Good, this is the first if...then added to fix a specific logic flaw, and this will not be the last of them, keep adding if...then until all the code base become an ugly mess and you still might have missed some scenario

You really don't need all these wasted efforts of patch making and testing if you just do a hard fork

That's the problem with current core's approach, they don't make things easier to understand, they make it more difficult to understand, thus any technical debate would easily extend to several thousand lines and just become TL'DR for average people, so no decisions can be made

In principle, changing the current way of signing the transaction is a simplification of the transaction format and can be considered an improvement towards simpler design (even that require months of discussion). However, by doing soft fork, it separate the signature and scripts into another block structure, that is a total complication of the bitcoin architecture and created many new added logic that make the system much more error prone

Of course you can make millions of tests and make the system work to a certain degree, but a slightest future change to the environment will make majority of those tests fail and certain parts have to be redesigned, but you can't do this when a financial system is on 24/7 live traffic. Complex system is fragile, you don't want to build a financial system on a rocket that can explode any time

The complex way of turning a page  
https://www.youtube.com/watch?v=GOMIBdM6N7Q

I think I understand Franky1's sense of security here, let me make a simpler example:

As soon as segwit code is out, some malicious miner create an illegal segwit transaction and mined it into a old style block, for example block 420001. All the old nodes will accept this block since they can not verify if the transaction is correct. So this block containing illegal segwit transaction will be berried deep down until the day segwit activates, lets say 425000

After segwit activation, all segwit nodes detected this illegal transaction (because they could not find the witness part at all) and illegal blocks,  they have to abandon all the blocks after this block and build from block 420001, since that is the latest block they consider to be valid, so network split into two chains from 420000 and all the transactions after 420001 disappear in segwit nodes after the activation

Notice the activation happens after the grace period. What prevent miners from switching back to original version after the activation? If there are so many anyonecanspend transactions to grab. You must collude with miners, so that is the problem with segwit, without miners cooperate, this system will fail, and bitcoin by design should not need devs to collude with miners to implement a change, that is too large degree of central planning

Even if you can collude with miners, in case the segwit upgrade failed, e.g. lots of security problem in live traffic after the activation, because of these anyonecanspend transactions, miners can not roll back to old version, since all these transactions will be spendable if they revert. In that case bitcoin is dead either way: Keep running segwit, customer losing money, roll back, customer losing money

And this one, replace the word "bitcoin" with "segwit" 
https://www.youtube.com/watch?v=4APcgsRdW6w

So the security is to make sure this output can never be spent by old nodes, how can you be sure of that?

Suppose two pools with 60% hash rate, and they are sitting on the side line watching the mined blocks, once 35% of blocks are mined as segwit support, they switch to mine segwit blocks and activate it, then there will be many segwit transactions out there, and then they switch back to original software

Now network are full of anyonecanspend transactions that is free for them to grab, and they have 60% of hash power so their chain become the longest, since majority of the nodes are still old style nodes (as in soft fork you don't need nodes to upgrade),  these nodes will build on their chain, and the rest of the segwit nodes will fork into a minority chain since now they reject old style blocks

So it is a mess after this accident, all the people who used segwit transaction will lose their money and they don't even understand what happened

The biggest property of segwit is that the more you talk about it, the more FUD you get  

If changing 1 to 2 require 18 months of talk, then changing several thousand lines would require many decades of talk if not centuries

A segwit  transaction is "anyonecanspend" type of transaction in the view of old nodes, but totally different in the view of segwit nodes, that is an attack trying to fool the old nodes. And if I'm a malicious node, I made an invalid segwit transaction with invalid signature data, it will be rejected by segwit nodes, but accepted by old nodes, since old nodes see it as "anyonecanspend" transaction without signature, they can not validate it

It seems you are not capable of rational discussion of the topic, think this way: Segwit is your god, pray 2 times a day to it

Don't make sure conclusions when you only see 10% of the whole picture. Check core's definition of soft fork, it is not that simple, adding rules is only one out of many possible scenarios of soft fork
https://bitcoin.org/en/bitcoin-core/2016-01-07-statement

I can change the transaction format in my node software, and that change will split the chain immediately, since the old nodes do not accept my transaction format and my node do not accept old transaction format. But that's only one out of 4 different scenario: I can reject old transaction format while make my transaction format acceptable to old nodes, I can accept old transaction format while make my transaction format unacceptable for old nodes, or I can make them acceptable both way

That gives 4 different scenarios, and depends on if the hash rate of new nodes are larger than old nodes, in each case there will be 2 different scenarios (except the last) : either one of the chain get orphaned, or both chain extends indefinitely. That's total 7 different scenarios, while in 3 of them the network will partition into two different chains

And this is only for normal soft fork scenario, not segwit like soft fork. segwit soft fork is a special type of soft fork that does not follow the above rules since it created a virtual bitcoin node which looks exactly like bitcoin node and talk to bitcoin nodes, and the old network thought that they are talking to a real bitcoin node while in reality they are not. So a segwit node can break any rules while old nodes still believes they are talking to a normal old node. This action in itself is more like an attack to the network instead of an upgrading, thus is criticized by knowledgeable people

It is consensus, not the software you run decide what is bitcoin. If core changed total number of coins to 42 million, would you still use core software because you want one web?

At what criteria do you define a new feature to be rule breaking? Segwit soft fork demonstrated by using soft forks you can always do rule-breaking changes (like change the transaction/block format) while still not being rejected by the old clients. That change is rule breaking thus it is a hard fork, just sold as a soft fork

I don't want to waste time in useless discussions, I throw in 10 bitcoin to bet that picture is wrong, call or fold? 

Obviously you don't understand how bitcoin transaction works, otherwise you won't ask so strange question. This proved my point, those who is supporting segwit typically don't understand how bitcoin works

I have explained, these wallet operators of course like to promote Segwit since they get a cut in transaction fee, but I'm afraid miners won't buy this

Old nodes only see one of the twin blocks, which is the original one, containing all valid transactions. They won't even receive new type of transactions and new type of blocks, just like old nodes can not receive the witness blocks

I give it in layman's terms: Imagine if you are running windows XP and you used to talk to another windows XP machine in your LAN, you can see the shares, and free space on that machine is 500G etc... But one day, that machine had an upgrade, it becomes a windows Vista, and the original windows XP is converted to a virtual machine installed on that Vista machine. You feel nothing, since you can still see the shares and free space on that machine is 500G, exactly as before. But in reality, that windows VISTA machine gets another hard drive with 3TB of free space, you don't see it because it exists out of your knowledge. This is similar to what segwit soft fork is doing, and that's the reason it requires thousands of lines to implement: It will build a new coin, and make original bitcoin a virtual coin inside this new coin

Wallet operators like it since it gives 25% discount fee on the signature data, which is specially interesting for wallet operators when they clear thousands of micro UTXOs

But this is plainly bribery at system level, such kind of bribery should not appear in the protocol level at all. If this becomes a trend, then miners will adopt a new protocol that especially benefiting them, then it quickly becomes a war of bribery, going down that route you will see this ecosystem is totally corrupted in a couple of years

Upside down, you don't need to read the code, every piece of code is just an implementation of an idea, you first check if the idea is valid. IMO, this design is not bitcoin, can be called Pieter coin or alpha coin or segwit coin etc... so the first question is: Why should we change existing bitcoin architecture to twin-block architecture, and what is the pros and cons of doing that?

There is a long discussion here
https://bitco.in/forum/threads/segregated-witness-sotf-fork-segwit-pros-and-cons.986/

In fact, this large scale of change is much larger than a hard fork, it totally changed bitcoin into something else, and raised level of complexity typically have many bad impact to the system's robustness and integrity

Of course, from a higher level of abstraction, software can implement ANY function if given enough time, anyway it is all data. But there are still many basic rules when it comes to system engineering. There are no good reasons that can explain why you should change the architecture to implement a useless function like fixing the transaction malleability (unconfirmed transactions are not safe, this is a limitation of POW, it has nothing to do with transaction malleability). Fixing transaction malleability is only to serve the LN, but to serve an outdated pre-paid card model (LN), which has already been abandoned by many telecom operators, you want to change the bitcoin architecture? This is like FED decided to adopt another Fedwire system to especially benefit Goldman Sachs, can you see the trend of centralization here?

I'm referring to this first slide:


If you don't understand where it is wrong, then you better spend some time to check how bitcoin transaction works. The description in the Satoshi white paper is also not the current design, but there are plenty of materials out there

Of course upgrading to anything is easy, you just need to download the software and run it, just like you download any virus or trojan, a couple of mouse click then you are set, isn't that simple enough?

That's exactly what blockstream devs are doing, since no one really understand how segwit works, they end up with propaganda and useless debate that does not make anyone more clever, finally: It is super simple, just download it and run it, everything will be fine...

By saying valid outputs and valid block, you refer to original rules run by original nodes. Segwit already demonstrated that original nodes have no idea about the existence of witness data, they think the signature data is missing thus the transaction is "anyonecanspend", while in reality all the signature data is in another witness block that is invisible to original nodes

Similarly, by setting new rules in a new implementation and make it invisible to old nodes, you will have new outputs in new transaction packed in new blocks that old nodes can not see. They still think every block is normal, but they can't discover there is another block attached to each block and the relationship is described in coinbase transaction of each original block

In one word: In new implementation, there are two linked blocks every 10 minutes, but the old nodes see only the original block

This parasitic behavior originally come from mastercoin. They tried to encode message into bitcoin transactions but later found out it is too expensive to send information that way. But segwit soft fork opened a whole new world of parasitic design practice, thus anyone who understand how it works call it a hack

It is witness block, not witness tree, check it again

Of course old nodes won't be able to see new coins, and new coins won't be able to sent to old addresses, but does that really matter if old nodes and new nodes have two different money supply? Like some of the people use both Euro and USD while others only accept one type

Why do you have added money supply while existing users can not do anything about it? That's the problem of soft fork. In a hard fork you would never have this possibility

Let me quote some text from Mike who hold the similar view: