It would be better if this package first used a local database lookup, and then only resorted to a whois lookup when the database lookup fails.
One instance of this package querying the RIPE whois server isn't so bad.
But multiple instances of this package all banging on the RIPE whois server is very bad indeed.
The current implementation suffers from:
- not being asynchronous; there is a performance penalty (response time)
- there are resource consumption issues (network bandwidth)
- there are scaleability issues
- there is the risk of being throttled or cut-off altogether by the RIPE whois server