> Here is the current method for resolving an index prefix: > > Start with the prefix for the prefix index. > Fetch prefix-date. > Read the first line. > This is your prefix. > > Example: > Prefix: "myindex" > Date: 986967900 > Fetch "myindex-986967900" > First line is "eiold" > Prefix = "eiold". > > So this is the current system. It's standardized. It's simple. Does the > DSK system have any benefits over this system?
Please elaborate. Can I currently use SSKs for the file that holds the prefix? Can I as the person creating the index specify a particular update interval for the file containing the prefix, so that clients are able to find it? If that's the case then the only advantage DSKs offer is the fact that inserting data looks a lot more like inserting a key. ============================================================= = Adrian Lopez <adrian2 at caribe.net> =========================
