Sometime in the second week of October, all Constantinople-implemented clients will fork the Ropsten testnet. Martin Holst Swende suggested aiming for October 9. However, the fork is not set for a particular date or hour, but rather for a to-be-determined block number. The developers agreed to identify the exact block number in two weeks, but there's already a proposal up on GitHub suggesting the fork take place at block number 4,230,000.
As it stands, Aleth and Parity have implemented Constantinople, and the rest of the clients are not far behind.
During the test period, developers will watch for and address any bugs or problematic inconsistencies between clients. Leading up to last year's Byzantium hard fork, there was a lot of uncertainty and several last-minute fixes. Days before the launch on the mainnet, developers discovered a denial-of-service vulnerability in Geth's code. Parity was also up against the wire, issuing three updates in quick succession prior to the hard fork to address consensus vulnerabilities.
There is no official date set for the launch on the mainnet, which likely depends on what the developers find on Ropsten. However, Hudson Jameson of the Ethereum Foundation estimated that it will likely take place in November or December. In any case, it will not occur prior to DevCon4, which is scheduled for October 30 through November 2.
In previous coverage, ETHNews has discussed some of the EIPs to be included in Constantinople, and the debate surrounding those proposals. The last EIP to be accepted was EIP 1234, a proposal to reduce block issuance and delay the difficulty bomb. Some in the mining community were not excited about this, because their profits are already under threat by low Ether prices and powerful mining rigs called ASICs – another subject of debate within the community. Some in the mining community feel ASIC mining rigs (as opposed to GPU rigs) threaten their profitability. To address this concern, they are arguing for a change from the existing consensus algorithm to ProgPoW, which would apparently eliminate the issue.
The developers ultimately rejected the inclusion of this algorithm in Constantinople, but agreed to consider it for Istanbul, the hard fork tentatively planned to take place in mid-2019, eight months after Constantinople. In today's core dev call, ProgPoW was again discussed. However, as has been the trend, the developers were not overly keen on the idea.