On Saturday, October 13, Geth and Parity clients brought the Constantinople hard fork to the Ropsten testnet. At first, things did not go as planned. Parity and Geth were unable to come to consensus, which led to a three-way fork in the testnet.
By Sunday, October 14, a handful of core developers, including Alexey Akhunov, Martin Holst Swende, and Wei Tang, identified the issue and fixed it. As far as we know, the issue affected Parity Ethereum and Aleth but not Geth. It is unclear if the issue affected other clients.
As of press, the Ropsten testnet is still forked three ways, but Parity and Geth nodes are now adding blocks to the same chain. The chain that had been mined by Parity Ethereum nodes affected by the consensus issue still exists, but according to commentary between Afri Schoedon and Nick Savers on the core developers' Gitter channel, the chain Geth and Parity Ethereum are both contributing to is the canonical chain:
Testnet bugs are to be expected; that's the point of the testnet. That's not to say they don't lead to community concern and speculation. In Friday's Core Devs Meeting, it was decided that if any major issues were discovered on Ropsten that Constantinople would be pushed back until January, and so naturally this leads to speculation about whether the issue was significant enough to cause a delay in the mainnet rollout.
The good news is that it didn't take too long for developers to identify and address the problem, even when the stakes were very low. The co-CEO of Absence, Abdullah Rangoonwala, piped up on the Gitter channel to understand the issue and ostensibly seek reassurance that this won't happen on the mainchain. Akhunov provided some comforting words:
But quick-moving, hardworking, proficient developers is a less interesting explanation than FUD or conspiracy theories, so here's a snack, trolls:
Funny coincidence or presidential overreach? Impossible to say.