<0​001dev:matrix.org> what is best way to contact core dev team?

<o​ne-horse-wagon:monero.social> Guest 68. Have you port forwarded 18080?

<j​effro256:monero.social> Is it related to a vulnerability ?

<j​effro256:monero.social> If so, use the Hackerone : hackerone.com/monero

<0​001dev:matrix.org> no. but is important

<0​001dev:matrix.org> jeffro256: is any contact except for luigi email shown on github?

<0​001dev:matrix.org> guest 68: windows 7 is out of security support. is not safe to run even on LAN. example: CVE-2024-30078

<r​brunner7:monero.social> 0001dev: If you mean with "core dev team" the devs that actually work on the codebase and are on top of developments: Well, this here is our channel / room. Seems to me you are already in the right place for any dev non-security related dev questions. What you write here will, sooner or later, get noticed by many such devs.

<r​brunner7:monero.social> Of course they are on Matrix.org ...

<o​ne-horse-wagon:monero.social> 0001dev: Current stats claim about 25% of windows users still use windows 7. Excuse is their apps won't work with newer versions.

<r​brunner7:monero.social> one-horse-wagon, you may have to update your info, Windows 7 probably stands at a few percents left: gs.statcounter.com/os-version-market-share/windows/desktop/worldwide

<o​ne-horse-wagon:monero.social> You're right. Thank you.

<o​ne-horse-wagon:monero.social> Was looking at old numbers.

o/ interested in hobby projecting an xmr based (clone+mods) coin, WOW is fun, enjoy the approach. Have developed hobby ERC20 tokens before, looking at the github and mastering monero, not overly skilled in C languages but curious. Wondering if anyone would recommend any specific resources outside of git, mastering xmr, and reddit posts for a

hobbyist?

<s​yntheticbird:monero.social> Zero to Monero 2.0 is a little outdated but is still a valuable resource

<s​yntheticbird:monero.social> for technical understanding

lovely thanks

<j​effro256:monero.social> The mininero implementation can be helpful for understanding how the math is implemented (just don't look at the old script file like I did)

woodser: just to confirm, this is 100% traced back to ubuntu 22.04 used for depends and no other issue? monero-project/monero #9371

what i mean is the same user built from 20.04 and 22.04 and one worked and the other didn't?

Is only cross-compilation broken?

<0​xfffc:monero.social> quick question, when we do `flush_txpool`, we just need to clean the txpool right? do we need go through `parse_and_validate_tx_from_blob`?

<0​xfffc:monero.social> matrix.monero.social/_matrix/media/…ero.social/DaocorQaHKfEmSnkcSnNswkr

<0​xfffc:monero.social> this is profiling results for `flush_txpool` command.

<v​tnerd:monero.social> Yeah it shouldn't be needed. Looks like it was just done as a convenience for using take_tx

<0​xfffc:monero.social> vtnerd: thanks. this is another optimization we can implement then.

Just make sure you only do it for the case of flushing the whole pool.

<s​yntheticbird:monero.social> what tool are you using exactly?

<0​xfffc:monero.social> github.com/google/orbit it is a basically fancy frontend for `perf`.

<w​oodser:monero.social> selsta yes I'm sure the binaries I cross-compile from 20.04 are incompatible with the binaries from 22.04

<0​xfffc:monero.social> is there any extra reason that we do use `validate` when we are starting monerod? this is `O(n^2)` and not efficient.

<0​xfffc:monero.social> take a look at this call order, starting from here, call order for our `O(n^2)` operation:

<0​xfffc:monero.social> 1. github.com/monero-project/monero/bl…e549b79a350bca53c454/src/cryptonote\_core/tx\_pool.cpp#L1802

<0​xfffc:monero.social> 2. github.com/monero-project/monero/bl…e549b79a350bca53c454/src/cryptonote\_core/tx\_pool.cpp#L632

<0​xfffc:monero.social> 3. github.com/monero-project/monero/bl…e549b79a350bca53c454/src/cryptonote\_core/tx\_pool.cpp#L756

if someone updates monerod and the txpool has different rules you want to validate it first I assume

<0​xfffc:monero.social> ( thanks to boog900 )

w​oodser: and non-cross compiled binaries work?

<0​xfffc:monero.social> selsta if we check the hardfork, can we skip the validation?

what's allowed in the txpool isn't necessarily hardfork related

for example we plan to drop timelocked transactions soon without waiting for a hardfork

with a large txpool it can take multiple minutes to start monerod so I thought about it before why it validates the txpool on startup but I don't think it can be removed

<0​xfffc:monero.social> I see, thanks . in anycase, even if we are required to do `validation` , we still doing `O(n^2)` for an operation that we should only do `O(n)` . so that `O(n^2)` can be optimized.

if it can be optimized then that would be nice

jeffro256: please open monero-project/monero #9343 against release branch

<r​ucknium:monero.social> You could also check if the node broadcasts the stale txpool to peers after startup. Some analysis of p2p logs suggests that it does.

<b​oog900:monero.social> We could separate the relay checks, call them on startup and only do full verification if the hardfork changes?

<w​oodser:monero.social> selsta I'm seeing the same result with make release-static from 22.04, where the binaries are incompatible

that's really weird, what hardware does the 20.04 and 22.04 device have?

<w​oodser:monero.social> they are parallels desktop vms running on mac m3

<w​oodser:monero.social> this only applies for arm64

<v​tnerd:monero.social> 0xfffc: are you going to try and patch the n^2 code?

<0​xfffc:monero.social> vtnerd: yes, I am looking into this.

<0​xfffc:monero.social> let me know if you have any feedback on this. thanks.

w​oodser: both are on the m3 mac and both are arm64 / armv8 ubuntu VMs? (20.04 and 22.04)?

<w​oodser:monero.social> yes

can you update the issue since it's not depends related

<w​oodser:monero.social> selsta yes sure, just udpated

<p​reland:monero.social> From my experience, make depends is a nightmare to work with unless you are using the exact runner that the GitHub workflow uses

<p​reland:monero.social> And I do mean *exact*; even a fresh chroot of ubuntu 20.04 errored out for me

<p​reland:monero.social> Albeit I was using make depends to cross-compile for Mac, so it was going to be an uphill battle from the start

<p​reland:monero.social> (For those who want to cross-compile, I have only been able to get it working locally on Debian bookworm; verified on both a chroot and in a Qube)