-
jakoson0
Did you know that all witdraw-buyer-seller-depoist chains are trackable in Monero? No? You should have read Breaking Monero. How many people are you endangering with your 'privacy' coin?
-
ArticMine
I updated issue 70 with the new proposal for scaling
-
ArticMine
-
sgp_
ArticMine: can you describe the dynamic minimum penalty free zone
-
UkoeHB__
thanks ArticMine, I will take a look and formulate a response over the next few days
-
ArticMine
sgp In a nutshell. The long term median is calculated first using fixed minimum penalty free zone of 300000 bytes. Then short term median is calculated using the long term median as the penalty free zone. So the long term median becomes the dynamic penalty free zone
-
sgp_
ArticMine: can you repeat that please? users are having issues with irccloud today
-
ArticMine
Thanks UkoeHB
-
ArticMine
Sure
-
ArticMine
<ArticMine> sgp In a nutshell. The long term median is calculated first using fixed minimum penalty free zone of 300000 bytes. Then short term median is calculated using the long term median as the penalty free zone. So the long term median becomes the dynamic penalty free zone
-
ArticMine
Meeting?
-
sgp_
.time
-
sgp_
sorry, I was off by an hour
-
sgp_
meeting pushed to 18 utc
-
sgp_
in 35 mins
-
ArticMine
ok
-
sgp_
sorry :/
-
sgp_
okay, meeting time
-
sgp_
these are issues with irccloud today, so many people will pop in and out
-
sgp_
the matrix relays appear to be working however
-
sgp_
-
sgp_
1. Greetings
-
ArticMine
Hi
-
sgp_
ArticMine sarang moneromooo knaccc dEBRUYNE vtnerd
-
knaccc
howdy
-
UkoeHB
hi
-
sgp_
pretty quiet crowd, but it normally picks up as the meeting goes on
-
sethsimmons
Hi all
-
sgp_
were there any outstanding items on the triptych discussion from the last two weeks?
-
ArticMine
The question I saw came down to timing
-
sgp_
looks like there aren't any outstanding items at this time, except to find out who will work on it
-
sgp_
my understanding is that this is waiting on optimizations and some consensus decisions from sarang (eg: related to output selection)
-
sgp_
unless someone else is willing/able to work on these
-
ArticMine
... and the issue of multisig
-
sgp_
are there any multisig decisions that need to be made?
-
UkoeHB
whether it must be supported if triptych goes live
-
ArticMine
It is more a question on how multisig will be handled and the time required
-
UkoeHB
yeah it sounds like a big project
-
ArticMine
It raised issue around changing the address format to ensure compatibility
-
ArticMine
I agree with UkoeHB it is a big project to deal with all the details and ramifications
-
sgp_[m]
Yeah, it's a big project
-
sgp_[m]
I don't think we should change address formats, but it's a potential option
-
ArticMine
It may be necessary to support multisig
-
sgp_[m]
In what way is changing address formats necessary?
-
ArticMine
It I recall it had to do with preventing the sending to
-
sgp_[m]
The legacy multisig "conversion" process is the other option
-
sgp_[m]
I lean towards that actually
-
ArticMine
of new format transactions to old addresses
-
ArticMine
My take on this is that we should not hold up BP+, CLSAG ring changes and Issue 70 over this
-
sgp_[m]
Yes, to spend triptych outputs sent to legacy multisig wallets, they need to be "converted" to a non-multisig wallet before they can be spent iirc
-
sgp_[m]
Is there anything new on triptych or should we talk about 70?
-
ArticMine
So one needs to prevent Triptych outputs from being sent to legacy multisig wallets
-
ArticMine
hence the address format change
-
sgp_[m]
ArticMine: not necessarily, those changes could cause more problems than they solve
-
sgp_[m]
I'd rather defer that conversation topic however and talk about 70
-
ArticMine
This is of course valid and in need of further dicussion
-
ArticMine
Sure
-
sgp_
-
sgp_
you have the floor articmine
-
ArticMine
In a nutshell we need to prevent drastic increases in fees
-
ArticMine
With the current situation if we had say 100x the level of adoption
-
UkoeHB
I think we discussed this last week, triptych multisig requires no conversions or different address formats. The main problem is it requires implementing complicated new crypto
monero-project/research-lab #72
-
UkoeHB
although if multisig isn't supported then sending triptych outputs to a multisig address is bad
-
sgp_[m]
ArticMine: I'm worried fees are way too low right now honestly
-
ArticMine
or more. Then a disruption to the Monero network either external or the result of an attack would cause a drastic fee increase
-
ArticMine
My proposal by the way has a fee icrease about 5x for the min fee
-
sgp_[m]
Okay, so getting rid of the current min tier effectively
-
ArticMine
Yes
-
sgp_[m]
So if I can try to summarize changes:
-
ArticMine
The min fee is set to the minimum fee that support scaling
-
ArticMine
THere are 3 critical changes
-
sgp_[m]
1. Higher initial minimum fee, but lower fee "slope" as transaction counts increase?
-
ArticMine
Correct
-
ArticMine
So at 1425000 bytes the min fee equals the current formula
-
ArticMine
... but the spam cost is higher
-
ArticMine
than now
-
ArticMine
The main consensus changes are
-
sgp_[m]
Is that crossover point ~10x increase in transactions compared to now?
-
sgp_[m]
(Roughly)
-
ArticMine
more like 30x or more
-
ArticMine
Give me a sec
-
sgp_[m]
How did you choose the "slope?"
-
sgp_[m]
Based on expected market cap?
-
ArticMine
It follows the minimum fee that allows for scaling
-
ArticMine
So it follows the inverse square of the long term median
-
ArticMine
So this is the natural slope
-
ArticMine
Back to blocksize
-
sgp_[m]
You changed a variable though right? 1.4 to 2?
-
ArticMine
currently the average is about 40k
-
ArticMine
Yes that is the consensus part
-
ArticMine
It is needed to allow for the current rate of growth
-
ArticMine
so the long term median does not fall behind
-
ArticMine
even at maximum
-
dEBRUYNE
Here, mostly reading
-
ArticMine
The other 2 consensus changes are
-
sgp_[m]
I'm a little skeptical about the 1.4->2 change in particular
-
ArticMine
2) putting a floor under the long term median for decay
-
ArticMine
This is critical
-
sgp_[m]
Is there currently no floor? No max rate of decrease?
-
ArticMine
since we cannot have a long term median that ahs been built up over years falling in just over 2 months
-
ArticMine
none
-
ArticMine
the other is making the minimum penalty zone dynamic
-
ArticMine
by setting it to the long term median
-
ArticMine
The rate increase from 1.4 to is needed just take a look at the growth in adoption
-
sgp_[m]
Okay, I think I am following on how it is dynamic
-
ArticMine
We cannot stymie Monero like Bitcoin
-
ArticMine
So there needs to be room on the max growth not be set so tight that say events like 2016 would cause problems
-
sgp_[m]
You're allowing the minimum penalty free zone to grow
-
ArticMine
Yes
-
ArticMine
The prevents a sharp drop in the short term median
-
ArticMine
and prevents a sharp increasein fees
-
sgp_[m]
Do you have a table where you play this out?
-
ArticMine
I am not sure
-
ArticMine
I have soem fe examples
-
ArticMine
fee
-
sgp_[m]
Sure what are those
-
ArticMine
I am looking at all the way from 300000 bytes to 1500000 bytes
-
sgp_[m]
Okay, so with some form of penalty under current circumstances
-
ArticMine
The penalty formula does not change
-
sech1
Will we have to say bye to our current 0.2 cent fees?
-
ArticMine
Yes they will become 1 centUSD
-
ArticMine
but if adoption increase theny can come back
-
ArticMine
they
-
sech1
But what if we just scale everything down 5x after these changes?
-
ArticMine
or 4x
-
ArticMine
The low fee will not scale
-
ArticMine
just as it does not scale now
-
sgp_[m]
You say the penalty formula doesn't change; I don't really get that
-
ArticMine
but there are many in the community that are concerned about the very low fee especially befroe the network grows
-
sgp_[m]
Doesn't it set a new ML so it does indeed change?
-
ArticMine
The penalty formula applies to the MS / MN
-
ArticMine
What changes is the penalty free zone
-
ArticMine
so MN does not fall below ML
-
ArticMine
much has now MN does not fall below 300000
-
ArticMine
Once MN is over ML the penalty formula does not change
-
ArticMine
You mean going from 1.4 to 2 in the rate of growth of ML?
-
sgp_[m]
That wasn't what I was referring to, but we can talk about that separately
-
ArticMine
There is nochange in ML then
-
ArticMine
The different is that now MN can go below ML and still require penalty
-
sgp_[m]
What growth rate is allowed with 2
-
sgp_[m]
1.4 to 2 is a pretty large jump
-
ArticMine
2x over 69 days
-
ArticMine
vs 1.4x over 69 days
-
ArticMine
The is maxing every this out
-
ArticMine
That
-
sgp_[m]
So with 2x, looks like a 7.4x allowable increase per year if my math is right
-
sgp_[m]
Wait
-
sgp_[m]
What was under 1.4
-
sgp_[m]
*that
-
sgp_[m]
10.6x with 2?
-
sgp_[m]
Or does it scale up even faster
-
ArticMine
5.3 with 1.4 vs 32 with 2
-
ArticMine
In 2016 we say 10X
-
sgp_[m]
What's the math to get 5.3 and 32?
-
ArticMine
1.4^5 vs 2^5
-
sgp_[m]
Got it
-
ArticMine
For it to work properly it is not run at max
-
ArticMine
not even close
-
sgp_[m]
Why not, say, 1.6? Closer to 10x, and you can still pay more to use the space than the min fee
-
ArticMine
It would have failed in 2016
-
sgp_[m]
It wouldn't have failed, transactions simply would have gotten more expensive for a short term
-
ArticMine
We would the be relying on the burst capability on ongoing growth
-
ArticMine
and setting up the stage for the scenario in issue 70
-
sgp_[m]
I need to run, but we can continue this conversation later
-
ArticMine
In any case the place for this is on the githu post
-
ArticMine
github
-
ArticMine
not tying up the meetig
-
ArticMine
meetig
-
sgp_[m]
Okay cool
-
sgp_[m]
Meeting over :)
-
sgp_[m]
Ty ArticMine!
-
ArticMine
We need to move on to the ring size question
-
sgp_
sorry, I needed to run immediately earlier
-
sgp_
ArticMine: can you post the logs on github?
-
midipoet
I vote for increasing the ring size.