My point is that unlike any "open" identifier system if the DOI authority go in a direction we (or our decendants) don't like (become unreliable or ask for too much money) DOI actually own the identifies (they are trade marked). Setting up an alternative resolution service may not be legally permitted. So our data may be tagged with GUIDs we are legally not permitted to set up resolvers for (e.g. A parallel handle system perhaps). Compare this to something like DNS where, should the existing system disappear tomorrow, we could set up our own system with our own root servers for free because the standards are open.
Roger
On 26 Nov 2008, at 11:02, Gregor Hagedorn g.m.hagedorn@gmail.com wrote:
On Wed, Nov 26, 2008 at 12:06 AM, Roger Hyam rogerhyam@mac.com wrote:
If they stop working in 10 years do we get our €1,000 back or do w e get actual compensation for injury caused because we have come to rely on them and now we have to find some other mechanism? Do they have a clause where they can come to us and ask for more money if they need it? Boy they would be in a good position to do so once we totally rely on them!
I am not sure that is relevant, since the DOI or handle will remain a neutral and permanent identifier in the way Dave V has specified, which can be resolved in multiple ways. The services that can potentially go down are resolution, not existance of an identifier. Should dois fail, would anyone sane try to ruthlessly reuse the identifier space in which a few million "historical" digital objects have been recorded?
There are advantages of coupling it into the existing doi system for resolution, one of the advantages being that it is well known and much more likely to be picked up by NCBI, EMBL, and Journal editors (outside of biodiv itself) than a homegrown system. Other advantages have been described by the funcationalities Roderic described.
Your comments, Roger, are they not really referring to ONE resolution method, where there could be others? Is it dangerous should the whole of DOI go down (unlikely) and we need to step in in 10 years? Is a few 1000 EUR is not a bad price for ten years or service?
Maybe I misunderstand something.
Gregor