You are viewing a single comment's thread from:
RE: Performance and Transaction Billing on EOS: A Cooperative Effort, Not a Competitive One
Great write up.
Do you think greylisting should have been tested on a testnet before being done in production? I've heard it caused some problems that had to be fixed in the latest version. I've been wondering how many other things hidden inside of EOSIO should be considered unsafe for production until it's fully tested out. Thoughts?
Not exactly true : it exposed a new issue in regards to transactions hanging around too long. This wasn't an issue with greylisting -- it was an issue long before that -- but having greylisting exposed the issue more visibly.
This wasn't found in tests simply because nothing can really simulate the mainnet transaction volume and organization right now. Until we have a testnet that emulates mainnet exactly, it's going to be extremely hard to test how changes will effect live production code.
Regardless, the issue was promptly fixed.
Bugs happen all the time, we should seek to find them and fix them as soon as possible, not have anxiety upgrading and apply the bugfixes...