parsing DOI does not always work

[pliz]

Here is what I get on some of the valid DOIs, which parser finds in the pdf automatically, so it is not a typo that causing the error:

Exception: Error on line 1 char 1: Document must begin with an element (e.g. <book>)
The operation underway was: "Parsing CrossRef XML. The DOI could be invalid, or not known to crossref.org"

Example papers that cannot be parsed:
http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6WNP-4HV74JW-1&_user=1550512&_coverDate=04%2F01%2F2006&_rdoc=1&_fmt=&_orig=search&_sort=d&view=c&_acct=C000053660&_version=1&_urlVersion=0&_userid=1550512&md5=e6177d02d58c7451a70ea2c0957661d3

and

doi:10.1016/j.jcp.2006.10.010

[pliz (pliz)]

BTW crossref.org parses the doi above correctly.

[John S (jcspray)]

Confirmed: crossref is now giving the error " Requests without user credentials are no longer accepted.<br> Please sign up for a free openUrl account to use the free service."

[Jeff Fortin Tam (kiddo)]

Did you contact the crossref guys about this?
I saw this page http://www.crossref.org/openurl_info.html
Also, the subscription page says it is only to monitor usage and troubleshoot problems (http://www.crossref.org/requestaccount)

How will this be solved? Having Referencer request a username + password saved in the gnome keyring?

[John S (jcspray)]

Workaround is to set this url in the preferences dialog. http://www.crossref.org/openurl/?pid=ourl_myUserName:myPassWord&id=doi:<!DOI!>&noredirect=true

As for how it is dealt with long term, we shall see. Patches welcome.

[Michael Banck (mbanck)]

Isn't this bug fixed in 1.1.0, by adding the login/password dialog for CrossRef, along with a link on how to register (btw, that link should probably be transformed to a proper hyperlink users can click on)