Segregated Witness, Segwit - Discussion

A block can fit 1MB of txs. As for how much in particular it depends on the transactions itself because not all txs are the same size. :slight_smile:

3 Likes

Make sense, thanks for responding :hugs:

Because the 4mB Block has to be distributed to all the nodes in the network, which I guess will take longer then a 1 mB Block? Maybe I am not right, but this would be my logic :slight_smile:

True that does slow down things a bit, but modern internet infrastructure fast enough it would be hard to find someone that wouldnā€™t be able to download a 4MB of data in 10 minutes :stuck_out_tongue:

Hi all, after SegWit, now that signiture is seperated on full nodes storages, to what information is currently signiture conntected to? And would changing the signiture even work? It would need to check with other full nodes to validate that change, but since they have different data they will not approve it. Right?

Iā€™m not sure what you want to know exactly. Signature is connected to an UTXO and its not part of the transaction hash calculation process.
Segwit nodes donā€™t need to check with other nodes for the data since they have it themselves and outdated nodes are unable to query that data, therefore they are unable to process segwit transactions.

1 Like

If you want at least imagine how many transactions can approximately fit into one block, I was curious about this as well in the past. I found out that for example for block 656259 there is 1991 transactions (btw 1991 is year when I was born, coincidence? :smiley: )

If you are interested in information about particular blocks, you can check them on https://www.blockchain.com/explorer where you can find many of cool stuff.

1 Like

Hey there Tom, :wave:

Pleasant to hear from you. I hope you are wellā€¦

Yes, what a coincidence for you in that particular block height, it also has 16 confirmations. Is that your birthdate? :grinning:

I also found that block at height 1992 only has 1 transaction :wink:

1 Like

Haha, unfortunately not :confused: :smiley: However I can see it already increased to 20 confirmations. I can wait a bit until it is 25 confirmations (and that is my birth DAY - what a coincidence!! :smiley: )

Yes, in that early time no one really used bitcoin. So if you can see that 1 transaction is basicaly only a bitcoin reward to miner who mined that block (which was 50 BTC in that time, crazy) :slight_smile:

1 Like

Maybe an elemental level question: I ve watched very closely the video on introduction to segwit and havenā€™t found clarity on the following:
Alice sends transaction to Bob and he is able to change the appearance of the transaction signature: that I think I get. Now, does he achieves this by him replacing his signatureID with another one of his signatureIDs, or modifying her signatureID? Obviously he wasnā€™t able to do this once added to the block correct? Thnx
Iā€™m gonna watch it for the 10th time:/

Bob modifies the witness part of the transaction thus changing the transaction hash. This was moved out of the calculation in segwit.
Its true that the transaction must not be confirmed yet and malleability is actually quite difficult to pull of in practice and must be timed well, but it is (was) possible. :slight_smile:

1 Like

Thank you Alko
:smile:The forum really helps clearing things up

Wow, did anyone else start thinking about civic and cultural reform during this last video, Segwit Implementation & Deployment?

  • yes
  • no

0 voters

I still do not understand how Bob can change Alice signature and still achieve to debit the coins from her account. So I guess there is far more complexity into how a transaction occurs than I can handle. I would have believed that without her signature there was no way to debit anything from her account.

@filip in Background Part 2 I have a question (lots of questions :slight_smile: ) about the Alice sends to Bob example.

Is there any evidence left behind visible in the block explorer that shows Bob changed the scriptsig?

Does Aliceā€™s txn make it to the mempool?

How can Alice prevent this happening? Donā€™t send the txn id to Bob until she sees that her transaction got mined and has at least 6 confirmations?

Are there many known cases of people doing this?

Thanks!
Brian

The only thing visible in the block explorer is the new txid that has been changed.

Yes, but it gets replaced by the new transaction with a different id. Once the transaction is mined the old tx gets dropped because the UTXOs used were spend and is no longer valid.

By using Segwit :stuck_out_tongue:

The transaction Alice intended to make never gets mined because it was dropped from the mempool.

Tx malleability is actually quite hard to perform and must be timed well. For humans this is not really an issue since I imagine the human Alice would check the address and would notice the balance has been updated. Robot Allice however relies on checking the txid and since the txid doesnā€™t exist anymore she will try to make a new tx thinking the one before was dropped. :slight_smile:

2 Likes

But it would not get dropped from the pool if she did not send the txn id to Bob which allowed him the opportunity to change the signature? :thinking:

Well she could also send the tx to Bob to propagate it to the mempool, then the malleability issue is even more critical because Bob has all the time in the world to change it. :slight_smile:

1 Like

Hi, what is the relation between the increase of the block size and the increase of centralization?
Also I do not understand very well what segwit is.
Thanks

Increasing the block size would make the infrastructure required to run a node more expensive so less people would be able to afford it. :slight_smile:

Did you watch all the videos on the academy about it?

1 Like