Henning Koch emailed with some comments on URL Escrow. One point upfront, I’m wary of free services. I’m always looking/waiting for the catch, the plea for money or the closure from lack of funds. I think URL Escrow should be a funded service. Service providers trust that there will be no abuse of domain name access and clients trust that the escrow service will be there when it’s needed. I think that level of trust requires solid financials that are rare in a free service.
As I envision the service, clients/providers would pay an annual fee for URL Escrow. The client/provider would register a URL with the escrow service. And if the provider goes dark, then the escrow service would step in to provide a temporary redirection service. There might be two tiers of service: redirection to a single URL and extended path_info redirection (old_base/path_info → new_base/path_info). After a few months, control of the domain name would be returned to the service provider and redirection would end.
Paying clients would have an existing account where they could enter their new target url. Paying providers would maintain a list of clients at the escrow provider. If the service went dark, then the escrow provider would contact the clients and establish accounts for them to enter a new target url. Add some software magic to translate the database to redirection instructions and it’s good to go.
The financials for the escrow service are a bit tricky. It is similar to insurance, a fallback position in case things go badly with a lot riding on correct estimates of the frequency of failure. I doubt that the returns justify the capital investment in new infrastructure. But an existing hosting provider with a solid reputation could gain an additional revenue stream from URL Escrow with a minimal investment.