Bitcoin Forum

A 3.7MB SegWit Blocks Was Mined On The Bitcoin Testnet

Segregated Witness remains a topic of some confusion among bitcoin enthusiasts. According to a testnet block explorer, a 3.7MB SegWit block we mined not too long ago. That is rather unusual, yet no reason for concern just yet. There are always exceptions to deal with and it appears very large SegWit blocks may occur every so often.

A lot of people are surprised to learn there is such as a thing as “oversized” SegWit blocks. In most cases, most SegWit blocks are around 2.1 megabytes. That correlated to the proposed size increase this solution offers. However, that doesn’t mean blocks can’t be larger. To be more specific, on a very rare occasion, one may come across a block that is a lot larger.

http://www.newsbtc.com/2017/04/06/3-7mb-segwit-blocks-mined-bitcoin-testnet/

I'm not surprised that there may be some oversized Segwit blocks
but I would love to know why is it bad to have oversized blocks as opposed to all of the blocks of the same size?
"What is even more interesting is how this particular block only contains 468 transactions. The rest of data all relates to SW information. If it had contained 3.7MB worth of transactions, that would be quite worrisome" - why is it a bad thing
isn't the block size increase the main point of BU hardfork suggestions?

It's a clever (or not so clever) interpretation that this journalist is presenting here.


The so-called "Segwit data" already exists in all Bitcoin blocks, today (well, the "wit" part does, but not the "seg" part does not, but it's not even data, just a way of organsing the data). And always witness data always existed in Bitcoin, Satoshi made it that way, Bitcoin couldn't work without it.


"Segwit" data is simply the signatures for the transactions. The signature proves the owner of the BTC sent it, Bitcoin couldn't work without "witness data" (which just means "signatures")

The only difference in Segwit is the "Seg" part, not the "wit" part. Segwit puts the signatures in their own block, so that the signatures can't be switched to their inverse "low-S" form (a bit like how 2x2=4 and -2 x -2 = 4 also).

That's it. There is no mystery about "worrisome Segwit data" when you actually know what it is.

"A lot of people are surprised" Indeed
I was under the impression segwit had hard caps just like the current system.
So this can happen with the current set up too?