in the preservation space you describe, the files would presumably be small
text files.  if that's the case it might be useful to ask, what would
keeping the metadata separate from the content files afford you?

On Mon, May 9, 2016 at 9:26 AM, Andrew Weidner <[log in to unmask]>
wrote:

> Thanks for your replies, Stuart and Brian.
>
> The information you provided got me to thinking more generally about what
> comprehensive preservation could look like for the digitized cultural
> heritage materials we are managing. As a content administrator for an
> access repository, I am primarily concerned with the ability to retrieve
> bitstreams and metadata from preservation cold storage in the event that we
> lose data and our intermediate backup systems fail. I've visualized one
> possible data restoration model in the following slides:
>
>
> https://docs.google.com/presentation/d/1x3YlQbQUitaRLH1qVVWYHZVonrffzoAKJTY905PpjK8/edit?usp=sharing
>
> I'm curious what the C4L collective brain thinks about such an approach,
> especially any pitfalls we should watch out for if we should choose to
> implement this model.
>
> Andy Weidner
>
>
>
> On Wed, Mar 23, 2016 at 8:59 AM, Brian Kennison <[log in to unmask]>
> wrote:
>
> > >>
> > >> How do others approach this problem? Are there recognized best
> > practices to
> > >> adhere to?
> > >>
> >
> > I’m still trying to put the CDL model into practice. <
> > https://confluence.ucop.edu/display/Curation/D-flat >
> > And Stanford has similar but different model <
> > http://journal.code4lib.org/articles/8482 >
> >
> > —Brian
> >
>