Andreas Schildbach, one of the developers of the popular Bitcoin Wallet app for Android and maintainer of alternative Bitcoin implementation Bitcoinj , calls for a block-size increase to allow for more Bitcoin transactions. Rather than relying on additional layers built on top of the blockchain, Schildbach believes the network should be designed to scale “on-chain”, too.
“I personally will not invest my time in lightning networks, at least not for solving the Bitcoin scalability issue,” Schildbach told Bitcoin Magazine .
The block-size dispute – which represents a trade-off between the number of transactions the Bitcoin network can handle and its decentralization – has taken a bit of a turn since a large segment of Bitcoin's academic and engineering community discussed the issue at the upcoming Scaling Bitcoin workshop in Hong Kong on the 6th and 7th of this month. Core developer and Blockstream co-founder Pieter Wuille’s Segregated Witness , in particular, is embraced by many within Bitcoin’s development community. As part of a scalability roadmap that needs no hard fork, the proposal could help allow added scaling layers such as the Lightning Network to be rolled out.
Schildbach, who maintains Bitcoinj alongside Mike Hearn (who’s also lead developer of Bitcoin XT) believes raising the block-size limit through a hard fork needs to happen sooner rather than later, however.
“The Segregated Witness proposal is very interesting, and could provide a valuable addition to the Bitcoin protocol once finished,” Schildbach said. “But blocks are already full. If we want everyone to use Bitcoin, then we need to make room for them. I believe this matter is urgent, given the time it needs to properly prepare for a hard fork. Bitcoin XT and Bitcoin Core developer Gavin Andresen has done extensive tests that show that Bitcoin scales linearly up to 20 MB blocks, and improved block compression solutions such as 'thin blocks ' will decrease redundancy in block propagation, limiting centralization risks.”
While most within the Bitcoin development community believe the block size needs to be increased at some point even with Segregated Witness on the table, opinions still differ on when and how to do this, exactly.
Andresen's proposed solution, BIP 101 , was implemented by Hearn into Bitcoin XT, and is programmed to increase the maximum block size to 8 megabytes if a threshold of 75 percent of mining power accepts the change. Once activated, this limit is set to double every two years for 20 years, ultimately leading to an 8-gigabyte block-size limit.
An alternative idea is Blockstream CEO and hashcash inventor Dr. Adam Back's “2-4-8” quick fix, which would incrementally increase to the limit 8 megabytes over four years.
Similarly, Bitcoin Core developer Jeff Garzik’s BIP 102 (or his similar “BIP 202”), proposes a bump to 2 megabytes. Combines with Segregated Witness, this could effectively provide for up to 4 to 8 megabyte blocks, too.
Schildbach, who maintains Bitcoinj alongside Hearn, indicates to be willing to accept almost any solution the Bitcoin community decides on.
“I support any proposal that increases the block size limit, as long as it doesn't introduce pitfalls like the ability to decrease it,” he said. “If the community agrees on a fixed limit of just 2 megabyte, then so be it. It would likely lead to the same discussion in one or two years, but hopefully then the community will have learned from the pains of the 1 megabyte limit.”
Schildbach is sceptical about any solution that does not involve a block-size increase, however. According to the developer, whose Bitcoin Wallet has been downloaded more than a million times, additional layers built on top of Bitcoin designed to scale the system would make things needlessly complicated – and take too long to develop.
“The Lightning Network and similar proposals will cause a complexity explosion on Bitcoin as a whole,” Schildbach said. “It will take quite a while to implement, test, fix and roll out the protocol. I'm talking of years, rather than months. I'm aware Blockstream is working on the core implementation. But work on necessary infrastructure, like wallets and the payment protocol hasn't even started, as far as I know.”