[OAI-implementers] OAI identifier resolver

Adam Farquhar adam.farquhar@alumni.utexas.net
Mon, 20 Oct 2003 12:03:26 -0500

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
  <meta http-equiv="Content-Type" content="text/html;charset=ISO-8859-1">
<body text="#000000" bgcolor="#ffffff">
Hash algorithms such as MD5 or CRC32 cannot be used to generate unique
identifiers.&nbsp; These algorithms will occasionally produce the same
output for different input strings (this is why hash tables require a
mechanism for dealing with collisions).&nbsp; Common approaches to
generating unique identifiers use some sort of a registration mechanism
to appropriately partition the space of possible values.&nbsp; Successful
ones will leverage an existing registration mechanism, such as DNS.<br>
Adam Farquhar.<br>
Xiaoming Liu wrote:<br>
<blockquote type="cite"
  <pre wrap="">On Mon, 20 Oct 2003, Young,Jeff wrote:
  <blockquote type="cite">
    <pre wrap="">- My hope is that these URLs will be as natural-looking as possible, which
is why I'm advocating the assignment of meaningful repositoryIdentifiers
during the registration process, even for repositories that don't use the
oai-identifier schema.
  <pre wrap=""><!---->
I think we all agree it's useful to uniquely identify a repository and all
its records in an URL-friendly way. Thus different service providers and
data providers can easily interoperate.

I just personally feel it's probably easier to agree on an algorithm than
a centralized registration mechanism. MD5 generated fingerprint is probably
too long, but other hashing algorithms (like CRC32) can generate much
shorter signature.