Skip to main content
10 events
when toggle format what by license comment
Nov 1, 2023 at 22:02 comment added LeaBit @UgamKamat Are you sure about you answer? bitcoin.stackexchange.com/questions/120298/…
Oct 20, 2021 at 9:00 history edited Ugam Kamat CC BY-SA 4.0
added 246 characters in body
Jun 28, 2021 at 20:16 comment added Dave Scotese I think that discussing a relationship between a transaction and an opcode ("CSV is a script level opcode that is used to lock the spending of a particular output of a transaction...") is misleading. I thought the script has no awareness of transactions, only of the specific UTXO that it's return value, if true/1, unlocks and which therefore makes the transaction to which it's an input valid. Ben Carman suggested nSequence apply to "at least one" rather than "the first" (input), and I thought "Why not use a bit or two in nSequence to identify what it applies to?"
Jan 22, 2020 at 13:43 history edited Ugam Kamat CC BY-SA 4.0
deleted 282 characters in body
May 7, 2019 at 12:04 history edited Ugam Kamat CC BY-SA 4.0
added 633 characters in body
May 2, 2019 at 8:51 vote accept Michael
Apr 29, 2019 at 17:00 comment added Ugam Kamat @PieterWuille I always had the opinion that it was to prevent DoS attacks. Earlier, no minrelayfee was required and hence that allowed nodes to spam the network at no cost by just incrementing nSequence. BIP-125 actually brought the protocol that the fees should be higher than the previously relayed transaction.
Apr 29, 2019 at 16:48 comment added Pieter Wuille The old nSequence semantics (where non-final transactions were held in the mempool until finalization) were not only a bandwidth issue, but also a memory problem. I believe that was the original reason for removing it, long before the mempool got an actual bounded size.
Apr 29, 2019 at 16:10 history edited Ugam Kamat CC BY-SA 4.0
added 576 characters in body; added 93 characters in body
Apr 29, 2019 at 11:19 history answered Ugam Kamat CC BY-SA 4.0