Gold collapsing. Bitcoin UP.

[albin]

The Core shills are doing their best to talk it down on /r/bitcoin. Of course bandwidth optimisations are a bad thing because weak blocks and IBLT are potentially better - only one problem - they don't bloody exist. Plonkers.

This relates to my absolute favorite bout of kettle logic: the "relay network is better so we shouldn't do any optimization" vs the "I'm ignoring the relay network's existence in my argument for small blocks" game.

 

[solex]

@albin,
You are right, but incredibly both Corallo and Maxwell were going on about how the RN is maxed out at 1MB and Corallo thought its efficiency would seriously fall off if the limit was raised. This was sometime last year, probably on BCT, been so much water under the bridge since. I can try and find the reference, but is it worth it?

 

[Roger_Murdock]

So, as someone who generally feels like the least-technical person on this forum and who hasn't really dug into segwit, am I crazy to think that it has almost nothing to do with scaling? In other words, I see people say that it's not "primarily" about scaling and that the effective increase in the block size limit is more of a lucky "side effect" or "bonus"... but even that seems wrong. The proposed formula could have been:

transaction data + witness data <= 1,000,000 bytes

which would have resulted in no increase in the block size limit, correct? But it would still have addressed segwit's main goals (i.e., fixing malleability). So it seems to me that the effective increase in the block size limit that we get from Core's actual SegWit proposal is more like a bribe: "hey, Bitcoin network, you know how you're now desperate for an increase in the block size limit because of the years we've spent stonewalling and goalpost-moving on that issue? Well, if you give us segwit (which we need for the Lightning Network), we'll give you a (token) increase in the (effective) block size limit. Oh, but not an honest, straightforward increase. No, part of the deal is that the effective increase must provide a huge (and completely arbitrary) discount for witness data, which -- it just so happens -- will provide an effective subsidy for Lightning Network settlement transactions (which will have more witness data than typical transactions)."

Is that about right?

 

[cypherdoc]

let's go:

 

[adamstgbit]

I'm always looking for arguments against segwit. That thing smelled extremely fishy to me since the beginning. I just don't like it. It's such a weird cumbersome package. Doesn't solve the issue and adds complexity, huge workload and also risks... not even talking about the unkown unkowns.

segwit feels like one of these 10,000 page document lawyers use, designed to berry some critical information they dont want the opposing side to know about.

And if it wasn't for the fact that it will increase effective blocksize, I bet it wouldn't be an easy sell...

Not only is it being rushed because its being used as a blocksize increase, its adoption will be rushed for the same reason, if it wasn't for this magical 2X effective blocksize, i bet there would be a lot more peer review done on it, and i'm sure some concerns would be raised, and possibly a major catastrophic flaw would be found and fixed before it was widely adopted.
[doublepost=1458010863,1458010255][/doublepost]

So it seems to me that the effective increase in the block size limit that we get from Core's actual SegWit proposal is more like a bribe

totally!
note how no one is talking about the implication of segwit's original purpose. Last i heard about TX malleability ( 2 or 3 years ago ) is that it was a feather not a bug... somthing tells me they dont want us to know anything about segwit beyond that it's goooood and we need to buy it NOW!

 

[Mengerian]

Is that about right?

Yes.

You have characterized the situation accurately, in my opinion.

 

[adamstgbit]

put dramatically, it's more like...

 

[rocks]

Come on guys, how can any of this be a surprise. Roger owns a lot of btc. People are already scared shitless of Ethereum, and now you're coming in with your pesky Satoshi coin. They could swallow XT and Unlimited because they thought their wallets were safe, but you are encroaching on big money now. Molecular has it right.

That may be why they are doing it, but it makes no sense and is wrong.

ETH is a threat to their wallets, along with any altcoin using a new Genesis block. Satoshi's Bitcoin is not a threat to long term holders though..

Long term holders of BTC have a stake in all forked branches because the coins will exist on both branches. Roger and all will own a significant stake in any forked version of Bitcoin that works.

It's almost as if none of these guys have a clue how Bitcoin operates.

 

[Bloomie]

 

[Roger_Murdock]

Yes.

You have characterized the situation accurately, in my opinion.

Well, shit. Ok, so SFSW seems to be basically a poison pill with a candy coating. In that case, I'm a little alarmed by how readily Bitcoiners on reddit seemed to be to swallow it. The sentiment among most "big-blockers," including people whose opinions I respect, seemed to be that "oh yeah, everyone likes SegWit, but we should do a simple HF to raise the block size limit first."

 

[adamstgbit]

whats not to like about segwit
https://bitcoincore.org/en/2016/01/26/segwit-benefits/

afaik, the only thing anyone can dislike about segwit would be how it's implemented, i mean there's alot of code for segwit and its complex, testing on the testnet forked the testnet ( thats to be expected , but still it highlights how complex segwit is and the NEED to get this RIGHT )

i havn't heard anything negative other then " its a complex update "

 

[Richy_T]

The accounting tricks are a negative and have been decried in this very thread.

 

[adamstgbit]

accounting tricks aside it's a worthwhile piece of code. modula.

I mean, I know very little about the details but it sounds all good, and as a bonus it helps scaling
and it helps scaling in a big way, segwit isn't == 2MB , segwith == 2X Block_size_limit
with segwit 2MB blocks "feel" like 4MB!
4MB feels like 8MB
8MB feels like 16MB
...
512MB feels like 1024MB

this is no "poison pill " ( as long as it actually works and doesn't break bitcoin... )
[doublepost=1458018194,1458017495][/doublepost]Maybe i'm overstating the scaling benefits because i dont understand it...
but there's a reason segwit is part of classics road map, and there's a damn good reason why its at the bottom of the road map, not the top.

 

[cypherdoc]

@adamstgbit

Go through my posts here for an extensive list of why I don't like SW.

 

[adamstgbit]

@cypherdoc

i'll need to start a new thread about it, sounds fascinating. this kinda thing shouldn't be buried inside a massive thread like this.

 

[Roger_Murdock]

@adamstgbit

Well, I think making a large change to the economics of transaction fees with this completely arbitrary discount factor for witness data is not a small thing. Similarly, I think the "complexity" criticism is one that should be taken seriously. I don't think we should be making complicated, fundamental changes to Bitcoin's basic structure without a very good reason, and without considerable deliberation and testing.

But yeah, "poison pill" may have been a little strong. So I'll revise my analogy. SFSW is a peanut-butter *cough* and turd sandwich: "Eat up, guys! I mean, everyone likes peanut butter, right?"

 

[cypherdoc]

@cypherdoc

i'll need to start a new thread about it, sounds fascinating. this kinda thing shouldn't be buried inside a massive thread like this.

if you prefer, go thru my tweet history of the last month. it's not long.

 

[adamstgbit]

@adamstgbit

Well, I think making a large change to the economics of transaction fees with this completely arbitrary discount factor for witness data is not a small thing. Similarly, I think the "complexity" criticism is one that should be taken seriously. I don't think we should be making complicated, fundamental changes to Bitcoin's basic structure without a very good reason, and without considerable deliberation and testing.

But yeah, "poison pill" may have been a little strong. So I'll revise my analogy. SFSW is a peanut-butter *cough* and turd sandwich: "Eat up, guys! I mean, everyone likes peanut butter, right?"

agreed.

Its quite disgusting how core is using segwit's block size increase property to push it out ASAP while keeping the discussion about segwit mainly focused on its increase block size sideeffect. when a perfectly valid, simple, easy, and ultimately unavoidable fix to blocksize is readily available (2MB).

but thats no reason to just throw away the idea.
[doublepost=1458021062][/doublepost]

if you prefer, go thru my tweet history of the last month. it's not long.

I will make a thread at some point called "SegWit Pros and Cons - a comprehensive list " i'll try to find some of the cons in your posting history.

 

[johnyj]

@adamstgbit
Looking forward to see your segwit list, in fact I wonder how come a fundamental change to the transaction format like segwit could reach consensus while a much simpler change of 2MB block size limit get no consensus, people will reach consensus on any thing that they don't understand?

 

[Christoph Bergmann]

@adamstgbit

Sorry, writing on a phone .... About sw
1. Nobody understands why we need to get rid of malleability.
2. It allows ca. 2 mb of usual tx, but an attack vector of 4mb. Of blocks are 8mb in throughput, you habe attacks with 16mb. Maaku already says after sw we cant have bigger blocks. Its no scalng, its anti onchain scaling
3. It creates anyonecanspend tx. This means its completely irreversible
4. It breaks the relation between fees and blckspace
5. It allows the devs to so many crazy things via softforks. That may be good or bad, depending on The trust in devs.