-
monerobull[m]
epic
-
ErCiccione
> <@escapethe3ra[m]:libera.chat> > <@erciccione[m]:libera.chat> Network upgrade (hard fork on mainnet): 16th July, Block 2668888
-
ErCiccione
>
-
ErCiccione
> is this final? want to post a short report
-
ErCiccione
I'd say so, yes
-
merope
Is the hardfork going to be a "single" fork or "double" fork with a 720-block transition?
-
moneromooo
Double.
-
merope
I assume the second fork height will be 2668888 + 720, right?
-
moneromooo
Not set yet.
-
merope
Alright
-
lousiF
-
rbrunner
Say again, how much will fees typically rise after the hardfork? In what ballpark is the factor of the rise?
-
rbrunner
I know I read the answer at least once somewhere, but could not find it again even with quite some effort
-
moneromooo
Restart with --offline --fixed-difficulty 1 after changing the HF table to switch to the next version on the next block. Mine a few blocks, then bc_dyn_stats 1.
-
rbrunner
Oh, that's so much work, and I am so lazy :)
-
rbrunner
Maybe it's really time to do this, also the question of transaction size is interesting and could be checked first-hand with this
-
moneromooo
I think it's at least 5x. Might be a fair but more though.
-
sech1
It is 5x
-
moneromooo
ty
-
rbrunner
Ok, thanks to both. That will need some gentle introduction to not put people into shock. E.g. by showing how low they will *still* be, even if 5 times higher
-
rbrunner
But not a dev problem :)
-
moneromooo
Those people are the reason I'd bump by another x2...
-
w[m]
30x, am I wrong?
-
w[m]
* I thought it was 30x. Am I wrong?
-
UkoeHB
no 5x sounds right
-
sech1
new code just removes "fee /= 5" from the code
-
sech1
so it's 5x more, not accounting for fee algorithm changes
-
moneromooo
That sounds... odd :D
-
jkirsch
Hey all. Can anyone help me with this information: When a node requests a seed peer list from a seed node, what is the criteria the seed node uses for choosing which nodes to include in its response? ie, Which nodes are sent back by IP address x if I launch the daemon with --seed-node x ? It doesn't seem to be either node x's white list top 250 or
-
jkirsch
the entire white list.
-
moneromooo
You can't request a seed peer list. I assume you mean a peer list ?
-
moneromooo
If so, then there are a few steps IIRC:
-
moneromooo
- take the white peers
-
moneromooo
- remove the ones that were already sent to that peer
-
moneromooo
- shuffle the remaining list
-
moneromooo
- send the first N in the list
-
moneromooo
This is from memory, it might be wrong, but it's the gist of it. See p2p_node.inl I think.
-
moneromooo
I think you can grep Yao to get to the code that does it.
-
jkirsch
Great that's very helpful thanks, I'm trying to figure out why the peer list the seed sends back may contain over 1,000 entries (sometimes over 2,000 :O )
-
rbrunner
Hmmm, my testnet daemon does not want to hardfork after I added two more entries in the "hardforks.cpp" 'testnet_hard_forks' array for 15 and 16 and mined the necessary blocks.
-
moneromooo
You might have found an asshole node, which tries to stuff your peer list with malicious nodes.
-
moneromooo
Or a bug.
-
rbrunner
What is the most probable stupid mistake I make?
-
moneromooo
The heights are wrong.
-
moneromooo
You're running the old binary.
-
rbrunner
I did not compile :)
-
rbrunner
No, I believe to have checked that all already. Hmmm.
-
rbrunner
Does it tell me something useful if the command "hard_fork_info 15" outputs: version 15 not enabled, 0/10080 votes, threshold 0
-
moneromooo
Try hard_fork_info 115
-
jkirsch
Very interesting. Across the ~4,200 unique IPs on my 11 nodes' white lists, ~2,200 responded with >1,000 nodes when I used them to seed. Gonna investigate.
-
rbrunner
Same result
-
moneromooo
Then it's not useful :)
-
moneromooo
Try hard_fork_info
-
rbrunner
Yup.
-
rbrunner
In any case, I don't have to "enable" hardforks myself, do I? The array entry should be enough, if done properly.
-
moneromooo
There's a number of asshole nodes. They're basically either some known jerk with a massive chip on the shoulder, or one of probably many spy companies.
-
rbrunner
Simply "hard_fork_info" gives me the expected 14: version 14 enabled, 10080/10080 votes, threshold 0
-
moneromooo
Does "status" tell you, forking in xx ?
-
rbrunner
I didn't do a "status" before I started to mine. Maybe I drop blocks and try again.
-
rbrunner
No, with 48 blocks below the supposed forkheight "status" does not tell me something. I think I recompile that bloody thing from scratch.
-
rbrunner
I think I found it: If you don't adjust the timestamps as well in that array so that they are in the future, it somehow might not work. Anyway, on v16 now.
-
moneromooo
Odd. It should insult you about it if you don't.
-
moneromooo
Check log again ?
-
sethforprivacy
Created a check-list for the v15 fork now that the height has been set:
-
sethforprivacy
-
sethforprivacy
Please watch that for updates and keep the items on it in mind as we get closer to tagging/testnet fork date.
-
rbrunner
I have now master plus BP+ plus higher fees plus ringsize 16 running; only view tags are missing, but their influence should be minuscule
-
rbrunner
I did a 1 in / 2 out tx and compared it with other such txs on the testnet blockchain
-
rbrunner
Result: Sizes 1526 versus 1417 bytes = 7.7% larger
-
rbrunner
Fee: 0.000077830000 versus 0.000014620000, about a factor of 5.3 higher.
-
moneromooo
Why is is larger ?
-
rbrunner
Higher ringsize?
-
rbrunner
16 instead of 11
-
moneromooo
Oh.
-
moneromooo
Thanks :)
-
rbrunner
I think BP+ mitigates quite a bit, but not completely.
-
moneromooo
Yes, you had said it. I read too fast -_-
-
rbrunner
Maybe I was the first one to make such a transaction? I am so proud :)
-
UkoeHB
lol congrants rbrunner
-
rbrunner
Gee, thanks.
-
UkoeHB
was the fee pr merged already? I can't keep track
-
rbrunner
No, I had to cherry-pick. The ringsize PR as well. Fortunately everything rebased already.
-
rbrunner
A true frankenstein release.