Chris made a fairly ascerbic (and fair) comment when I complained about the disappearance of the digital identity of one of my papers. Amongst other things, he found the current(!) doi for it …

It’s still possible I’ve got the doi wrong, but given the previous doi resolves and the subsequent one as well, it suggests it’s just broken. This next example is instructive, same pattern DOI, same vintage, nearly works, yet things are not as they should be:

10.1256/smsqj.52407 resolves to the right paper, but if you look for the doi on that page, you see the doi reported as 10.1002/qj.49712152408.

Right next to that last doi is a link to what Wiley thinks about Dois…

... there is the added problem that ownership of information changes, and location of electronic files changes frequently over the life of a work. Technology is needed that permits an identifier to remain persistent although the links to rights holders may vary with time and place ...

In this case the publishers (Wiley) seem to think that the first part of the doi (before the slash) has to point to them in perpetuity, rather than understanding it’s only purpose is to give them the authority to ensure that at assignment, they can assign the right hand part uniquely. Once they’ve done that, the publisher can move with impunity … which is the whole point of the system.

Wiley is never going to assign anything in the 10.1256 space, but it has inherited lot. Dear Mr Wiley, YOU DO NOT NEED TO REASSIGN, you’re just screwing with the system.

Dear Met Society, you should instruct your publisher to behave!

And now back to my paper … which seems to have disappeared into the long grass during the reassignment (which sounds rather Orwellian). I guess it does have a new DOI, but on the evidence thus far, I might as well not bother with it either …

comments (1)

Dom (on Tuesday 09 June, 2009)

You might be interested in comment 8 in this article (after pausing to admire the beard):

http://network.nature.com/people/mfenner/blog/2009/02/17/interview-with-geoffrey-bilder