When I looked at taking the Databricks Certified Developer for Apache Spark 3.0 certification last year, the FAQ said that certificates do not expire, as they are tied to a specific Spark version. Now, the Academy FAQ states that certifications expire after two years.

I have the following questions:

1. When and where was this change announced?
2. What happens to people who already got certified last year? Do they keep their infinite certification term or do their certifications now expire after 2 years? Some people might have paid for the certification expecting it would be valid indefinitely, as described.
3. Why was this change introduced? If the rationale was that certifications are tied to a specific Spark version (since supposedly features do not change), what justifies now having to re-certify every two years for Apache Spark 3.0?

Thank you!