The ajax request is being sent to html page hence you should use html as data type. Using xml might break in case there is un-escaped characters such as "<" or "&" in html. In fact it was causing problems when trying to retrieve bibtex data for particular item which included link with an ampersand.
It was mentioned in comments section in website by Tobi:
> Tobi: Hi Martin,
>
> it seems like bibtexbrowser cannot handle & (ampersands) in URLs. Entries that contain & confuse > the bibtex-block-script, which displays bibtex information in a new site instead of loading the > block of information just below the entry. I can't seem to find the bug, but hope that you can.
>
> Tobi
> (May 27, 2014 - 6:33 pm)
Also I have found some random page which is using bibtexbrowser and you can see the bug here http://www.imp.kiev.ua/~kord/papers/ by clicking on "Strong electron pairing at the iron 3dxz,yz orbitals in hole-doped BaFe2As2 superconductors revealed by angle-resolved photoemission spectroscopy" article's bib anchor.
