Up to 78.5% over the last rolling 24 hours now.

Indeed it has achieved its aim which is to threaten miners into submission. To that end it has been successful and must continue to exist and appear to have support until segwit really does get activated.

Here's the original document on PPLNS:
https://bitcointalk.org/index.php?topic=39832

How ridiculous the face saving code has become is nicely summarised by Samson Mow.

Instead of just signalling BIP141... One day history will look back upon our blockchain and facepalm. Heck I think it already is.

Meanwhile support just hit 64% over the last 24 hours.

Pretty much this. A lot of the wallets, both online and standalone are not that smart with respect to fees and have been manually adjusted to higher levels and haven't been adjusted back down. Anyone running bitcoin core/qt which has a dynamic fee estimator would have noticed the drop in recommended minimum fees to just over 100 satoshi per byte over a week ago.

This site has been showing it for a while too if people have been paying attention instead of using some braindead default:
https://bitcoinfees.21.co/

They're infamous for doing shit like that purely to save face from having to admit they're backpedalling on their BU stance. It's just a coinbase message so it doesn't cost them anything in terms of risk to leave it in there. BU was always just a bargaining tool for them in the first place once UASF was mentioned. Now that they have a way to make it look like they're not actually agreeing to core, even though they actually are, they can finally get rid of it but they're proving their usual level of inertia with previous claims. Most of the other pools that are either covertly or overtly under bitmain's control are still doing the same thing. These morons need to get off their arses and be clear about their message and stop adding conflicting messages to their coinbase. The joke got tired a long time ago.

Segwit2x support over 24 hours is now over 55% making it higher than any activation signal to date and looks set to easily get to the 80% required to start signalling in mid July. Adding the existing BIP141 signalling pools to the total means segwit is a surety unless something happens for them to start backpedalling on their agreement.

This whole approach was really stupid in the first place by the mining agreement and I'm just glad someone stepped up to the plate to make their incompatible bit4 nonsense compatible with existing core nodes and segwit activation.

Basically the ones that didn't sign the agreement. Many of the pools that were already signalling segwit haven't added NYA to their signature (they didn't sign the agreement in the first place), and some that haven't made any decision still haven't made any decision.

Absolutely no change in any way. The only thing it means is if you are on a pool that isn't signalling segwit after mid July, and more than 80% of the network is signalling segwit by then, your pool's blocks will all be invalidated.

What I see happening is BIP148 user activated soft fork will be a miserable failure but it won't matter. By that stage most pools will be signalling segwit thanks to segwit2x. Segwit will get activated and then fees will remain low, blockstream will not commandeer the blockchain, lightning network will not centralise transactions, nothing bad will happen and everyone will wonder what the fuck all the fuss was about. There will be no need for antpool with Jihan at the helm to create an aggressive hard fork against UASF; that was always a bullshit powerplay "I want my lollipop back" from Jihan as a counterthreat to UASF which hasn't got anywhere near enough traction to be relevant anyway. So nothing will happen to diff, there will be no fork, there will be no change of PoW. Forget the talk of diff drop to 2014 levels; that's off the cards entirely. The confidence that consensus will instil in the market will be reflected by yet more BTC value increases... till the next phase. The fun part is now that most of the pools are agreeing on what to signal next, transaction fees have suddenly plummeted due to total transactions plummeting - which means it's almost certain that someone amongst the pool groups was the architect behind the transaction flood which made it look like there was a transaction block space crisis. There's no economic reason for the total number of transactions to crash the way they have over the last week with BTC value being so strong right now.

After that people will be left scratching their heads wondering if we even need a 2MB block size increase at this time, but because all the pools made such a fuss about it, we'll have our next debate, but the pools have already insisted they want a block size increase and signed the NYA. I'm not sure how core will respond at that stage because that will be the true test of whether the miners can effect a change in consensus - but despite all the anti-core rhetoric, core dev has always overwhelmingly said they would do a block size increase but with a much slower time frame than that proposed by segwit2x simply because the regression testing required to confirm a hard fork is safe is substantial. I predict we'll get a 2MB base blocksize increase as well, though how and when I can't tell at this stage.

By the way, the current segwit2x "signalling" is just a show of support in their coinbase signature and doesn't change the way anything operates. However what it is saying is that they will be actively signalling from mid-July for segwit.

What's also amusing is the corresponding drop in transactions back to normal basal levels again now that all those pools are in agreement about what to do next. This adds circumstantial evidence that someone was injecting all that shit into the transaction biosphere to make it look like there was a transaction volume crisis there that didn't exist.

By the way, if you've all only been remotely watching the scaling debate from a distance, segwit2x appears to truly gaining enough traction to bring segwit activation to the fore.

The pools that signed the agreement just started signalling for it and it is taking off faster than any other scaling signalling. See:
https://coin.dance/blocks

Unlike the original proposal, it includes BIP91 meaning it is compatible with core's segwit now. The rapid adoption of segwit2x, which with the addition of BIP91 means that miner activated segwit is now likely on the horizon at last, and since there is overwhelming support for it and the associated 2MB hardfork in the future, that may also follow.

Unlocked since the original discussion is still relevant.

I'm going to call it now, segwit2x will be responsible for activating segwit at last due to overwhelming miner support, BUT great thanks to James Hilliard for introducing them to BIP91 which makes segwit2x compatible with the other segwit activations.

Watch this for the very rapid rise in segwit2x support that will happen fast now:
https://coin.dance/blocks

Segwit may well get a miner activated soft fork before BIP148's UASF which remains in limbo in terms of meaningful support but seems to have had the desired effect of forcing miners' hands in one way or another.

What happens after segwit though is a mystery, but there is enough miner hashrate support for 2MB that it might happen too.

Yes it's correct. Mining bitcoin on GPUs and CPUs became obsolete 4 years ago.

Hah that looks like the imagination episode. Let's all imagine what it's going to be like when we start finding blocks.

Whatever the answer is, one thing's for sure: Mr Shit Arse has more hashrate than us.

Let's not labour on what other pools are doing shall we?

Pretty sure I edited out for false advertising.

Who's barry?

I certainly did not support his agreement's proposal in the form they agreed to, only the concept of 2MB+segwit, but that's all offtopic here.

Original poster hasn't come back and question answered enough. Locking thread.

Have you been mining here for 4 weeks? The value is an exponentially decaying average with a time constant of 7days which isn't quite the same as your average for only the last 7 days. It's close enough to the same value normally that I let people assume that's what it is.

What's not merged, BIP148? Of course not, there isn't remotely universal acceptance by core devs.