Hacker News Reader

The docs list 5 conditions that all must be satisfied for the journal to be considered hot: https://www.sqlite.org/atomiccommit.html#_hot_rollback_journ...

I believe they would all be satisfied.

I don't see any mention of checking IDs. Not saying you're wrong - I think the docs could very well be wrong - but could you provide a citation for that behavior?

View on HN

Please read the entire document instead of just picking out sections; you will then be able to see where your misconceptions are occurring. You have attempted to make this same point three times, so I will say it for a third time; that section is about CORRUPTED databases, not database that are consistent after fsync.

The text in this document also directly contradicts what you're saying. Put another way: the presence of a hot journal is how SQLite determines the database might be corrupted.

https://sqlite.org/lockingv3.html#hot_journals

I read the whole document. It doesn't mention IDs anywhere.

If you're not going to provide citations for your claims, yet criticize me for "picking out sections" when I provide citations, then continuing this conversation won't be productive.

Reply on news.ycombinator.com