16:56:52 <m-relay> <0​xfffc:monero.social> Does anyone know why after doing `pop_blocks` we get transaction verification failure error?
16:57:02 <m-relay> <0​xfffc:monero.social> image.png
16:58:17 <xFFFC0000> ( image stuck at the Element side )
17:47:02 <m-relay> <j​effro256:monero.social> If your node is still connected to the network and you do `pop_blocks`, then when peers notify you of new transactions, they will fail to verify because your blockchain is too short
17:53:15 <m-relay> <r​brunner7:monero.social> Sure, but why then doesn't this happen all the time? Because while still in syncing-to-top mode, however that's implemented, you don't start yet to verify something?
17:54:38 <m-relay> <r​brunner7:monero.social> Well, the daemon of course starts to sync (again) almost immediately after the popping, but *something* must now be different.
17:56:52 <sech1> Don't quote me, but IIRC it is because pop_blocks moves transactions from popped blocks back to mempool, then starts syncing and gets the same transactions back in synced blocks, and somewhere in between there is a conflict between them
18:01:56 <m-relay> <r​brunner7:monero.social> Found the code that puts them back into the pool. The writer felt compelled to write some comment there, this may be quite tricky overall: https://github.com/monero-project/monero/blob/cc73fe71162d564ffda8e549b79a350bca53c454/src/cryptonote_core/blockchain.cpp#L645
18:03:09 <m-relay> <j​effro256:monero.social> rbrunner7: probably because `pop_blocks` does not set the value of "sychronized" to `false.` So it doesn't "know" it's not synchronized for a little bit
18:04:03 <m-relay> <0​xfffc:monero.social> Makes sense. When I was debugging it I noticed the failure happened in incoming_transaction method IIRC.
19:56:55 <vthor> Holly hell, where is the janitor of the code? :D Who can tell me how to convert the `Monero key image export\x03` into the json format of the monero-wallet-rpc (offset unsigned int, array<string, string> {bin2hex( keyimage), bin2hex(signature)} (I guess))  and back, or point me on the point of code where I can figure that out.
20:00:50 <vthor> I'm jumping there around in the source and it makes me dizzy (I read now on SO that the hole file is encrypted with the view key (except the header), is this correct?) With https://xmrchain.net/rawkeyimgs I can't also not verify the key, but after browsing in that source it seems that is an older format because the magic header is there: `Monero key image export\002` 
21:23:41 <vthor> I found it, but now I'm more dizzy, seems even more complicated then encrypting the seeds for a standalone PoC....