Published on: 2018-09-05T03:00:39+00:00
Gregory Maxwell, a developer at Bitcoin core, has been working on increasing the size of blockindex objects in memory to accommodate signed testnet mode. However, this approach did not prove successful as it increased the size of objects even when not in signed testnet mode. An alternative solution proposed by jtimon involves turning one of the fields, such as the merkel root, into a union with a pointer to a look-aside block index used only in signed block testnet mode. A new implementation is currently being worked on, which includes a global mapping of block hash to signature that is (de)serialized in the block header. The code for this implementation can be found on GitHub.Ryan, a contributor to the bitcoin-dev mailing list, suggested implementing small blocks on testnet to simulate a fee-market-like situation. However, there are concerns about how long the current situation of testnet spam and full blocks will last. Gregory Maxwell proposed creating a signed blocks testnet with a structured reorg pattern as an alternative to the current testnet. This would serve as a middle ground between regtest and a full testnet, providing useful testing opportunities without being subject to the network's whims. Concerns were raised about the increased memory usage due to the larger size of blockindex objects, but this could potentially be addressed by turning one of the fields into a union with a pointer to a look-aside block index.The idea of a signed blocks testnet with a predictable structured reorg pattern was proposed as an alternative to having two testnets with different block sizes. The current testnet is considered too stable when some instability is desired, and too unstable when stability is needed. A signed blocks testnet would provide a middle ground between regtest and a full testnet, allowing for more controlled testing opportunities. Although patches were previously developed for such a testnet in Bitcoin core, concerns about increased memory usage have been raised. However, turning one of the fields into a union with a pointer to a look-aside block index used only in signed block testnet mode could potentially address this issue.On the bitcoin-dev mailing list, a user named shiva sitamraju asked about a potential reset for testnet in the next release due to its length and long sync time. Peter Todd disagreed with this suggestion and advocated for testnet to be a larger blockchain than mainnet to identify size-related issues first. He also suggested using regtest as an alternative for testing, as private regtest blockchains offer better control over block creation. There was a proposal for having two testnets simultaneously, with one having a smaller block size for faster testing. However, Peter Todd favored the idea of a larger testnet. While a public testnet remains useful for referencing transactions in articles, there are various alternatives available for testing purposes, including regtest and private blockchains.In response to a user's question about the lack of multiple testnets, Peter Todd suggested having a larger blockchain for testing instead of resetting the current testnet. He also recommended using regtest as an alternative, as it allows for easy setup and control over block creation. The user mentioned that the current testnet is over 1.4 million blocks long and takes at least 48 hours to fully sync.In an email to the Bitcoin development mailing list, a user raised concerns about the increasing length of the Testnet, which now stands at 1,411,795 blocks, and the significant time required for a full sync. They inquired about the possibility of a reset in the next release or any reasons not to do so. Another member of the mailing list argued in favor of keeping Testnet as a larger blockchain than mainnet to identify size-related issues early on. They also suggested using regtest as a more suitable alternative for testing, as it offers easy setup and better control over block creation. The email was signed by Shiva S, the CEO of Blockonomics, a company specializing in decentralized and permissionless payments.The sender of the message is seeking information regarding the potential for a testnet reset in the next release or any reasons against it. They highlight that the current testnet's length and sync time are causing delays and increased disk overheads for testing purposes. The email is signed by Shiva S, the CEO of Blockonomics, and includes links to their Telegram group and Welcome Guide for further engagement.
Updated on: 2023-08-01T23:45:31.105467+00:00