RDAP fixes the main weaknesses of WHOIS. Answers are JSON with defined field names, so software can read them without guessing. Queries run over HTTPS, status values and event dates are standardised, errors use HTTP status codes (a 404 means no such object), and servers can offer different levels of detail to authenticated users.
Clients find the right server through bootstrap files that IANA publishes: one lists the RDAP base URL for each TLD, others do the same for IP address ranges and AS numbers. A client downloads the file, looks up the TLD and sends a request such as /domain/example.com to the listed server. The registry answer often links to the registrar's RDAP server for additional data.
All gTLD registries and registrars must provide RDAP. Country-code registries decide for themselves, so a number of ccTLDs are missing from the bootstrap file and can only be queried through classic WHOIS. As with WHOIS, personal contact data is usually redacted in public answers.
Example
GET https://rdap.registry.example/domain/example.com
{ "objectClassName": "domain", "ldhName": "example.com",
"status": ["client transfer prohibited"],
"events": [{ "eventAction": "registration", "eventDate": "1995-08-14T04:00:00Z" }] }