Paul J Stevens <[EMAIL PROTECTED]> said: > > Jesse Norell wrote: > >> > >>This can be very interesting to do. Of course, we still need to store > >>the message in its original format, but to store some information on > >>mime-parts would be very benificial for performance. > > > > This would/could be another application for making a more generic > > per-message data cache, rather than solely for message headers. > > How about some serialization of meta-data like python's shelve or php's > type-serializer and introduce a cache-field to store results from > introspective queries. >
What does this mean in English? Presumably serialized data could no longer be plain-text searched. That missed seem to miss most of the point of the cache fields, which is that we want to have the data directly in the database as messageid/key/value tuples (possibly with joined tables, but that's another debate -- although joins won, iirc, and I recanted my claims that joins would be slower than overstretched indices). If instead we have messageid/serialized-data pairs, then we'd have to parse the serialized data and search it from within dbmail. By using key/value pairs, we can let the database do the searching... which is what databases do! Aaron --
