It is not fun to edit by hand.  For now, this is OK. We can deal with these
limitations.  Because it meets our needs, is easy to validate, and is easy
to integrate into Portage and other code (Web sites, databases, etc.) it is
very strongly recommended for use in our current tree and Portage 2.x.

> - We need ways to have people or herds only responsible for only a certain
>   versions of an ebuild. For that reason the restrict attribute is in the DTD

Agree.

> - The changelog. This probably will be a topic of debate.

Implement herds first. Hold off on ChangeLog stuff until we have excellent
tools to enter/edit ChangeLogs without seeing raw XML.

We need herds today. We don't absolutely need XML ChangeLogs today. Herds
implementation should not be held back on the account of XML ChangeLogs.
They can be added later when the ChangeLog editing tools are polished.

> - Long descriptions.

Agree.

> - Internationalisation.

Agree.

> - packages allways have herds, but may also have maintainers.

Don't care either way. Let's start using herds and we'll find out if we made
the right choice here.

> - Multiple herds per package.

Don't care either way. See above.

> Thanks for reading this overly long email.

Thanks for posting it :)

Once you collect this feedback, please post a suggested herds roll-out plan.
Good work :)

Best Regards,

-- 
Daniel Robbins
Chief Architect, Gentoo Linux
http://www.gentoo.org