that one's not inherently fatal, in the sense that it shouldn't stop the search. On 16 January 2012 17:13, erik quanstrom wrote: > 2. we're ignoring status codes that we should be treating as fatal (like > Srvfail)