Lists Home |
Date Index |
A problematic use-case (if that's the right phrase) is accessing a, say,
500M corporate metadata repository. How many minutes should I be prepared
to wait while it's parsed before I can use it, and what's the point of
parsing it, anyway? The VM management issues in many people reading a file
that large are also problematical. Wotcha reckon?
Direct access is the go; I can't imagine anyone putting up with SQL if all
access was sequential, which seems to be the analogy here. Incidentally, I
can't think why people put up with an access method as grossly inefficient
as SQL anyway, but that's a whole other story. Thanks, Dr. Codd (not!)
Guess it comes down to user/responsiveness issues. Batch processing of xml
files; who cares how long it takes. Users waiting for search results; it
had better be quick!