<bridgerton[m]> "<Soiled> Monero needs to fund..." <- Theres no internet...

<Soiled> Go to South Korea, give them money and tell them to do it

<Soiled> Buy 3d printers to print guns

<Soiled> Food to give to North Koreans

<Soiled> Miners to mine into da farlands

Meeting scheduled here in just over an hour

Starts at 1600 GMT?

Correct

<carrington[m]> "Correct" <- It's first time for me and new within the community. Its going to be over voice, video or text? Using Element as a client for chat.

It is just a text-based meeting in this channel

Meeting time!

Hi

yo

hi

1. Introduction

Hello and welcome to this regularly scheduled community meeting. This is an opportunity to discuss recent event in Monero, crowdfunding proposals and updates as well as any other relevant topics which people want to talk about. It is an open meeting so feel free contribute. We will try to stick to the agenda on the github issue.

Hi

2. Greetings

Hi 😁

Hello. Matrix seems to be running very slow for me still.

yeah, bridge seems slow

3. Community highlights

carrington[m]: Participating first time. Hi!

As far as community highlights I think the biggest news is that all the PRs for the upcoming network upgrade are submitted and it looks like most of the issues raised at the last dev meeting have been resolved.

welcome hansams :)

You can try refreshing this log link to see messages quicker :

Are there any other recent news items people would like to discuss? The best way to keep up with news in the Monero community is to follow on of the community news sources listed in the github issue or to join in with monero communities on Matrix/IRC, Reddit, Twitter etc.

I was pleased to see monermoo pick up coding again the other day and release something to reduce unsigned transaction file size by 40+% , @bbqcore knows more about it

yeah, i've been talking to Andy at AirGap about making a QR standard for monero offline txs

moneromooo has pushed a PR that addresses the data size bottleneck - which was the biggest obstacle to making QR codes possible

From what I gather, this makes it more feasible to do some clever stuff with QR codes?

now we just need to test and gather data size from different wallets - if anyone has experience doing offline txs and wants to help test so we can get feedback that would be great

Very cool. It would be nice if this could be implemented in Sidekick from Monerujo

carrington[m]: monero output export data is usually too big to fit in QR, even animated ones

youd have to scan for like 2-3 minute

this PR really reduces the size and would improve the qr UX

And it looks like Sidekick has been getting an uptick in donations

yeah. great time for offline monero txs

carrington[m]: agreed, the standard needs to be made first. the team at AirGap is really interested in making this happen

bbqcore[m]: 40% smaller blob can be represented by qr code, but 40% larger can't be ?

wtf ?

only 40% and no problem ?

come again

Which brings us onto

4. CCS/fundraising updates

- 99% smaller in some tests 👀

a. Monero Afghanistan Expansion Strategy by spirobel

plowsof[m]: yeah we need more eyes on this thing

test it in a bunch of wallets

bbqcore[m]: tx representation in the same between all wallets, what do you suggest to test ?

s/in/is/

> <@bigbklynballs[m]:libera.chat> > <@bbqcore[m]:libera.chat> test it in a bunch of wallets

>

> * tx representation is the same between all wallets, what do you suggest to test ?

were looking at the size of "outputs" export file

carrington[m]: Is there any other expansions in other countries/cities/regions or is Afghanistan the forst focused one?

every wallet is unique and size can vary depending on usage

s/forst/first/

carrington[m]: feels quite unclear what the scope of this ccs is

It looks like spirobel isn't around, but people can comment on the issue on gitlab. I don't think there has been a reddit post for this proposal, so that would be a good idea to get more eyes on it.

Yes I agree the deliverables are a little vague.

ii seems that they are doing a study to see what is actually feasible

Actually they made this reddit post with a similar topic

it looks like a research paper, yes

with someone who is there

no idea where spirobel[m] is located

tora bora

*someone on the ground who is greatly affected by recent events

Well anyways if people could leave a comment there it would be appreciated.

b. mj part time coding 2022-03

The first question is how people in that country are earning money right now

mj-xmr[m] mj-xmr

Hi. Thanks for pinging.

I would like a comment from Rucknium[m] on why tsqsim is preferred to existing open source projects.

And why monero can be better than that way

Give me a minute to read...

jberman

bigbklynballs[m]: half the country was recently prevented from working

nioc: external sanctions or internal restrictions by new govt ?

new gov, Taliban

It would be hard to compete with guys with guns

then

usually is

In that case people will prefer anything that reduces risk to be killed and monero via electronic device isn't the safest solution

mj-xmr[m]: we need monero developers and funding someone almost full time for an external tool doesn't make much sense in my opinion :/

There is also a related reddit post

unless Rucknium[m] can explain why his research wouldn't be possible otherwise with existing tools

bigbklynballs[m]: I am not suggesting at all that it is comparable, but as a principle you are referring to any government in the world. The problem exists no matter the country. In Europe they just use other measures: restrictions, fines, imprisonment etc. That is why I was asking, if this Afghanistan initiative is a pilot or there are others in other countries. There might be similarities and good practices.

<selsta> "I would like a comment from..." <- He has made a comment already, since such a question was asked. I can't find it, so I'll paraphrase:

He said that he could perform this already with existing tools, but he preferred to be in close contact with somebody who has experience with a time series analysis, in order to quickly respond to required changes. This was commented some 4 months ago.

* The problem (guys with guns) exists no, * In Europe/US/Canada they, * imprisonment etc. And finally they bring out guns. That is

Lately I've reviewed his document, and OSPEAD requires about 50 thousand iterations in a single data portion to find an optimal setting, so I personally think that a tool, that focuses on speed of execution would speed up the research. All the other tools serve more as discovery tools, rather than robustness check tools.

hansams[m]: as far as I'm aware there are no targeted Monero outreach proposals like that one, and none have ever been funded by the CCS

selsta: Unless the tool is finished. Then I'm more free for the internal development.

mj-xmr[m]: yes, I think it was fine for the last CCS to port the tool but continuing to work full time on it for months now just isn't a good way to spend dev focus on my opinion

mj-xmr[m]: But still, the decision is yours.

in your last CCS it sounded like the tool is already coded and you would just port it

so I was surprised that you want to continue focus on it for 3 months

Yes and this happened. But still, these were just 3 months. Is this a lot for such a tool, you think?

One could work with the tool already somehow. This was my goal. But quite a few features can still be ported. I could tell more about them, if there's such a need.

If we want to fund someone for months to work on a tool there have to be clear benefits on using it over existing tools. That's why I wanted a comment from Rucknium[m].

Otherwise, here's the development plan:

There are some new requests from Ruck.

> <@mj-xmr[m]:libera.chat> Otherwise, here's the development plan:

> github.com/mj-xmr/tsqsim/wiki

> There are some new requests from Ruck.

What's the final goal of this ?

The tool for ... ?

Initially it was meant to predict the expected number of transactions in a given time step (for OSPEAD - decoy algo), but after various discussions, many people have found other uses of the tool for predicting also other values than the transaction numbers. Examples include: Seraphis CPU drainage and input for adaptive mining.

It would be beneficial is there was someone else who has reviewed your work in depth who could comment on the need for this proposal. Although I realize it is difficult to get reviewers.

mj-xmr[m]: machine learning model that predicts number of transactions, and your tool is for training such model or evaluation of such model ?

@carrington I think too, that Rucknium would be the best person to comment. Let's maybe give him some time and comment under the proposal, when he finds time?

+1 for mj-xmr suggestion ^

mj-xmr[m]: What does it mean "Seraphis CPU drainage" ?

> <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> Initially it was meant to predict the expected number of transactions in a given time step (for OSPEAD - decoy algo), but after various discussions, many people have found other uses of the tool for predicting also other values than the transaction numbers. Examples include: Seraphis CPU drainage and input for adaptive mining.

>

> machine learning model that predicts number of transactions, and your tool is for training such model or evaluation of such model ?

Preferably classical TSA (AR/MA, ARIMA, SARIMA, etc.) for simplicity. Coding ML can be left for a plugin dev. The tool serves as both training tool and evaluation of the model via Walk Forward Validation (or Optimization)

Yes the gitlab comments have only been available for a week so there is plenty of time

c. j-berman full-time 3 months part 2

support ^

> <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> Initially it was meant to predict the expected number of transactions in a given time step (for OSPEAD - decoy algo), but after various discussions, many people have found other uses of the tool for predicting also other values than the transaction... (full message at libera.ems.host/_matrix/media/r0/do…c8ac35f03c9a1d9f41e9664c3ee5e70f1c3)

(Collecting my thoughts on this) I will say that my own pace on OSPEAD and related matters hasn't kept up with mj-xmr 's pace on developing tsqsim, since I have not yet arrived at the computational time series portion of the work.

It looks like jberman[m] is not around right now, but this CCS proposal was also only made recently

I support jberman's proposal

> <@mj-xmr[m]:libera.chat> > <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> Initially it was meant to predict the expected number of transactions in a given time step (for OSPEAD - decoy algo), but after various discussions, many people have found other uses of the tool for predicting also... (full message at libera.ems.host/_matrix/media/r0/do…95fa763b253dcc791070a41fb3663bcee23)

will left a comment on gitlab

I had to carry out some BCH work that I had committed to before OSPEAD ( See read.cash/@Rucknium/update-on-cashf…team-phase-one-flipstarter-4dc6e95f if you are curious on that at all. )

> <@mj-xmr[m]:libera.chat> > <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> Initially it was meant to predict the expected number of transactions in a given time step (for OSPEAD - decoy algo), but after various discussions, many people have found other uses of the tool for predicting also... (full message at libera.ems.host/_matrix/media/r0/do…f18ba6618e9a26db89f9a17423b7e679936)

> <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> > <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> Initially it was meant to predict the expected number of transactions in a given time step (for OSPEAD - decoy algo), but after various discussions, many people have found other uses of the... (full message at libera.ems.host/_matrix/media/r0/do…23d8ee30444ea072185510666a6de2039ec)

Also for a while I have been struggling with some strange behavior in my "dry run" of OSPEAD that seemed "simple" since it _sort of_ involved the order in which to apply logarithms, but I believe in the last few days I have resolved the problem, or at least found an alternative route.

> <@mj-xmr[m]:libera.chat> > <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> > <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> Initially it was meant to predict the expected number of transactions in a given time step (for OSPEAD - decoy algo), but after various discussions, many... (full message at libera.ems.host/_matrix/media/r0/do…863958027d5edda250c8b669286c9b0f5b1)

It's true that as formulated today, OSPEAD is computationally expensive. Many time series techniques are written by statisticians, not programmers, and are therefore slow. So a fast C++ implementation of things may be important to make it viable.

Thanks for the update Rucknium .

> <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> > <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> Initially it was meant to predict the expected number of transactions in a given time step (for OSPEAD - decoy algo), but after various discussions, many people have found other uses of the... (full message at libera.ems.host/_matrix/media/r0/do…1cf392d45ff474854eef4d3dddeae7b54d9)

> <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> > <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> > <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> Initially it was meant to predict the expected number of transactions in a given time step (for OSPEAD - decoy algo), but after... (full message at libera.ems.host/_matrix/media/r0/do…76b88a4c8c98b4fd7a32d575265b562cd78)

may be

rucknium[m]: Any example of inefficient code written by statistician ?

nioc: May be important, but will surely be faster, if this is of any value.

Anyway, ideally I would be more advanced in my OSPEAD work so that I could catch up with mj-xmr 's place that he is at with tsqsim. I think with the logarithm issue mostly resolved, and therefore imminent public release of (part of) the OSPEAD technical specification, I may be there shortly.

I'm not 100% ready with my tool either. It's good but not great yet.

value would be cloning jberman :D

"I'm sorry. I'm not Neo."

bigbklynballs[m]: I'm not sure since I haven't been able to delve into things. Broadly, poorly optimized statistical code written in a slow language can be tens or hundreds of times slower than well optimized statistical code written in a fast language.

are there some benchmarks that comapare your tool to existing tools mj-xmr[m] ?

Well perhaps it makes more sense for mj's proposal to be delayed for some amount of time? Until OSPEAD has progressed a bit further

rucknium[m]: Do you know only slow languages ?

selsta: I'm sure that it beats other ones, if not on the face value, then surely due to its parallelism (including network parallelism), so I didn't need it, but I can prepare some.

> <@bigbklynballs[m]:libera.chat> > <@rucknium[m]:libera.chat> bigbklynballs[m]: I'm not sure since I haven't been able to delve into things. Broadly, poorly optimized statistical code written in a slow language can be tens or hundreds of times slower than well optimized statistical code written in a fast language.

>

> Do you know only slow languages ?

This is quite rude. Please give me a tool you want me to benchmark, so we can talk on some level.

I.e. I mean increase the "expiration date" on the CCS proposal

carrington[m]: I'd be fine with this.

But again that proposal has only been up a week so maybe it is too early to make decision like that

In case people were not aware, since the last meeting there was also a large proposal for funding the Haveno dex which was ultimately withdrawn after lots of discussion on reddit and gitlab

I'd love to know until the end of the month.

In the case that it should be delayed, shall I prepare an alternative one (in the way like selsta proposed)? If so, I'd like to start the alternative one in March anyway.

bigbklynballs[m]: I mostly know R, which is generally a slow language. An experienced R programmer can avoid some issues that slow R down, but that only gets you so far.

On the agenda there is also

d. 'Work in Progress' projects discussion

I'd say to prepare a new one, yes mj-xmr

Do we have a meeting now?

so you don't have to wait until Rucknium catchs up

started 1h ago msvb-web

carrington: my ccs.getmonero.org/proposals/mj-part-time-2021-q4.html is complete with 3 reports. Just not paid yet.

The ccs.getmonero.org/proposals/mj-compil-time-reduction.html is scheduled for funds redistribution.

After several months of inactivity, xmrSale has an update repo.getmonero.org/monero-project/ccs-proposals/-/merge_requests/246

I was a bit upset with the 'anonymouse team' for this update, claiming that something was impossible but we already have a few examples of it. BusyBoredom the creator of AcceptXMR has offered to help them. (which does the thing they claim is impossible)

But at least theyre alive 👍️

> <@carrington[m]:libera.chat> On the agenda there is also... (full message at libera.ems.host/_matrix/media/r0/do…6b1a364b2ab5c088e212b1f30cf85374a36)

Keeping up with in-progress CCS work seems to be getting more difficult with the high number of projects in that category

Several of the CCS proposals are nontechnical, so it would be immensely helpful if anyone can take the time to verify their work or lend a hand.

I also took a look at this 2~ year old proposal in limbo ccs.getmonero.org/proposals/36c3.html

and made a suggestion to 'pay out' some of the funds to the volunteers involved. ajs_ attended , and is owed funds from it for some expenses + to pay off his ticket)

there was budget for volunteer expenses so Ajs is saying 50 euro per day * 5 = 250 (food / drink etc). + 30 euros to fully pay for his ticket. 290 is 1.0562639 percent of the google docs price of 27,455.26. 1.0562639 % of 560 xmr is 5.91507784xmr to be awarded to Ajs. Who wishes it to be donated to monerokon

do we know if these funds were paid out or not in the first place?

maybe it simply didn't get written into the ccs system

ajs_: also has fund left over from xmr.radio and i quote "If is 5.915 XMR and there is a little over 17 XMR left from the Monero FM CCS, luigi1111 luigi1111_ would it be possible to reallocate these funds (~22.92 XMR) towards MoneroKon's CCS?"

selsta: Diego said the funds have never been paid out

IIRC luigui was going to roll those fund to general fund

plowsof[m]: it says 280 were paid out

the remaining haven't?

then it could be used to fund monerokon if core sees fit

The community donated for volunteer expenses and those volunteers where never paid

let that sink in

Ajs out of the kindness of his heart wishes for his funds to be donated to monerokon

That is very generous of them. It would be good if you each could drop a comment in the gitlab to make the situation clear.

Anyway we are over the usual 1 hour, so I need to point out that I won't be available in 2 weeks to help with the meeting. If someone could volunteer to chair a meeting then it would be appreciated.

carrington[m]: OK. Will look into the projects. Somehow it would be helpful to use some kind of tagging in the projects to make it more clear what it is all about.

this was discussed over #monero-events carrington

selsta: i see.. we nee d clarification on this

lol luigi probably has no idea himself

according to Diego Salazar there was no payout 🤷

up until the time that Diego was still working for core

Ajs never received a cent for expenses (5 day/night food / tickets) (along with the other volunteers)

but very much doubt anything has happened since then

Thanks everyone for attending the meeting! Ending it here for the purposes of logs.

ty

thanks carrington

thanks carrington

hansams[m]: usually there are not so many all running at once

Reminder that there is a meeting in #monero-events:monero.social in around 45 minutes for Monerokon 2022 planning

SerHack: work in progress has taken a back seat as he is busy with mastering monero 2

General question by newbie like me. So, the disussions are carried out in IRC channels, Reddit, Twitter? So, if there is nothing which catches eye, then it does not happen or exist in the community? Ot there are other, like dedicated work groups and discussions?

also everyone is welcome to ask questions about the translation workgroup since we didn't get that far during the meeting

hansams[m]1: this keeps an overview of many channels monero.observer

my brief report is that there a new PR for monero gui waiting to be merged

and I'm trying to make screenshot context work on weblate

Looking forward to another edition of Mastering Monero!

still getting 404 erros when uploading them

s/there are/there is

Yes sorry we seemed to run out of time but I have to end the "meeting logs" somewhere. Discussion can of course carry on.

no need to apologize

i'm always around

thanks netrik182

Thanks for the meeting. I am still getting use to all of it and learning and trying to get trough all this technical lingo (at leas to understand, what is important for me as non-techie to understand what not), but hopefully will be actively involved in close future.

Good meeting, thank you.

> <@mj-xmr[m]:libera.chat> > <@bigbklynballs[m]:libera.chat> > <@rucknium[m]:libera.chat> bigbklynballs[m]: I'm not sure since I haven't been able to delve into things. Broadly, poorly optimized statistical code written in a slow language can be tens or hundreds of times slower than well optimized... (full message at libera.ems.host/_matrix/media/r0/do…b5c2f97fd1ff5b9990a70f139ad96586270)

Some python libs support parallel model training, but gpu libs are likely faster than even parallel cpu training

and your lib support only cpu training

> <@bigbklynballs[m]:libera.chat> > <@mj-xmr[m]:libera.chat> > <@bigbklynballs[m]:libera.chat> > <@rucknium[m]:libera.chat> bigbklynballs[m]: I'm not sure since I haven't been able to delve into things. Broadly, poorly optimized statistical code written in a slow language can be tens or hundreds of times... (full message at libera.ems.host/_matrix/media/r0/do…5d62f91c8a2e4ccbfe54cf71a2cbbaa8a4e)

in order to keep up with industry state it would be good to compare with all of them

just to know that our lib is true state of the art in the industry

s/our/your/

bigbklynballs[m]: GPUs are only good when you deal with low computational complexity and huge amounts of data. But I can benchmark it against one such lib.

bigbklynballs[m]: This sounds like a proposal of its own.

These libs are already written and maintained by someone, so in order to not pay twice it's mandatory to learn what is already available

bigbklynballs[m]: Then let's make it so: I prepare a test case on something that's already been written and tested (my tool), and you prepare the equivalent tests on all of the above ones, except one, that I'd do, that deals with GPU processing?

> <@mj-xmr[m]:libera.chat> > <@bigbklynballs[m]:libera.chat> These libs are already written and maintained by someone, so in order to not pay twice it's mandatory to learn what is already available

>

> Then let's make it so: I prepare a test case on something that's already been written and tested (my tool), and you prepare the equivalent tests on all of the above ones, except one, that I'd do, that deals with GPU processing?

If you will not able to understand how some of those libs are working and quickstart guides will not help then you can ask me here

I'll try to help you

<plowsof[m]> "according to Diego Salazar..." <- So this is slightly incorrect. There was a first up front payout

The second half is the one that's up for grabs

sorry, should have clarified

For the chills: imgur.com/a/vRp3R

<bigbklynballs[m]> "I'll try to help you" <- I will not need help with this one benchmark. It's fine. The rest is on you, since you already know them.

<Torr> "For the chills: imgur...." <- looks like it would be easy to steal power from neighbors from down there, mine crypto lmao

<chesterfield> "I do think it's important to..." <- And what if the already available tool is simply deficient? The performance boost, that I'm talking about deals with being able to do certain stuff at all in reasonable DEVELOPMENT time, not just run time. If run time was the only cost, you'd just let it run for weeks while doing other things. In our case, you need to have reasonably fast feedback while you're still developing the

models. I'd argue, that using generic tools, not being written to help our case in the 1st place, will not have this advantage.

Ah ok thanks for clarifying Diego Salazarany idea what it paid for out of that google docs thing. I thought volunteer expenses would be the easiest place to start as this wasnt paid for. all i know is that ajs_'s hotel room was paid for by RIAT (maybe the ccs covered that) but his expenses not.. just attempting to chip away at it because its just not acceptable that a volunteer paid out of pocket when the community made sure that

wouldn't happen (i know its not your personal doing). it would be nice if other volunteers came forward to get their 5*50 euro (% adjusted) expenses (food/drink - seems an acceptable figure) + ticketcost at least 😀 I will try my best to keep bringing this up until something happens 🤷

monerobull[m]: Sharp focus, eh. Zero missed opportunities.

It was paid out before the conference since doing it all on RIAT's dime initially wasn't an option

they needed some kind of money to work with

<chesterfield> "I do think it's important to..." <- And what if the already available tool is simply deficient? The performance boost, that I'm talking about deals with being able to do certain stuff at all in reasonable DEVELOPMENT time, not just run time. If run time was the only cost, you'd just let it run for weeks while doing other things. In our case, you need to have reasonably fast feedback while you're still developing the

models. I'd argue, that using generic tools, not being written to help our case in the 1st place, will not have this advantage.

(had to repost, because I removed it accidentally from Matrix)

<bigbklynballs[m]> "Some python libs support..." <- Ew python

<Encore> 👍

ok but like

where do i buy these

<ian.niculescu> theblockcrypto.com/linked/133904/pa…tm_source=twitter&utm_medium=social

<ian.niculescu> this is a sentralized exchange

<ian.niculescu> (SEX)