21:51:43 <Nausicaa_> transaction pending for half an hour even though the wallet says there's no backlog at priority 2 and above. looking at the explorer i notice a lot of them paying a fee-per-byte that's just above a preset level, e.g. 21 instead of 20, 84 instead of 80, 328 instead of 320. these guys are cutting in line!
21:53:38 <Nausicaa_> * nanoneros per byte
21:54:09 <m-relay> <s​yntheticbird:monero.social> Wouldn't it be better for the protocol to just ignore fees that are out of the preset ?
21:54:20 <m-relay> <s​yntheticbird:monero.social> drop transactions*
21:55:59 <Nausicaa_> iirc rucknium had an idea for fee discretization
21:56:15 <m-relay> <s​gp_:monero.social> To some extent, this phenomenon is unavoidable because payments can be made to miners out of band. However, it's probably sensible to have a better fee scale
21:57:40 <m-relay> <r​ucknium:monero.social> AFAIK isthmus originally suggested fee discretization. Then koe implemented it in the proposed Seraphis code. But AFAIK there hasn't been much analysis of possible unintended effects of fee discretization.
23:59:34 <m-relay> <a​rticmine:monero.social> The basic idea behind having set fees was to disincentive fees being used to fingerprint transactions. This in turn could help with privacy.
23:59:35 <m-relay> <a​rticmine:monero.social> My take is that the impact of this after FCMP is going to be a lot less than now. There is much less impact if a subset of users choose to dox themselves after FCMP than before. 
23:59:36 <m-relay> <a​rticmine:monero.social> Fees are not and cannot be consensus. So we can review fess after FCMP is online. One option may be to increase the number of fee tiers.