Crown development meeting minutes 24/02/2020

Just the facts, ma’am

The longest dev/strategy meeting in recent memory

Present: Alexander, artem, Crownfan, crowncoin-knight, pjcltd, walkjivefly

1. Artem spent the weekend working on the registration issues opened by walkjivefly. Since he couldn’t connect to testnet (for reasons unknown) he worked in his sandbox. He has reproduced some of the issues, identified the problems and will try to push a new build tonight.
   He thinks the testnet instability problems are probably not related to the NFT framework.
2. Ashot was not present but sent a message via crowncoin-knight saying he would soon resume work on the codebase update.
3. * Testnet stability has improved since the last rewind and blocking of “unknown” nodes. Defunctec did some analysis which suggests the much increased superblock frequency in testnet could be a contributory factor to the instability, probably more so than the old version nodes which connect sometimes.
   * walkjivefly asked if Artem could push a new build which reduces the testnet superblock frequency to the same as mainnet and refuses connections from old protocol nodes. If such a build proved more stable it would give us confidence that Emerald should not suffer the instability problems seen in testnet.
   * He also asked if having chosen an activation block for disconnecting old nodes, mainnet could refuse NFT transactions before that block is reached in order to prevent any invalid transactions sneaking into the chain and perhaps causing issues later. Artem prefers to enable the NFT functionality with a spork because a skilled attacker could fake an NFT transaction without using the nftproto or nftoken RPCs.
   * pjcltd has a pool of 20 additional virtual machines ready to start as soon as he’s given the build number to use. This will increase the size from about a dozen nodes to just over 30 and should help improve the stability for the remainder of the testing period.
   * Alexander and pjcltd are working on including new testnet DNS seednodes in the next release which will further improve the testnet stability and ease of adding new nodes.
   * We may create a completely new testnet (with a whole new datadir) for the codebase update.
4. walkjivefly asked how confident are we of being technically ready for a very provisional timetable of Emerald release on 2 March, NFT activation and protocol enforcing on 16 March. Artem wants to address all of the known NFT issues before release. Given our resource availability situation a more achievable plan is to aim for a 9 March release date and protocol enforcing on 23 March.
5. Crownfan explained the Emerald marketing campaign is in development and close to being ready. It will be deployed in stages with each one building on the previous. We need to be certain that the code is ready to go at the advertised block height/date before launching the campaign. Once launched much of the message will take on a life of it’s own due to organic/viral growth and cannot be altered or stopped. If the cutover doesn’t happen on the announced date the campaign will be much less effective or even wasted.
6. We have put out feelers to exchange partners about the update. We need to give them two weeks notice of the protocol enforcing cutover point and provide the repository link to the release candidate code in Gitlab.
7. After lengthy discussion of all the factors we have settled on a provisional timetable of Emerald mainnet release on or around 9 March, protocol enforcing and NFT functionality on 23 March. This gives us a bit over a week to resolve the necessary stability and NFT issues, two full weekends for people to upgrade their nodes, sufficient notice period for our exchange partners, and a short window for contributors to rollout the first NFT use-cases, including Crowncards, before Coinfest 2020 in Manchester on 3–5 April.