-
br-m
<rbrunner7> Meeting in 1 hour
-
uncle_rae
hurray \o/
-
br-m
<rbrunner7> Meeting time. Hello!
monero-project/meta #1471
-
br-m
<sneedlewoods_xmr:matrix.org> Hey
-
br-m
<jeffro256> Howdy
-
br-m
<rucknium> Hi
-
selsta
hi
-
br-m
<rbrunner7> Let's start already with the reports from last week.
-
br-m
<rbrunner7> I reviewed the PR for Polyseed in the GUI from selsta yesterday. Nice.
-
br-m
<sneedlewoods_xmr:matrix.org> More fixes for #9464, also rebased and fixed conflicts for: #10232, #10233 and #10819, but I plan to push one after the other and see what issues come up through AI reviews, to prevent getting duplicates.
-
br-m
<rbrunner7> That's this one, in case anybody wants to have an early look:
monero-project/monero-gui #4729
-
selsta
v0.18.5.3 got released, no bug reports so far
-
br-m
<rucknium> I PR'ed this to defeat the Nyx eclipse attack:
monero-project/monero #11527
-
br-m
<rbrunner7> According to plan that's the last 18.x release, right?
-
selsta
yes
-
br-m
<rbrunner7> Splendid
-
br-m
<rbrunner7> That might be the longest and most detailed Monero PR description ever, @rucknium:monero.social :)
-
br-m
<sneedlewoods_xmr:matrix.org> IMO the description is very helpful
-
br-m
<rucknium> It keeps going if you follow the link at the bottom :D
-
br-m
<rbrunner7> Might well be this will stop some similar attacks right in their tracks, without us knowing.
-
br-m
<rucknium> @rbrunner7:monero.social: I have a small question for you about a line of code you wrote for the subnet deduplication PR. Whenever you have time.
-
br-m
<rucknium> A lot of attacks use the small white_list limit as a starting point.
-
br-m
<rbrunner7> You can drop it here, or in a DM, and will look at it in a quiet moment in the next few days.
-
br-m
<rucknium> Ok. I will ask at the end of the meeting.
-
br-m
<jpk68:matrix.org> Hello, sorry for being late ;)
-
br-m
<jpk68:matrix.org> I forgot what I did this week
-
br-m
<rbrunner7> Stressnet is still pretty smooth sailing?
-
br-m
<rbrunner7> Maybe 10 different things. Happens to me sometimes in the day work, and then I don't know how to report hours done :)
-
br-m
<rbrunner7> Alright, looks like we are already through with the reports. Anything else to discuss beyond those?
-
br-m
<sneedlewoods_xmr:matrix.org> nothing from my end
-
br-m
<rbrunner7> Doesn't look like it. So, short meeting, thanks everybody for attending nevertheless, the Monero adventure continues, and read you again next week!
-
br-m
<sneedlewoods_xmr:matrix.org> thanks everyone
-
br-m
-
br-m
<rucknium> The bot review (
xmrack/monero-review #1088) says that filtered.insert(filtered.begin(), peer); is less efficient than filtered.push_back(peer);. The efficiency would matter more after the total size of the white_list would increase.
-
br-m
<rucknium> Is there an unsaid reason why you used filtered.insert() instead of filtered.push_back()? I noticed that mooo's pruning PR used filtered.push_front() in that spot, before you cleaned up that area of code.
-
br-m
-
br-m
<rucknium> And I don't know why push_front() was used in that spot (which is used only when we have pruning enabled) instead of push_back().
-
br-m
<rucknium> Whatever it was trying to do, the order opf placing peers into the container doesn't matter with my PR since I always take a uniform random draw instead of the get_random_index_with_fixed_probability() probability mass function that drew more heavily from recently-seen peers.
-
br-m
<rucknium> I think I am safe to replace filtered.insert(filtered.begin(), peer); with filtered.push_back(peer); there, but I wanted to check.
-
br-m
<rbrunner7> Hmm. I don't really remember, but pretty sure I just stayed with the logic, assumed that there must be reason to insert in front without getting to the bottom of it, and did likewise. Wondering a bit however why I did not also use push_front ...
-
br-m
<rbrunner7> You explained something with a fence recently if this happens, right?
-
br-m
<rucknium> I think Chesteron's Fence logic encouraged you to copy the code logic as before. Now that I am changing the logic upstream, the effect of the fence is different. Or it is better to say that the fence has no effect since it's a uniform random draw.
-
br-m
<rbrunner7> Your explanation that after your changes it doesn't matter anymore anyway sounds logical, so now we will probably never know :)
-
br-m
<jpk68:matrix.org> Fun fact: épée is a term in fencing (get it, epee?)
-
br-m
<rucknium> Ok. Sounds good to me. Keep in mind that I am not very good at reading C++, and even worse at writing it :)
-
br-m
<jpk68:matrix.org> Just since we were talking about fences :P
-
br-m
<rucknium> Actually it is the fact that I am so bad at writing C++ that this PR is so simple. The simplest solution is implemented because I can only implement simple solutions :)
-
br-m
<rbrunner7> Well, if it does the job, which is sure as well as it gets after all your simulations, nothing to worry about
-
br-m
<rucknium> Definitely an exemplar of Grug Brained development
grugbrain.dev
-
br-m
<rbrunner7> Uff, that the hell is that? :)
-
br-m
<rbrunner7> I am afraid that this dev retired recently, swept away by coding LLMs ...
-
br-m
<rbrunner7> Which pretty much trample on all those nice "leasons learned"
-
br-m
<slowbeardigger:matrix.org> @rucknium: Omg