💬 andrewtoth commented on pull request "Don't empty dbcache on prune flushes: >30% faster IBD":
(https://github.com/bitcoin/bitcoin/pull/28280#discussion_r1660341788)
Done, nice catch!

💬 andrewtoth commented on pull request "Don't empty dbcache on prune flushes: >30% faster IBD":
(https://github.com/bitcoin/bitcoin/pull/28280#discussion_r1660341870)
Updated comment.

💬 andrewtoth commented on pull request "Don't empty dbcache on prune flushes: >30% faster IBD":
(https://github.com/bitcoin/bitcoin/pull/28280#discussion_r1660342034)
Mentioned in another comment, we need GetFlags or we will have to change a lot of test code in `coins_tests.cpp`.

💬 andrewtoth commented on pull request "Don't empty dbcache on prune flushes: >30% faster IBD":
(https://github.com/bitcoin/bitcoin/pull/28280#discussion_r1660342135)
Added a comment in the commit message.

💬 andrewtoth commented on pull request "Don't empty dbcache on prune flushes: >30% faster IBD":
(https://github.com/bitcoin/bitcoin/pull/28280#discussion_r1660342186)
Done.

💬 andrewtoth commented on pull request "Don't empty dbcache on prune flushes: >30% faster IBD":
(https://github.com/bitcoin/bitcoin/pull/28280#discussion_r1660342456)
The std lib uses this pattern though, with `std::unordered_map::erase`. It mutates and returns next. Surely we should be following std lib conventions, no?

💬 andrewtoth commented on pull request "Don't empty dbcache on prune flushes: >30% faster IBD":
(https://github.com/bitcoin/bitcoin/pull/28280#discussion_r1660342534)
Updated comment.

💬 sipa commented on pull request "cluster mempool: cluster linearization algorithm":
(https://github.com/bitcoin/bitcoin/pull/30126#discussion_r1660345204)
Yeah, it is to reduce the ability for peers to construct cases that just happen to be worst case for the exact search order they know we're going to try.

💬 sipa commented on pull request "cluster mempool: cluster linearization algorithm":
(https://github.com/bitcoin/bitcoin/pull/30126#discussion_r1660349106)
Done, named as `MAX_SIMPLE_ITERATIONS`.

💬 sipa commented on pull request "cluster mempool: cluster linearization algorithm":
(https://github.com/bitcoin/bitcoin/pull/30126#discussion_r1660349184)
I've added comments.

💬 sipa commented on pull request "cluster mempool: cluster linearization algorithm":
(https://github.com/bitcoin/bitcoin/pull/30126#discussion_r1660349260)
I have no idea! Added a comment that it was found empirically.

💬 sipa commented on pull request "cluster mempool: cluster linearization algorithm":
(https://github.com/bitcoin/bitcoin/pull/30126#discussion_r1660349319)
Done.

💬 sipa commented on pull request "cluster mempool: cluster linearization algorithm":
(https://github.com/bitcoin/bitcoin/pull/30126#discussion_r1660349373)
Added a comment to `Linearize` to reflect this.

💬 sipa commented on pull request "cluster mempool: cluster linearization algorithm":
(https://github.com/bitcoin/bitcoin/pull/30126#discussion_r1660349427)
Done.

✅ maflcko closed a pull request: "WIP: Fix coinstatsindex overflow issue"
(https://github.com/bitcoin/bitcoin/pull/26426)

💬 maflcko commented on pull request "WIP: Fix coinstatsindex overflow issue":
(https://github.com/bitcoin/bitcoin/pull/26426#issuecomment-2199292529)
Closing for now due to inactivity for 1.5 years. #https://github.com/bitcoin/bitcoin/issues/26362 remains open and can be used for discussion. This change can be reopened or recreated as a new pull request.

💬 maflcko commented on pull request "test: p2p: check that connecting to ourself leads to disconnect":
(https://github.com/bitcoin/bitcoin/pull/30362#issuecomment-2199328998)
ACK 5d2fb14bafe4e80c0a482d99e5ebde07c477f000

💬 maflcko commented on pull request "scripted-diff: Use LogInfo/LogDebug over LogPrintf/LogPrint":
(https://github.com/bitcoin/bitcoin/pull/29641#issuecomment-2199403598)
> I suggest [#29256 (comment)](https://github.com/bitcoin/bitcoin/pull/29256#issuecomment-2197461775) instead.
>
> It may also make sense to finish consensus and work on the simplest possible consistent user-facing API and developer API before doing a mass migration.

I don't think the changes here conceptually conflict with any open pull request. There are discussions around the naming and functionality around `LogWarning` and `LogError`, but I don't think they are changed or touched in th
...

🤔 glozow reviewed a pull request: "optimization: Switch CTxMemPool::CalculateDescendants from set to vector to reduce transaction hash calculations"
(https://github.com/bitcoin/bitcoin/pull/30325#pullrequestreview-2150632999)
Seems like a reasonable thing to do, but this conflicts with / is already done in #28676. See the "Select transactions for blocks based on chunk feerate" commit there which gets rid of this call.

I think our current policy is to avoid making refactoring changes that would need to be redone or removed in cluster mempool.

✅ maflcko closed an issue: "Add "maxuploadtargettimeframe" to change the timeframe considered by "maxuploadtarget""
(https://github.com/bitcoin/bitcoin/issues/30176)