Orchard could be describing word for word my own need: "Once I’ve acquired feed data, I need to store it in some form usable by my other programs and by some method as agnostic as possible toward the actual contents of the feed." The rest of the post analyzes four approaches: fine-grained relational DB tables, triples in an RDF store, XML database storage and coarse-grained persistence. None of these does everything; each has weaknesses. I'm pretty much in the same situation, complete with code littering my website (only in Perl instead of Python). What I won't do is buy something off the shelf or start using a packaged designed by someone else. If I can't build it, I won't understand it, and if I don't understand it I'm not very useful as a pundit and critic.CRLF
Today: 12 Total: 121 [Direct link] [Share]
