Hello, I am trying to make a Monero Flutter plugin so that monero wallets can more easily be made in flutter projects.

I am able to make wallets, restore them, and send transactions so far, but I have run into a pretty serious native crash.

When I restore a wallet sometimes it will crash with the following error.

*** *** *** *** *** *** *** *** *** *** *** *** *** *** *** ***

Build fingerprint: 'google/taimen/taimen:11/RP1A.201005.004.A1/6934943:user/release-keys'

Revision: 'rev_10'

ABI: 'arm64'

Timestamp: 2022-05-05 13:29:07+0800

pid: 4323, tid: 4415, name: DartWorker >>> com.example.flutter_libmonero_example <<<

uid: 10435

signal 11 (SIGSEGV), code 1 (SEGV_MAPERR), fault addr 0x0

Cause: null pointer dereference

    x0 00000077dc66e27c x1 0000000000000000 x2 0000000000000020 x3 00000077dc90dfc0

    x4 0000000000000020 x5 00000077dc66e29c x6 6ba45c57ecedc839 x7 7b693b97342475af

    x8 000000000029fdc3 x9 000000000000001f x10 0000000000001b08 x11 00000077dc30e040

    x12 f23b9fa7f5b2b920 x13 5ab56bcd7a095991 x14 0018b6584180a81a x15 000000000a516280

    x16 00000077f1367f90 x17 0000007b077e8000 x18 00000000000bb9d5 x19 0000000000000020

    x20 0000000000000000 x21 0000000000000000 x22 00000077f1c6fc28 x23 0000000000000020

    x24 00000077f1c71000 x25 00000077f12e9578 x26 00000077f12e9328 x27 00000077f12e94c8

    x28 00000077f1c6fc10 x29 00000077f1c6fab0

    lr 00000077f0844d00 sp 00000077f1c6fa80 pc 0000007b077e7eb0 pst 0000000080000000

use a pastebin

I can trace the error back to the wallet2::get_cache_file_data function, and I think it is occuring at github.com/monero-project/monero/bl…master/src/wallet/wallet2.cpp#L5802

it seems that in memcpy it tries to derefrence a null pointer, when the wallet is being serialized which causes the crash, however I dont really have an idea why this would happen.

alcapaca: Try building with -O0 -g and getting a stack trace.

Next dev meeting: monero-project/meta #703

If you can't get a stack trace, printf all over the place repeatedly is what I do.

I have got a question about the upcoming Monero view tags

If you have the wallet A, and you want to transfer money to wallet B, the view tag included in the tx is A's or B's or both?

And wouldn't it decrease privacy?

> <@xxfedexx:matrix.org> I have got a question about the upcoming Monero view tags

> If you have the wallet A, and you want to transfer money to wallet B, the view tag included in the tx is A's or B's or both?

> And wouldn't it decrease privacy?

Not a topic for this channel, but hopefully this answers your Qs: localmonero.co/knowledge/view-tags-reduce-monero-sync-time

Thanks, but it says:

> add a 1-byte “tag” to each transaction using a shared secret known only to the sender and receiver of that transaction.

That secret gets regenerated on each transaction with the wallet keys?

* Thanks, but it says:

> add a 1-byte “tag” to each transaction using a shared secret known only to the sender and receiver of that transaction.

That secret gets regenerated on each transaction with the wallet keys?

How can the receiver know it too?

* Thanks, but it says:

> add a 1-byte “tag” to each transaction using a shared secret known only to the sender and receiver of that transaction.

That secret gets regenerated on each transaction with the wallet keys?

nevermind, the localmonero post is not clear but the github issue monero-project/research-lab #73 explains better

Hello. Do I need to add the p2pool module for flatpak?

BigmenPixel[m]: the problem is that the way p2pool is implemented it tries to install the binary in the same directory as the GUI

package managers don't like this

hyc: my gitian builds are running but will take a couple more hours, i don't have anything cached

selsta: Then I can put the binary file there in advance?

might work

selsta: Thanks.

selsta: ok

<selsta> "might work" <- Sorry for stupid question.)) Is this p2pool necessary?:

that's the software, yes

> <@bigmenpixel:matrix.org> Sorry for stupid question.)) Is this p2pool necessary?:

> github.com/SChernykh/p2pool

No

🤔

is it 100% necessary to integrate in all distributions of monero-gui, no

selsta: In flatpak, can’t integrate it manually without rebuilding, so it’s better to do it.

if you can easily integrate it then good, otherwise I'll add a compile time flag to disable p2pool for package managers

selsta: Thank you for the second time. :)

Btw, any chance that the rpc_pay feature gets "rewritten" to use p2pool instead of solo mining? As in, the client mines p2pool shares on behalf of the node, rather than full solo blocks

Fits much better with the "small but consistent payouts" model

(Perhaps point them directly to p2pool-mini even, rather than the main p2pool)

wouldn't that require having p2pool in the daemon, or somehow routing the rpc_pay stuff to a p2pool daemon?

merope: the problem is that there is always incentive to run free of charge nodes for malicious actors, they need the data and not the hashrate

oh I don't mean that as a solution for the malicious node problem, it was an unrelated suggestion about promoting rpc-pay

and as long as there are free nodes people are unlikely to use rpc-pay

<merope> "Fits much better with the "small..." <- p2pool is just an illusion of "small but consistent payouts"

ooo123ooo123 not the place, and also wrong

what i meant, as it currently is i don't see rpc-pay being adopted

(even if it somehow uses p2pool)

Right - but think of it in terms of "send a tip to your favourite public node to support them". Most people don't do that now, because the odds of finding anything are so low that it's not even worth their effort to figure it out

But now that we have the p2pool integration, the odds of finding something on p2pool mini are significantly higher

merope: how much significantly in numbers ?

relatively to solo mining

Main chain diff is ~300G these days, while p2pool mini typically hangs around 100M-200M

so 1000x-3000x better odds of finding something

Of course, the payout is proportional - but it's something

do you consider 1e-15 as something too ?

can you just use numbers instead of words ?

Irrelevant question, but yes

they are ambigious

endor00[m]: There are no numbers in p2pool README too, only words "more frequent payouts"

Therefore it's an illusion

edit: it uses word "Regular"

How do p2pool payouts compare to the differential fee from spending a p2pool output?

Fees are rather high, id say it's also rather easy to figure out the true sender

Since p2pool addresses are public, you could cross reference them with transactions that have tons of inputs. Unless your ring includes some other people miner address sending a large transaction it would be very easy to pinpoint you as the real sender, right?

s/people/p2pool/

<merope> "so 1000x-3000x better odds of..." <- Are you sure that 1000x-3000x multiplier is applicable to equal amount of "something" from pool mining / solo mining / p2pool mining ?

Yes

merope: You're good at math then

<merope> "Yes" <- Can you help me to calculate it ? I believe it's at most 12x

relatively to solo mining

This is not the right channel for this discussion

Do these cmake flags look correct for building static binaries?... (full message at libera.ems.host/_matrix/media/r0/do…1c667cbd256e84b98ab4ebfb453ff86181a)

From github.com/sethforprivacy/simple-monerod-docker/blob/main/Dockerfile

> <@sethforprivacy:matrix.optoutpod.com> Do these cmake flags look correct for building static binaries?... (full message at libera.ems.host/_matrix/media/r0/do…0009cc9ee3d7497ce7c99252919e5a25949)

yes, use depends like here github.com/monero-project/monero/blob/master/Dockerfile#L31

STATIC=ON will build most libs statically, if found.

Note "if found".

Though not giving you a static binary, pedantically.

Yeah, `file monerod` is still showing dynamically linked

I want a unified Dockerfile for multiple arches so depends is not ideal...

It's the best you're gonna get. Just make sure you don't have stuff you'd think should be static in ldd output.

Thanks

Like 0mq etc.

aren't all the dependencies in our depends builds static?

that's what gitian is for, setting up the env so nothing else is present

my daemon on master branch had a segfault but ulimit -c was set to 0 (even though it was set to unlimit from the last segfault..)

so I wasn't able to find a core file

have it running in gdb now

that's too bad. I've been running on master from (v0.17.0.0-67e5ca9ad) for ~9days now

so far all hashes match but still 3 operating systems to go to go

cool

if there were any timestamp discrepancies they'd certainly show up by now, anyway

btw all testnet node operators should compile master so that they end up on the correct chain

whens that fork?

less than a week\

Testnet hard-fork: May 16th, 2022, block 1982800 and 1983520

Stagenet hard-fork: July 9th

Mainnet hard-fork: July 16th, 2022, block 2668888

ok, just compiled master and running testnet on xmrchain

hmm, need to spin up testnet p2pool by May 16th

lol cookie

Will it be tagged before then? I'll push a Docker build as well once there is a tag, or I can do a special one if not.

no tag

selsta: what kind of time-line should I aim for with the 'good to haves' remaining here monero-project/monero #8237 ?

I don't know, I don't see much point in not merging it just because some arbitrary dated passed.

The current multisig is broken anyway so as long as we didn't tag it should be fine.

well it looks like you are wanting to tag in 5 days, when would the next tag be? or maybe I'm missing something

no we won't tag in 5 days

testnet hardfork doesn't require a tag

stagenet then? 21 days?

wait that's a month

yep

plan is to release ~1 month before mainnet hardfork

ok I will try to get them PR'd by next monday

or the first one anyway (and third one rebased)

this coming monday*

this is about the ccs: i accidentally upvoted some ccs proposals with my gitlab github account eventhough i already upvoted them with my gitlab only account, I think i reverted everything but in case i missed one i wanted to make it clear that im not trying to manipulate any votes

.merge+ 8312

Added