well most of the time if you have HTML code that you past in article body and it has link to some pdf file (searchabel) and its located anywhere on the same server where KB is hosted, program must index that file contentes too why only attachments and body contents are indexed what if we have many linked files in HTML codes comming from E drive since C was small for that... and you dont wants to attach them since its a burden and extra floating copy...

expample code is -
<P><IFRAME name=inlineframe src="http://kb.*********.com/PDFs/CSI Weekly Procedures 0001.pdf" frameBorder=0 width="100%" height=600></IFRAME></P>

where i need "CSI Weekly Procedures 0001.pdf" this file should be indexed located on diff drive on same server or any network location.