-
ndorf
am i alone in the opinion that there should be some easy daemon command that includes the age of the top block in its output?
-
ndorf
i'd suggest `diff`, since that one already includes the top block hash
-
luigi1111w
status ?
-
ndorf
it seems that it goes hand in hand with the top hash, which is not in `status` already, and adding both might be too much?
-
ndorf
`diff`, while perhaps unfortunately named, includes the top hash already
-
luigi1111w
why is hash better matched than height?
-
moneromooo
Age is not always super useful since depending on when you start your daemon, it might be either "when you received it" or "whatever timestamp the miner put".
-
moneromooo
The "jitter" will be large compared to block time.
-
moneromooo
It might be useful if it's really old, like 20 minutes ago, so that the jitter decreases relatively. But then you have a notifier for that IIRC.
-
moneromooo
That said, if you want to add "time of last block" in get_info, I'm fine with that.
-
moneromooo
Oh, in the console, not RPC. Then I guess you'll know when you started, so the difference between received and timestamp doesn't matter that much. The diff command seems fine.
-
moneromooo
Oh, I might have that already actually, I made a patch to display recent block rate probabilities.
-
ndorf
luigi1111w: `diff` includes both, height and hash
-
moneromooo
Ah, not quite. You ask "how many blocks did I get in the last N seconds"...
-
luigi1111w
ok. fine by me
-
ndorf
moneromooo: is that going to be a new command or also a `diff` change?
-
moneromooo
It is a separate command. Not sure why I didn't PR it. Maybe it doesn't work...
-
moneromooo
Or maybe it wasn't useful enough.
-
ndorf
ah. ok, thanks
-
ndorf
FWIW it seems a pretty compact thing compared to add to `diff` relative to what's there already, e.g. 'AGE: 1m30s'
-
vtnerd
iDunk the ASM cannot work in windows, they have to be re-written to support the windows call convention. every other OS uses the same call convention / ABI
-
vtnerd
the `if (UNIX && ..` might pass for true within a cygwin environment, I was testing via linux cross-compile
-
hyc
reverse the test, do if (_WIN32 ... )
-
iDunk
With that test, supercop code is not selected in neither MSYS2 nor Ubuntu (mingw cross-compiler).
-
iDunk
If I remove UNIX from the test, amd64-64-24k is selected, but results in non-working code for reasons mentioned earlier by vtnerd.
-
hyc
well yeah, you still need ABI-compliant function prologues and epilogues
-
xmr-pr
Mladia opened issue #6748: Cannot sync up with blockchain
-
xmr-pr
-
KnifeOfPi
hello
-
KnifeOfPi
we have a user with some very strange issues
-
KnifeOfPi
everywhere he's loaded his wallet - including the CLI - it's shown transactions that don't actually exist
-
KnifeOfPi
-
KnifeOfPi
56db0b493add26a1130be63a27c7da085070dcf38ba2582e37e669da0eba9214
-
KnifeOfPi
he did not ever use his own node
-
KnifeOfPi
moneromooo, are you able to help?
-
KnifeOfPi
-
moneromooo
So.. a stranger's node ?
-
moneromooo
Maybe the node lied. Put dummy data as txids.
-
moneromooo
Or maybe there is a bug.
-
KnifeOfPi
it's moneroworld node, and he's used multiple different nodes
-
KnifeOfPi
cake wallet official node reported the same results from his cake wallet
-
moneromooo
Is that user willing to apply log patches and rescan ?
-
KnifeOfPi
i doubt it, he is really not a technical user, and i obviously can not ask for his seed and do it myself
-
moneromooo
Well, dunno then.
-
KnifeOfPi
if you would explain to me how to do that, i can possibly try, but it's gonna end up being a game of telephone
-
moneromooo
Once I give the patch:
-
moneromooo
- save it to /tmp/patch
-
moneromooo
- cd /path/to/monero/tree
-
moneromooo
- patch -p1 < /tmp/patch
-
moneromooo
- make
-
KnifeOfPi
he is running windows and there is really no way i can instruct him on how to do something like that.
-
selsta
we can apply the patch and create bins
-
selsta
but then he has to trust our binaries
-
moneromooo
Oh good point, github makes test binaries.
-
selsta
github does not currently, but I can make them or someone else can
-
moneromooo
I'll make a patch then.
-
moneromooo
I assume this is deterministic, right ? Always the same txid at the same hieght ?
-
moneromooo
Or at least same height ?
-
KnifeOfPi
Yes, it's a few different txid's. He sent these transactions and none of them were received but they show up whenever he restores; they still don't appear on the blockchain.
-
moneromooo
Do you know the daemon url that got used ?
-
KnifeOfPi
node.moneroworld.com:18089
-
KnifeOfPi
that is the one used with his CLI
-
selsta
moneroworld is a node aggregator, so it could be any node
-
selsta
but cake wallet node had the same issue so most likely not a lying node?
-
moneromooo
Are the amounts as expected ?
-
KnifeOfPi
the amounts are what he sent, yes
-
moneromooo
Including the incoming one ? Check the ts of that block (which isn't the one on the chain for the same height).
-
KnifeOfPi
the heck
-
KnifeOfPi
maybe it's caused by a restore from wrong height????
-
moneromooo
That would caused missed txes, and change being seen as receive. Not bad txids.
-
KnifeOfPi
what could be causing it to restore the same bad txids every time?
-
moneromooo
The daemon being malicious.
-
moneromooo
gingeropolous: how many daemons in your aggregator ?
-
moneromooo
And did that number go up a lot recently ? :)
-
KnifeOfPi
it would require for every node there to be malicious, run the same malicious code, and also for our cake node to be affected
-
moneromooo
Oh, your node was used, so likelyknow good ?
-
iDunk
Have him check currency format in regional settings. Idk if Monero CLI uses it, but balance looks weird there.
-
moneromooo
You think that could alter printing txids ?
-
iDunk
I've seen weird shit in Windows caused by things I would never have guessed.
-
vikrants
If our node was bad, we’d see 100s or thousands of these cases no?
-
moneromooo
Are you guys the same team ?
-
moneromooo
I mean, one case of bug, or two ? :)
-
vikrants
One guy .. multiple transactions
-
vikrants
Knife and i talking about the same thing
-
moneromooo
Do you have the daemon that got used ready ?
-
KnifeOfPi
what do you mean?
-
moneromooo
Can you run console or RPC commands on that particular daemon ?
-
moneromooo
(the one you're sure got used)
-
vikrants
I’ve asked Mykola to come on
-
vikrants
He’s getting on here
-
moneromooo
Already there.
-
caralho
My does the car think it gets to eat just because somebody else eats?
-
caralho
Cat*
-
KnifeOfPi
w a t
-
moneromooo
Well, why would it be any different ? Might as well try, righ t?
-
mordsid
what exactly to do ?
-
moneromooo
I'm not sure how complex cat thought processes are, but "food is here" and "I'm hungry" definitely trigger "let's eat!".
-
moneromooo
mordsid: Can you run console or RPC commands on that particular daemon (the one you're sure got used) ?
-
KnifeOfPi
lol
-
mordsid
yeah i can run
-
moneromooo
print_block 2139310
-
moneromooo
The log patch, to apply to the wallet rescanning:
paste.debian.net/hidden/efe50762
-
moneromooo
simplewallet must then be run with --log-level 0.
-
selsta
is this on top of master or release branch? or should it not matter?
-
moneromooo
Hmm. It's on top of some random branch, let's check...
-
selsta
I will add it on top of release
-
selsta
manually worst case :P
-
moneromooo
19th of july, pretty recent.
-
moneromooo
Is it known what version the buggy wallet was running ?
-
KnifeOfPi
0.16.0.1
-
moneromooo
Then apply on top of that please.
-
KnifeOfPi
We will need a compiled version if that is ok.
-
moneromooo
If someone's feeling nice.
-
selsta
yes, I will do it
-
moneromooo
^ someone nice
-
mordsid
-
moneromooo
Looks like mine. Thanks.
-
moneromooo
Running a scan with the patch above against that daemon would be nice, whether the original reporter or not. Can be Linux too.
-
moneromooo
If it reports the right txid, then it must be run by the reporter to. But maybe it'll show a bad txid already.
-
moneromooo
Rescanning for ~2k blocks before 2139310 is ok.
-
moneromooo
(ie, not from genesis)
-
moneromooo
I just want to see the logs from that patch. Remember --log-level 0.
-
selsta
windows binary will probably take half an hour KnifeOfPi
-
KnifeOfPi
this is fine, i have an exam in 2 hours so as long as its before then we are probably good
-
selsta
mordsid: what is the cake node url?
-
mordsid
exactly this one is xmr-node-uk.cakewallet.com:18081
-
KnifeOfPi
yes
-
KnifeOfPi
so - will the user have to use a different wallet-cli or is everything done on the daemon side?
-
selsta
yes
-
selsta
a different wallet-cli
-
selsta
with the patch applied
-
KnifeOfPi
ok
-
iDunk
(The first patch)
-
» iDunk hides
-
KnifeOfPi
lol
-
moneromooo
Actually, I'll add some more logs if that's no bother...
-
KnifeOfPi
ok
-
moneromooo
(I've run the patch here and I get the expected data)
-
selsta
didn’t start the build yet anyway
-
KnifeOfPi
just to be clear, does the node also require a patch?
-
selsta
no
-
KnifeOfPi
ok
-
moneromooo
selsta:
paste.debian.net/hidden/7a12f102 if you don't mind :) Thanks
-
moneromooo
Hmm. monero-wallet-cli links against libpython for some reason :S
-
moneromooo
Not emacs yet I think...
-
iDunk
Not here
-
selsta
hmm can I select a branch with gitian-build.py?
-
selsta
I thought this was possible by simply specifying it instead of version number
-
iDunk
Yes, it's the last item
-
iDunk
But it's the remote branch, IIRC.
-
iDunk
I can fire up a depends build quickly, all my deps are already built.
-
selsta
-
selsta
this would be the commit
-
selsta
I did this before but I can’t get it to work now... hmm
-
selsta
./gitian-build.py --setup --docker --url
github.com/selsta/monero selsta cake
-
moneromooo
I've tried a restore with the patch and the daemon above, and I get correct results. So it'll have to be run by the reporter.
-
selsta
ok solved it, looks like gitian-build.py does not like a different URL after already setting up a different one
-
selsta
KnifeOfPi: uploaded the monero-wallet-cli here (with signed hash)
gui.xmr.pm/files/cli/cake-issue-tmp
-
selsta
they have to run the cli with --log-level 0
-
selsta
and also make sure to use the xmr-node-uk.cakewallet.com:18081 node
-
KnifeOfPi
thank you selsta, very cool
-
KnifeOfPi
i will let you know what happens
-
selsta
and then please send the logs to mooo
-
moneromooo
ty selsta
-
dEBRUYNE
Anyone an idea here? If I recall correctly, the spend proof should still be available even without the private transaction key, as it basically simply proves you signed the rign
-
dEBRUYNE
