Jesse Norell wrote:
But I am still not sure how much the performance gain is by
putting it into separate tables compared to searching on the
message blocks.
Consider a hypothetical user who passes 2-5MB documents back and
forth between colleagues all day long, and for whatever reason (he
is part of management?) saves a copy of all mail sent and received.
(He must be using a native webmail program, it looks like :).
If he wants to search for a Subject including "Jan-Promo.doc" in
all his messages, there's going to be a huge difference in the two
setups.
If I understand everything correct, the headers goes into the
first messageblock and the rest of the message goes into the next
message blocks.
With a flag on the message block containing marking it as a
header block will we only search the header blocks and these are
usually not that many kilobytes.
But it remains to be implemented.
My $0.02 is that we should make a simple solution first, that
maybe isn't that effective, but mostly good enough. We can then
refine it and add separate tables.
Magnus
