Tech 24 February 2018

Jasmine Solana

Too soon: More testing needed for OP_GROUP

Andrew Stone from Bitcoin Unlimited has proposed OP_GROUP as a method for implementing representative tokens – also called “colored coins” within the Bitcoin community.  But some members of the development community are urging caution and more time to decide whether OP_GROUP should be added to Bitcoin Cash software implementations.  Recently, Jonald Fyookball published a thoughtful piece explaining why OP_GROUP is potentially dangerous, and should only be activated after taking appropriate “time to test and analyze extensively” its impact. Fyookball’s post has generated significant favorable reaction from the developer community on Reddit.

May 2018 upgrade is too soon and Coingeek.com mining hash will not support this proposal

At first glance, the OP_GROUP proposal appeared to be a small change, given that the opcode is comprised of only 50 lines of code. However, OP_GROUP is capable of changing the Bitcoin design at a fundamental level. It alters how unspent transaction output (UTXO) sets are interpreted and makes token validation a function of miners, in stark contrast to other solutions where this validation is not part of the base bitcoin code.

For other op codes, the script runs only inside the script interpreter, acting like a firewalled black box.  But OP_GROUP changes the way transactions are interpreted outside the interpreter, and thus interacts across the boundary between the interpreter and the rest of Bitcoin.  Because OP_GROUP crosses that firewall boundary, it merits sufficient testing and analysis to ensure it is safe and does not change the very nature of Bitcoin in an undesirable manner.  There is not sufficient time before Bitcoin Cash’s May 2018 protocol upgrade to fully conduct reasonable testing and evaluation of OP_GROUP.

Furthermore, OP_GROUP could product negative impact on the Bitcoin Cash ecosystem.  The opcode would make the miners more responsible to ICOs, which could lead to possible legal issues globally with governments shutting down miners they believe are too involved in the ICO process, much like how exchanges have been continuously targeted in countries like China.

This is why the proposal needs to be examined more thoroughly, taking into account the effect it could have on miners, incentive models, and even other changes that could happen in the future. Pushing for a May 2018 implementation, sans tests and sufficient documentation, is too soon.  The proposal, after all, is a radical change for Bitcoin Cash and for these reasons the Coingeek.com mining hash will not support the OP_GROUP proposal at this time.

Note: Tokens in the SegWit chain are referred to as SegWit1X (BTC) and SegWit Gold (SWG) and are no longer Bitcoin. Bitcoin Cash (BCH) is the only true  Bitcoin as intended by the original Satoshi white paper.  Bitcoin BCH is the only public block chain that offers safe and cheap microtransactions.
Note: Tokens on the Bitcoin Core (segwit) Chain are Referred to as BTC coins. Bitcoin Satoshi Vision (BSV) is today the only Bitcoin implementation that follows Satoshi Nakamoto’s original whitepaper for Peer to Peer Electronic Cash. Bitcoin BSV is the only major public blockchain that maintains the original vision for Bitcoin as fast, frictionless, electronic cash.

COMMENT

Brett Williams

I agree. The size of the code change isn’t the problem, the problem is the size of the change to the model. The more I think about it, the more it seems to me that this OP_GROUP proposal would create a subtle change to the basic model. That’s not to say that it’s bad, just that it’s a very uncertain change that should be treated with extreme caution.

The basic task of miners is to validate scripts to see whether it’s valid to transfer BCH. This has a basic self-same harmonious sense to it, because BCH is also the credit the miners charge for that service. They’re concerned with whether the scripts validate because they care whether they get paid. It’s all of one piece.

In theory I guess you can attach other demands for computation and people can say that they refuse to pay miners unless that computation is done to standards, that doesn’t seem theoretically impossible. But it’s a different model. The miners are then validating those transactions as a service because it’s collectively demanded of them, not because it’s completely directly in the line of whether they’re paid. I’m not sure whether that works or not. We certainly can’t be sure by May.

Add a Comment

latest news

WWF monitoring food products via blockchain

Tech 4 hours ago

WWF monitoring food products via blockchain

World Wildlife Fund (WWF) Australia seeks to promote environmentally sound practices by the monitoring of food products with blockchain.

Read More
IBM blockchain to track mineral production and distribution

Tech 5 hours ago

IBM blockchain to track mineral production and distribution

Vancouver-based MineHub Technologies will be tracking mineral concentrates from production to delivery with the IBM Blockchain Platform.

Read More
Fake movie on Pirate Bay injects malicious code to steal crypto: report

Tech 16 January 2019

Fake movie on Pirate Bay injects malicious code to steal crypto: report

A fake movie file on The Pirate Bay torrent tracker is infecting computers and stealing cryptocurrency, in the latest example of cryptocurrency-based malware to sweep through online communities.

Read More