Leadership · Retrospective
Maintaining a public chain after its original developers stopped

Trivechain's main network has been producing a block roughly every minute since its genesis on 1 March 2018. The developers who first built it stopped working on it in 2019, and I took over maintaining the protocol. I had joined after launch to build the mining pool, and my work grew from there into leading development. Maintaining a chain someone else designed is a different job from building one, and this month I have been thinking about what it asks of me.
Who still depends on it
A chain does not stop mattering when its developers move on. These people still rely on Trivechain working:
- Masternode operators, who lock collateral, run nodes, and vote on proposals funded by the governance treasury.
- Miners and pools, including the pool I built.
- Exchanges that list TRVC, three of them named in our README, which need deposits and withdrawals to keep working.
- Holders and wallet users, most of whom never think about the protocol at all, and should not have to.
- Certificate issuers and recipients. At the TriveAcademy workshop with the Malaysia Blockchain Association last December, certificates were issued on Trivechain. Checking them depends on the chain and its explorers continuing to run.
None of them signed up for a change of maintainer. They depend on the network either way, and many of them would not know whom to ask if it stopped.
What I think a maintainer owes them
Code people can read and run. In March I open-sourced the TriveAsset protocol and our HD wallet library. In April I published our adaptation of the Dash RPC client under the Trivechain organisation and added tests to the hashing libraries for our proof-of-work algorithm. Opening the code was partly about sharing the load. A protocol that depends on one maintainer has a single point of failure, and at the moment that maintainer is me. Public code with tests means someone else can find a bug, propose a fix and check it without waiting for me.
Releases that do not surprise people. Pinned dependencies, so that a build next month uses the same code as a build today; in April I switched the TriveAsset protocol's builds to install exactly what its lock file records. Small fixes treated as real work: in April I also corrected a check in the core node that raised an incorrect warning about a version bit. Version bits are how nodes tell their operators that a rule change they have not installed may be activating. A false alarm in that place teaches operators to ignore the real one.
Status that matches reality. The core repository's README is where many people look first. In March I updated it to mark which explorer and which wallet are actively developed, and to bring the list of mining pools up to date. In May I updated the supply figures that exchanges ask for. Our Medium publication, by contrast, has had nothing new since December 2019. That is a gap, and I would rather say so than let the silence speak for us. A short, accurate note is worth more than a long hopeful one.
What I cannot promise
I cannot promise new features on a schedule, and I would rather not pretend to. What I can offer is narrower: changes that are visible in public, tested where they can be, and explained to the people who have to act on them before they need to act. That is less exciting than a roadmap. For a chain whose users are mostly operators and holders rather than developers, I think it is the more useful commitment.
There is also something a maintainer should settle in quiet weeks: who exactly to contact, at every exchange, pool and large node operator that relies on the chain, if something goes wrong. That list is part of the job too.
Most weeks, maintenance is a dependency update, a README correction or a fix nobody notices. It makes no announcements. I think that is what the people relying on this chain need from me.