May 15, 2024

The usability features are already saving me tons of time in testing and debugging, since I can often corrupt the store by halting it in the debugger or creating a bug, but then it just comes right back from snapshot in a minute or two. The runtime diagnostics for the tables are essential for sizing and bucketizing. And the dynamic log leveling and other info helps a lot. Also, with the addition of stepped parallel block download the times have become very consistent - within a minute or two difference between runs when syncing the entire chain with Electrum over the Internet. Reliability remains extraordinary. The only issues I ever see are ones that I introduce with new code.