pablomartin has quit [Remote host closed the connection]
brunoerg has joined #bitcoin-core-dev
zeropoint has quit [Quit: leaving]
jonatack has quit [Ping timeout: 252 seconds]
<bitcoin-git>
[bitcoin] kevkevinpal opened pull request #30068: test: assert can't activate snapshot based chainstate more than once (master...assumeutxoActivateTwice) https://github.com/bitcoin/bitcoin/pull/30068
<bitcoin-git>
[bitcoin] fanquake closed pull request #30068: test: assert can't activate snapshot based chainstate more than once (master...assumeutxoActivateTwice) https://github.com/bitcoin/bitcoin/pull/30068
abubakarsadiq has joined #bitcoin-core-dev
aleggg has quit [Ping timeout: 268 seconds]
aleggg has joined #bitcoin-core-dev
brunoerg has quit [Remote host closed the connection]
<sdaftuar>
not much of an update from me this week either. i did push a rebase up to the draft PR, and i'm working on some slightly esoteric things right now. still planning to udpate the branch with better benchmarks and tests in the near future.
<sipa>
i'm getting close to having some PR'able low-level cluster linearization code.
<sdaftuar>
i think as glozow said, #28984 is probably the best thing to be reviewing/working on as a next PR
<instagibbs>
glozow suggested it, people can yell at me to close if they don't think it's necessary, whatever aids review
<josie>
achow101: started with it, was curious if there were any bdb wallets from when you and murch were doing wallet simulations?
<josie>
or more generally, if anyone here has old bdb wallets that can be used for testing that PR
<achow101>
josie: I might've deleted them
<glozow>
I think there would be a conflict. I just figured it might make the PR smaller
<achow101>
josie: since it uses the test framework, they all get cleaned up at the end of the script
<achow101>
but yes, if anyone has old bdb wallets, testing them on 26606 would be useful
<josie>
yeah was more just curious to try it on a variety of wallets in the wild. i remembered the simulation wallets actually had some interesting txs and behavior, which is what made me think of them
<Murch[m]>
josie: Sorry, also don’t have any
brunoerg has quit [Remote host closed the connection]
<cfields>
achow101: current test failure relevant? It's wallet-related.
<sipa>
josie: i think expected issues would be more due to being created with old versions, having gone through many (but maybe not all) version upgrades, ...; not so much the content (in terms of addresses/transactions) of the wallet
<sipa>
maybe very large wallets might trigger edge cases
<achow101>
cfields: i probably have to rebase. that ci task is the one that tells us about silent merge conflicts
<fanquake>
cfields: a number of wallet tests just fail semi-regularly, looks like one of those
<josie>
sipa: yeah, iirc we were simulating with very large wallets and some of them had been around forever, which is why i was hoping for old versions/upgrades etc. agree that the tx content itself isnt that important for the bdb parser
<theStack>
achow101: the bdb parser in the test framework assumes a pagesize of 4096 bytes. i wondered where does number comes from? did earlier versions of bitcoin core set a fixed pagesize? (is this even possible?)
<achow101>
It's the bdb default
<theStack>
ah, i see. IIRC it was 16K on my machine using the depends build when i last checked
brunoerg has joined #bitcoin-core-dev
<Sjors[m]>
sipa: IIRC you mentioned a while ago that you have a WIP improvement to the Transport class and friends
<sipa>
Sjors[m]: i do not recall
<Sjors[m]>
For Stratum v2 I'd really like to kill its dependency on CNetMessage
<Sjors[m]>
(and CSerializedNetMsg)
<sipa>
oh, i had an idea for an improvement, but never implemented
brunoerg has quit [Ping timeout: 272 seconds]
<achow101>
theStack: the page size is stored in the wallet, so if it's different, we'll be able to handle it
<Sjors[m]>
sipa: can you project these thoughts as a comment on #29432 or a new Github issue? I might take a stab at it if you don't have time.