moneromooo: combining the metadata so that I can consolidate all the transaction details?

Or is there One Secret Trick that lets me concat those files or something?

For all I know technically it should be traightforward to write a tool for merging meta-info from a group of wallet files for the same private key. Maybe would need one or the other new method in `wallet2` to support, but nothing wild I would say.

I'm confused about how transactions get added to the pool with dandlelion, when I send a transaction to my node it displays the "Transaction added to pool" message twice, one seems to be triggered by /sendrawtransaction and the other one by the levin packet NOTIFY_NEW_TRANSACTIONS.

Looking at tx_pool.cpp there is a set called m_txs_by_fee_and_receive_time and the transaction gets emplaced into that set in tx_pool::add_tx, this is getting called once for each "Transaction added to pool" message and the problem I encountered is that the receive time, used to order the transactions in the set and calculated with time(nullptr)

everytime tx-pool::add_tx is called, might vary from one call to another and therefore the same transaction can be found twice in that set.

After mining a block to confirm the transaction it's is going to be considered twice and a "key images already seen" message will be shown, then if I decide to mine another block one of those repeated transactions will still be alive in the set and because of that I encounter different behaviors for everytime I run a test case, from "failed to find

tx meta" because the transaction was removed from the db to having the same transaction in the db with invalid values which look like UB (i.e. extremely big block weight).

Did someone see anything related to this?

I don't understand the question.

Oh, you want to merge several wallet files ?

Starting to get really complicated. Just use the same in the first place.

Interesting, I did not know the meta not found was caused by dandelion, I wanted to look at that at some point.

The txpool code became really hairy with dandelion, I don't think I really understand it anymore :/

It is supposed to handle this though. Buggy, I guess.

Gonna look at it for a bit longer and open a PR if I can find a decent solution to the problem

OK. You'll want vtnerd to look at it otherwise, he groks that new code well.

Hi guys. What is the contrib folder used for? Specifically the .mk files inside contrib/depends/packages?

depends build system

for cross compiling / reproducible builds

So only if you do make depends?

or for release-static-linux-x86_64 too?

if you use gitian it will use depends

Alright thanks

release-static-linux-x86_64 will not use it AFAIK

contrib is for external stuff, typically helpers, glue between monero and other software, packaging stuff, etc.

It's also got some source monero needs for historical reasons.

Good evening

thanks for that

bysnack : on the first /sendrawtransaction call, the tx is considered to be in the dandelion++ stempool, and therefore when a node eventually announces the tx to you it gets upgraded to the normal txpool

unfortunately, these are the same txpool object which might make it harder conceptually for people reviewing the code at-a-glance

I cant parse out what you are saying about mining blocks exactly

I read it as "it tries to mine it twice".

does the monero gui solo miner use XMRIG?

If you mean monero-wallet-gui, no.

ok, thank you

I'm curious why it wouldn't, but that's probably something best researched on my own

Because xmrig is 3rd-party miner, it's not part of official monero binaries

Rip fireiceUK

sech1: ahhh okay that makes total sense. it's all clicking now. the donations, why it's its own thing. thanks I feel less oblivious now

xmrig can also mine like 20 other coins, so it's not monero-only miner

oh cool! I thought the xmrig was "monero rig"

which still looks to be the case. but I see now how it's able to mine other coin

luigi1111: do you have to do merges this weekend?

there are a lot of PRs in the merge bot so rip

time*

.merges

yikers

ikr

code freeze in 4 days

yea that’s not going to work

lol

but merges would help

?

code freeze in 4 days will not work

there are still large unreviewed PRs

or is code freeze no new feature PRs?

Bug fixes are obviously ok, otherwise there wouldn't be any point in waiting before releasing.

Well, the stated code freeze is August 17

If this is not the case, it should be addressed

sarang: what is your definition on code freeze?

I assumed no new feature PRs

that’s okay

Otherwise why bother setting a code freeze?

no new feature PRs should work, I don’t think there is anything planned for next HF that is not PRed yet

Code freeze!

sarang: is there a link with the HF timeline? I didn’t find it

It was in meetings

Code freeze Aug 17, release Sep 17, fork Oct 17

If this is not the case, I would like to know ASAP

This was to provide accurate information to hardware wallet teams

There's a hardware wallet?

sarang: should be good

.merges

clsag PR still needs an approval on github, I don’t know how thorough the reviews have to be because the code is audited

or is only part of the code audited?

sarang: is approval something you can do for the CLSAG PR?

.merges

sorry for testing. it's broken.

Kinda sexy, huh dsc_

will fix later (+v requirement)

what's up with the release notes for 0.16.0.3 in monero and monero-gui? is this urgent?

luigi kinda only half uploaded them

kpcyrd: it’s a placeholder, the website has release notes and they should also get uploaded to github soon

selsta: the PR should receive additional scrutiny IMO

selsta: ok, thank you :)

I wrote some of that code and should not be reviewing it

I can haz PR?

sarang: can you review the commits you did not write?

also which commits were part of the audit?

The audit reviewed all the commits. I did squash a little bit when rebasing, but the conflicts were minimal.

ok that’s good

the paste wasn't working sooo

not true

ok maybe true

I must be quick before it expires