If you wanted to take something fairly boring and make it sound like a gritty South African crime drama, you could do a lot worse than naming it Segregated Witness. However, once your target audience realises that the name is the most interesting part, and worse still, the plot is full of holes, your dream of bridging the bitcoin divides may be over before its begun. Harsh? Luckily for SegWit, it’s not up to me, and for the sake of a community in crisis, I’m willing – eager, even – to be proven wrong.
The case for the defence
I suppose I should start by saying that I admire Segwit’s spirit enormously. Nearly a year in development, very carefully considered (not carefully enough maybe, but I’ll come to that) and down-to-earth practical, it’s doing its best to offer something to everyone. The Ed Milliband of bitcoin upgrades if you will.
Proposed and presented by Pieter Wuille, co-founder of bitcoin heavyweights Blockstream, it’s safe to say SegWit arrived to a flurry of excitement among the bitcoin community. And at first glance, it’s easy to see why. SegWit offers a gentle way to resolve the block size debate by coming online as a soft fork. In other words, you can upgrade if you want, but failing to do so won’t exclude you from the network. Most importantly, whatever you do, the blockchain won’t split, and you have a whole year to voice your support or rejection of the project. If SegWit can garner approval from 95% or more of users, it will become the standard – if it doesn’t, it disappears.
Instead of tackling the block size issue head-on, SegWit takes a novel workaround approach. Again, at first glance, quite an elegant solution. In short, the transaction signature is removed from the ‘from’ field, and either into the coinbase or the input of a generated transaction. This causes the block size to appear far smaller to nodes, meaning they can double their capacity – or even quadruple it in some cases.
So far so good, and there’s more. SegWit also fixes the transaction malleability issue which has been hanging around causing problems for an embarrassingly long time. Cited as a major cause of the withdrawal problems leading to the collapse of MtGOX, this bug allows the unscrupulous to resign transactions, change Ids in the process, and make off with your bitcoins. I don’t think anyone’s against closing this loophole. So why all the gloom?
May the jury’s attention be brought back to the most fundamental question in this case: how to resolve the block size debate. Interesting and practical it may be, but SegWit does not solve this problem. If anything, it prolongs it. Maybe that’s what our community needs right now, more time to argue… SegWit could see us through another year or two or three, but at some point we all have to face up to the limits our current block size places on bitcoin’s viability. Releasing it as a soft fork may only compound this issue. It feels lacking in courage. Yes, a hard fork carries high risk. But so does an unresolved block size debate. Projects fork, and projects recover. Maybe bitcoin just needs to get over itself, trust that it’s a robust and pioneering ecosystem whose users won’t let it fail.
Trust – which brings us to our second piece of vital evidence. Moving the transaction signature to the coinbase might sound good in theory, but in practice it’s a different story. Miners regularly use the coinbase to send important business information about mining history and bitcoin improvement proposals (BIPs). Perhaps more importantly, it’s also the first part of the block to be written. Moving the signature into the coinbase means the whole flow of transaction processing is interrupted. Do we expect miners to just suck it up and re-write their code? If miners aren’t getting on board, SegWit is dead in the water. In fact, ViaBTC – a mining pool accounting for around 9% of network activity – has already voiced its rejection of SegWit, preferring instead the braver Bitcoin Unlimited fork.
The jury’s out
According to CoinDesk’s mid November report, only 1 in 4 users has upgraded to SegWit since its late October release. Not a bad start you may say, but when you factor in the other proposals fighting for attention, it doesn’t appear realistic that it will achieve 95% uptake by the November 2017 deadline. In fact, I strongly suspect the jury will be released from their duty well in advance. This hearing is now adjourned pending further evidence.
This article represents the personal opinion of the author and is not a recommendation to buy or sell Bitcoins.