As an experiment while writing a test, I played around with solr.xml and defined it this way: <solr/>
Starts up and runs just fine in core discovery mode since it defaults to core discovery mode in the absence of a <cores> tag. Of course all the defaults are used and you better have core.properties files in the right place and all that, but... So does it make sense to officially support a "no solr.xml" option? Removing solr.xml entirely barfs with "solr.xml does not exist in /Users/Erick/apache/trunk/solr/example/solr/solr.xml cannot start Solr" and an empty solr.xml results in an XML parsing error. I don't have strong feelings either way, but thought I'd throw it out for people to kick around. Worth a JIRA? Erick
