On 9/16/07, Danek Duvall <danek.duvall at sun.com> wrote: > On Sun, Sep 16, 2007 at 07:34:30PM +0100, Peter Tribble wrote: > > How does this work without a repository? > > Without a repository? For the moment, we assume one, even if it's local. > We'll probably have to move beyond that, but I don't think it's going to > prove to be a common occurrence.
Oh, I would assume that not having a repository would be the most common case. Most software suppliers wouldn't have one; most delivery mechanisms wouldn't support one. The repository model works fine if you're (a) talking about bundled software, and (b) you're prepared to trust the repository to make decisions for you. (In this I include the model where the customer can establish their own repository which can proxy or aggregate software according to local need, so they can determine what's included in the scope of the repository and what the policies are. And there would be the need to supply the capability for anybody to do that.) The problem arises if you don't do that. MegaCorp may have some heavily protected download system you have to go through that doesn't fit into a repository model; individuals building up packages to send to their friends shouldn't have to worry about the complexity. Distribution may be via ftp, e-mail attachments, disks on magazine covers, memory sticks given out at trade shows. And then many IT departments build their own deployment mechanisms that you have to submit to. So the repository is a nice optimization; it cannot be a critical component. -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/
