I am trying to understand every bit of mathematics done in source code. I have read zero to monero but it not exactly tell about every functions.

Try ztm 2

So wht is use of sc_mult() and ge_double_scalarmult_base_vartime()

Have have read the version 2.0

As bulletproo

As proveRangeBulletproof() is the one responsible for creating Commitment and proving range proof

But when it passes control to bulletproof_PROVE(amount,mask), why bp_PROVE usessc_mul() and ge_double_scalarmul_base_vartime()

Try this question in

#monero-research-lounge or lab

In a wallet with about 100,000 unspent outputs it seems it takes 2 minutes for wallet2 to choose 7 or 8 of them for a transaction; choosing 1 is almost instantaneous. Is that known?

Yes ofcourse it would be fast

<rbrunner> "In a wallet with about 100,000..." <- Running on master 0.18 wallet? Or 0.17.x

0.18

So no DNS nonsense :)

Maybe the original issue we thiught rhe dns was - the cache

Is anyone there for my queries.

ofrnxmr[m]: Yeah, maybe it's this problem.

Just tell me why sc_mul() do

Guest53: Try mrl.. thats where they hang / keep notifications on

Yup, if it's about that, you probably won't have luck here: "But when it passes control to bulletproof_PROVE(amount,mask), why bp_PROVE usessc_mul() and ge_double_scalarmul_base_vartime()"

Wht is mrl

#monero-research-lab channel

Ohk

Where our eggheads lurk :)

😅

The crypto wizzards' den

rbrunner: is mrl active or not

I think usually at this time of day not too much, the US is still sleeping. You may need a little bit of patience. It's also usually quiet here morning UTC.

Okay

You are from which country

Switzerland

(Not doxxing myself, I am not anonymous.)

rbrunner: are you crazy. I was just asking coz when you told USA is sleeping than i thought you belong to countries near me so just asking my neighbour nothing more.

That wasn't an accusation, just a clarification.

But nice to meet you. Today only I joined this chat so my problem get solved as i was reading source code. But

rbrunner: okay i was just saying in friendly way just joking leave it.

Misunderstanding

Alright :)

😃

Rbrunner means

"Her everyone. Before you accuse me of doxxing myself or giving out too much info, im not an anon. Its ok. I dont care if people know my name or country"

Exactly.

Have you read monero source code

He wasnt saying "im not telling you more because I dont want to doxx myself"

But I can understand how it could have sounded like that

Yes, but if I happen to read crypto stuff I usually don't understand half of it

Ya that maths calculation is just something else in theory in crypto books in compared to source code

But anyway, if I really want to find out something, I may ask, or I debug. That can work wonders, if you can really watch what the program does. Can be really powerful.

First i understand the maths required but now in source code i am not getting any maths.

Okay ya from debug i remembered one thing

In any case, when in doubt between a book or the code, the code is right. Why it is so is of course another question.

I was compiling monero but some dependency were old when I downloaded them debian told they were corrupted and can't be downloaded

Ya code is right but code logic is very easy to understand but not when maths is done.

Just think, if everybody would understand, where would the challenge be in this :)

😃😅

But what about solving challenges

I not imagined that chat on monero could be so active as monero.stackechange is rarely active.

Yeah, it's not too obvious that these IRC channels here are the place to be, but anyway, you found it now. Just that time zone problem, planet is not rotating fast enough

What is the issue you mention about outdated dep? Has to do with the crypto? Or just monero in general?

Were you building from master/v0.18?

The most current version the new guy fork

Sorry it's git fork

No the issue is related to dependencies required in compiling monero source code i compiled many time it not happened but after installing new os the problem is there now

Depends build? Or using your system libs

The cake gives error "Couldn't find required header zmq.h" but zmq5 is installed on my system

Its looking in a different location than your os has itbstored.

Logs? I dont think debian should be having issues

But it's giving issues. libzmq5 is installed and dpkg and apt are showing it, yes libzmq3-dev is not installed i tried to install but debian giving package corruption error.

You can always use the make depends or, I prefer gitian

I am also on Debian 11, without problems. I also think it can't be a general problem, otherwise more people would report it.

Never had a "package corruption error" in my life ...

Parrot Os 5.0 on debian 11

That exists?

Older version of parrot was not giving error but this is giving

After reinstalling the newer version of parrot on system after that problem is occurring

Maybe that ultra-secure system stumbles over its own feet somehow. Would not be the first time, tight security sometimes can go wrong

Sounds like your os repos are broken

+1

So wht should i do

Shout at the parrot :) No, seriously, maybe try to contact any Parrot people

Okay i think i should try contacting them. Shout at parrot 😂🤣

I don't even think it's possible for *Monero* to do something wrong and the result is a "package corruption error" ...

Double check your sources, or check parrots online website repo to see if the package is properly maintained,

Or go Google "parrot os 5.0 libzmq3-dev package corruption on install"

Thanks a lot. I am going to check it now.

See you both later. Thanks once again first time got community feelings.

Welcome.

I tried to fix libzmq3 by directly downloading from website instead of apt or dpkg and manually downloading all their dependencies after dependencies as one depend on other and other on other. But now it is done.

rbrunner: i am that same person whom you and that another one helped to solved my problem about dependencies and now it is solved but source code one is now also unsolved hope research lab see it soon.

I cleared the cookies as i thought libera would have stored me from my ip address as it uses behind my username but it is not doing anything like that so new name as a new guest.

Butnnow it is showing error related to source code of monero "error: find_embedded_file" was not declared in this scope

That error does not tell me something, sorry

Is there a way to decode the outputs_hex from export_outputs - getmonero.org/resources/developer-g…des/wallet-rpc.html#export_outputs?

I would like to see human-readable information from the outputs_data_hex..

It's a quite complicated binary serialization, I don't think that will be easy. Probably much easier to add code to the Monero code to log what exactly gets written.

binaryFate: CLI only please, GUI not yet

hyc: randomx fix confirmed working, at least it doesn't crash

Structs in ge.h these are the curve points so why they are 40 bytes in size as curve point should be (x,y) (32byte,32byte) and when stored as compressed should be of only 32 bytes but example like struct ge_pe3  have four things in it X Y Z T why there are four points and why each is of 40 bytes instead of being 32 bytes

selsta what fix?

see latest merged randomx pr

ah ok

selsta: ok good to know. so jit works again, no more UMASK needed?

no more UMASK needed, but don't know if it uses JIT or not. they just said no crash without setting UMASK

Structures in ge.h are actually ecc pints (x,y,z) but why are these of 40 bytes instead of being 32 bytes as in book it is writinen points are of (x,y) =(32,32) so why 40 bytes

judging from the hashrate he reports, JIT is not working old.reddit.com/r/monerosupport/comm…4iihg/segmentation_fault_11/ihi1nql

Is there anyone to solve my queries, ecc points are of 32 bytes than why structure in ge.h containing curve points are each of 40 byte or int Fe[10].

sickhrt: curve points are not stored in fe structs, finite elements are (elements of the curve field with order q)

It's ge struct in ge.h for points

sickhrt: check ztm2 section 2.3 footnote 9

And size of q is 32 -something bytes, so every point will be less than that

I updated CLI hashes in DNS records

UkoeHB: i checked that but i am not getting why it wants to say.

See there is a Fq where q is 32-something bytes and every point choosed is within that, am i correct?

So when these points will be stored in ge struct they should be 32 bytes not 40 bytes

sickhrt: curve elements (ge) are multiple finite elements (fe) using different coordinate representations (Jacobian and extended Jacobian iirc)

Just tell me one thing struct ge_p2 is used for what.

`void ge_scalarmult(ge_p2 *r, const unsigned char *a, const ge_p3 *A);`

scalar multiplication lol, for example; it's used all over the place

as an intermediate representation when doing curve ops

Ohk so you want to say there is 40 bytes of space just for sake of multiplication

no, a ge_p2 is 120 bytes

Yes i know

One variable is of 40 bytes i am saying so

sure

yes

So there are not for storing curve point but for calculation

correct

Thanks

A lot

you deserialize with ge_frombytes_vartime() and then reserialize with ge_tobytes()

And that curve point after multiplication they become public key and other all stuff for whatever purpose are of 32 byte cause of compression.

uh yes?

not sure what you're saying lol

🤣

No problem but i got wht you wanted to say

Thanks a lot

sure thing

Can anyone tell me about bulletproof range maths as ztmw don't have description about it and i am not getting wht other external source are saying.

your best bet is github.com/AdamISZ/from0k2bp

Only thing I understood in papers i read is first it takes amount and than bitwise multiply with 2n and than some other maths but wht if someone made range proof of wrong about how security for that is verified i am not getting that part.

UkoeHB: i want to know wht if someone made range proof of fake number than how security will verify it.

From wht i read i am thinking range proof is not secure thing.

If someone makes an invalid range proof, then they can inflate the currency. The point of these proofs is it shouldn't be possible to make an invalid proof except with negligible probability.

You guys can read at www.moneroinflation.com I have been trying to make a good introduction to these concepts too.

rbrunner: github.com/monero-project/monero-gu…installers/windows/WelcomeImage.bmp

can you update this for v0.18? i grepped the source code for oxygen orion and missed the installer image as it's an image :/

Oh, sure, no problem. Pretty long already since last time :)

don't we have some commandline text-to-image tool that can insert the proper text for us and be automated?

pbmplus / netpbm prob has a tool to do it

yes having the graphics be producible by a tool would be ideal

Well, yeah, once every year or so :) And you need a special font if you want to stay in style.

PR opened.

what font is currently used for the image?

I use a Windows font that I picked up somewhere called "Century Gothic" which is close enough to the font the original design work used

It's complicated: monerooutreach.org/monero-typography.html

do we really need this font for anything other than the word "Monero"?

I mean, that's part of the logo, but a release name is just ... auxiliary

rbrunner: please add closes monero-project/monero-gui #3984

Hope I did it correctly

yes

hyc is trampling my artistic freedom. I am offended.

:)

:P well we have the font definition, I'm sure it's no bother to use it

sech1: can you squash monero-project/monero #8381 ? and did you test it locally for a while?

i've been running it on my nodes but I don't mine so my testing seems useless

rbrunner fwiw, using the MoneroGothic font in the image looks weird. Since it's bold face and you're using Century Gothic / not bold

Btw what point size did you use for the text? I eyeballed it as 40 pixels high

Wondering how much we can automate this anyway, if we wind up with a very long release name we'd have to do some weird word wrapping or widen the image