Originally posted by: doof moof
all the users will be accessing the same data. It will be a database of lots
of artist profiles wich users can then search and check out different artists.
Here is a site which is a similar model to the one i am creating:
www.purevolume.com I still don't know if I'd opt for using shared objects on this one, for a
couple of reasons:
1. I'd want to repurpose that data at some point. If it's all stored in rso's,
any application that wants to use that data will either have to make a
connection to FMS, or decypher the rso data to reuse it.
2. I'd probably want to store images along with the data, and you can't store
binary blobs in an rso (not in a useable format anyway)
3. I'd have to build a separate admin tool for managing the rso's.
4. I'd be putting additional load on the FMS server, and I'd really prefer to
use those resources for critical functions rather than wasting it on pushing
data my http or remoting server would handle more efficiently.
5. The users don't need to share the data in real time. Although they are
using the same data, they don't need to worry about changes to that data in
real time.
What I might do is have a couple of rso's on the server that contain things
like new additions and lists of most popular records. I would then have FMS
run a query on the database (via remoting or LoadVars) every so often to update
those lists. The idea here is that you want to use rso's for data that might
change while the user is connected, and use your http or remoting server for
data that is just sitting in a database.