Keep in mind while its pretty cool to keep images and binary data in a database these solutions may be prone to scalability issues

lol, this statement made me laugh quite a bit for one fact: does perlmonks not keep user images in the database? Each user image appears to have it's own node_id, so is this the case? Or does the node_id in this case just point to a file that the perlmonks code slurps in and outputs?