Re: RFC: simple proposal for Internet-scoped IDs

Johan Corveleyn <jcorvel_at_gmail.com>:
> Just to be clear, this entire proposal / discussion is about a feature
> that only makes sense for public repositories, forges, open source
> projects, etc, ... right?

That's the use case in mind, but see below.

> AFAIU, there is no need for something like
> this in corporate repositories, where you normally have a closed set
> of users, with centrally managed accounts (usually in a central
> directory), where SVN is just one of several IT services based on this
> same user account.

Probably not. Though it wouldn't surprise me a lot if, once this
capability is implemented, people in that world find uses for it.
Easy repository mobility, in particular, would have uses even if none
of the other global-addressing properties mattered.

> So I guess what I'm saying is: whatever comes out of this shouldn't in
> any way hinder the usage in corporate environments.

Nothing in my proposal changes present behavior unless a repository
administrator deliberately throws a switch.

> Also (but I guess that has already been said): if it ends up being
> some client-side configuration, it should obviously be settable per
> repository (so I can work with some forge with a certain attribution
> id, and with the corporate repository without one, all from the same
> pc).

Agreed.

-- 
		Eric S. Raymond