<victor5577:matrix.org> Hi

<sophianawttyzoobabymatrix.org:matrix.org> Hello

<victor5577:matrix.org> What's uo

<sophianawttyzoobabymatrix.org:matrix.org> Good you?

Hello, I would like to know what is the min amount of time for monero mining in a web browser?

<elongated:matrix.org> johny_: Browser mining died half a decade ago, your are too late

it is not possible to mine monero with web browser?

<elongated:matrix.org> johny_: You maybe able to mine with 1-2h/s, so no you can’t mine in web browser

what is that github.com/NajmAjmal/monero-webminer ???

<rbrunner7> That may be the webminer that works. With an expected income of maybe 5 cents per year through browser mining. Does that sound interesting to you? :)

<elongated:matrix.org> johny_: A toy to mine, but it’s ineffective

<elongated:matrix.org> @rbrunner7: Maybe less than that

actually I want to use monero PoW for something else but maybe I should use another crypto currency

<elongated:matrix.org> It was fun for web developers while it lasted

<elongated:matrix.org> Randomx killed it

<elongated:matrix.org> johny_: You can use randomx pow for verification, but not mining income

I want mining income + PoW to kill 2 birds with one stone

<elongated:matrix.org> Maybe look at some other coins, but those ain’t going to make you much mullah as most of those coins are dead

johny_: mining income LOOOOL

from browser?

<elongated:matrix.org> Cindy: Botnets work

not browser users though

<rbrunner7> johny_: Making money through mining for normal people with normal hardware is basically dead for a few years already. For all coins. The times where you could mine Ethereum with your GPU for maybe USD 1 per day are gone for good.

Cindy yeah I want to use the browser to mine crypto

<rbrunner7> I have a nagging feeling that dear Johny does not believe us ...

exactly I am not talking about a single user

RandomX is very inefficient on browsers

due to the fact that WASM doesn't have custom floating point rounding modes yet

<rbrunner7> Oh, you have a website with a million visitors per day, and you want to let them all mine for you? Good luck. Google will declare your website dangerous in 5 minutes or so and block it.

the RandomX WASM implementation you see around here uses a custom-made library that emulates the floating point rounding modes, which leads to less hashrate

rbrunner7: that is if the browser even lets you allocate enough memory for the dataset

Google does not also website that mine but it is ok for them to catch all personal data?

browsers cap memory allocation, and will kill your tab if you go over it

and the cap is not even a gigabyte

<rbrunner7> johny_: The world is a bad place, it's so unjust.

rbrunner7: ETH nowadays can only be staked by rich people

<rbrunner7> You should start an AI coin and then do a rugpull. Could net you millions if you are lucky. You also will be a criminal afterwards, but isn't there always a problem?

<rbrunner7> Role model: Trump family

though i will have to say, RandomX.js has improved beyond any other ports of RandomX

including having a WASM JIT, rather than just blindly using the very slow interpreter

rbrunner7 actually extranet does not need to be indexed on google

I mean member area stuff like that

I need to find out the most profitable coin to mine in a web browser

none

the algorithms that are easier for browsers (like hashcash, equihash, cryptonight) are dominated by ASICs

and will net you next to nothing due to the extremely high difficulty

is it good idea to create new crypto designed for web browser mining? I have to invent new algorithm?

you can modify RandomX to get rid of the floating point rounding mode operations

If you want coins - maybe try selling your skills/possessions for some?

but the problem is, your crypto will be worthless unless you sell it to exchanges or something

BlueyHealer you are out of topic, PoW is useful for many things but wasting cpu for nothing is a shame

better yet, how about you pester the WASM board to implement floating point rounding modes

that's the one thing that's massively slowing down RandomX in browsers

TBH I never even considered mining not in a dedicated application...

still a proposal

ok I will have to look at it, I don't have lots of knowledge about mining

if this proposal becomes apart of webassembly (and implemented by major browsers)

it would speed up RandomX.. a lot

but unfortunately as of now, you'll have to deal with like 1-2H/s

btw, how much do you think the guy at banmonero.com earns from his "burn address"

wtf is that for a page

<milas900:matrix.org> After swapping on cake wallet Xmr account shows zero why ?

milas900: probably wait until tx confirms or something

it needs some confs maybe

yea

anyone know of some ACTUAL burn addresses for monero

<milas900:matrix.org> It generated some new addresses when you swap . I’m afraid money goes to other addresses

its a sub adress maybe

<milas900:matrix.org> It’s really a bug it says synchronized and shows ur litecoin as zero..

klick the TX and check how many confirmations are there

sync is ONLY blocks SYNCED

it has nothing to do with confs

i might modify moneropy to generate an address with the public keys being 0 or something

<milas900:matrix.org> Where to find the Tx

you swapping things and dont have a clue about things

milas900: monero or litecoin??

next time you should read more into it maybe milas900 :P

<milas900:matrix.org> Swap from ltc to xmr

did it give you a TXID?

or something to identify?

there is something like a "history" buttin you can click

button :P

again, cakewallet might be filtering transactions based on the number of confirms

just wait like 20 minutes and get back to us

true, i agree

if not, show us your TXID

so we can check it on our own at the blockexplorer

that would be super useless

my first swap on cake was crazy too

but sure :P

just to verify everything is ok

:P

<milas900:matrix.org> How many confirmations are needed ?

10/10 maybe

most wallets use 10 confs

<milas900:matrix.org> Ok got it thanks

<milas900:matrix.org> The GUI is not friendly at all

i agree on that

are you using desktop version?

<milas900:matrix.org> btcdwed: Are u developer there ?

no, just wondering ;D

<milas900:matrix.org> Any version should be a user journey and experience where you track your money like human being

i dont like it too

<milas900:matrix.org> Not like ur balance becomes zero and no pending or anything

the official monero wallet shows all transactions

pending, partially confirmed and confirmed

yea

cakewallet is weird

<milas900:matrix.org> Cake wallet also shows

<milas900:matrix.org> But slow

cake is just showing unblanced amount when you do something

its kinda confusing

<milas900:matrix.org> Ur heart first stops

first time i thought my money is gone :P

<milas900:matrix.org> Then u panic

true

<milas900:matrix.org> Then you get a shock

we are here to help you

no shock at all

<milas900:matrix.org> Then you say thanks god

hehe

<milas900:matrix.org> Yes but in order price to shoot up developers should improve that cake wallet

<milas900:matrix.org> Is it open source or what ? An issue should be opened

youre free to comment an issue on github etc

<milas900:matrix.org> Is there a limit by the way to swap ?

what limit

the only limit is time, just relax and msg us if you still unsure in 30mins

:P

<milas900:matrix.org> Amount I mean

you meant ltc to xmr?

<milas900:matrix.org> Yes

i cant see any swapping limits @ cake

i made a burn address by using the public view key derived from zero, and zero as the public spend key

is this valid?

maybe i should try generating the stagenet version of this and sending some stagenet XMR

to test if the network doesn't freak out

i heard that like in some ed25519 implementations, zero causes an error in validating a signature

Cindy the real killer for wasm is using 2 GiB there are recent projects that do workable things for the float part

So you are limited to light mode

Use HashToPoint tbh Cindy that will get you a random point element within the curve that you don't know the discrete log

HashToPoint(Burn Spend Key), HashToPoint(Burn View Key)

hashtopoint?

alright alright

Though you might want to make a proper valid view key

So you can monitor burns

But not spend them

but the view key must be something that you can easily know

like uhh FF

255 as the private view key

or maybe all FF's

You can use this

Hash(spend key)

Spend pub key*

It cannot be all FF

true i thought so too

It must be an element of the scalar field for ed25519

i could have swore i saw something like that in a ed25519 page

that's why i hesitated

Monero generates points as generators like that, for example

Used in all new stuff

<datahoarder> mrelay.p2pool.observer/p/orHhys0KUlNJNTlF/1.txt (code snippet, 11 lines)

<datahoarder> Result:

<datahoarder> mrelay.p2pool.observer/p/orHhys0KUlNJNTlF/2.txt (code snippet, 4 lines)

<datahoarder> Actually, right result:

<datahoarder> mrelay.p2pool.observer/p/287kys0KRE9fTWFY/1.txt (code snippet, 4 lines)

that can load up a view wallet you don't know spend private key for, but can be monitored and even used to generate subaddresses

DataHoarder: bpa.st/GSKA

this better?

also wtf is that A prefix of the address you generated

you can't just use hash

Cindy: testnet

ah

you need to do mod l, where l is the scalar element modulo

monero already has defined methods to do this in code, I used the same on mine

oh so that's sc_reduce

it also ties spend + viewkey together

sc_reduce in moneropy does intToHexStr(_hexStrToInt(key) % l)

biased hash to point = hash_to_ec / hash_to_p3 / biased_hash_to_ec

which is what get_view_key calls internally

ScalarDeriveLegacy = sc_reduce(keccak(data))

so i just call sc_reduce on the spend key?

you also want the domain separator

instead of just the view key?

to ensure that these keys are only generated for this context

which is why I made obvious the purpose of them

"Monero Burn Spend Key" and "Monero Burn View Key"

this way you only need these two strings to generate anything

14:47:29 <Cindy> so i just call sc_reduce on the spend key?

no

that means you are generating a private key lol

you are generating a POINT

that must be handled properly or learning the dlog of it allows spend

oh i see

see git.gammaspectra.live/P2Pool/consen…/master/monero/crypto/generators.go

this is how monero generates points when it must be infeasible to learn dlog

"HopefulHashToPoint" is keccak(point

but this fails 7/8th of the time :D

wtf is unbiased

is it usually biased

BiasedHashToPoint is what is used in monero key images

ah

unbiased is new one for carrot/FCMP++

i'm talking about the original cryptonote

not really carrot yet

that's HopefulHashToPoint

and BiasedHashToPoint

oh

BiasedHashToPoint uses elligator2

which is the ge_fromfe_frombytes_vartime

just uncommented, unexplainded

kayabanerve went through the effort of reverse engineering it and making a proper impl of it

nvm, someone else did it

unless this is too old

because it says something about the generator being different nowadays

that's the hopeful one

for GeneratorH

key images did use BiasedHashToPoint which is what you want

new ones like Bulletproof / Bulletproof+ used BiasedHashToPoint git.gammaspectra.live/P2Pool/consen…o/ringct/bulletproofs/generators.go

(yes, I have implemented them!)

moneropy is lacking

though bulletproof uses BiasedHashToPoint(Keccak(data)) lol

i can't find any hash to point generators

which ends up with double keccak

github.com/bigreddmachine/MoneroPy ?

yes

8 years old lol

lol

i know

probably before RingCT and bulletproof and shit

the other monero python library is just a glorified RPC library

key images?

and not useful

does it handle that?

well, you can make C code that uses Monero's function to generate keys

no

then just import keys in hex to moneropy

moneropy and not github.com/monero-ecosystem/monero-python ?

monero-python is the RPC library i'm talking about

moneropy implements the stuff behind monero.. at least the monero of 8 years ago

lol yeah

they hardcoded H

just use my consensus library :')

well unfortunately i can't use go in a interpreter :P

finding stuff half implemented or old is why I ended up just doing it in-house

it doesn't make it easy to mess around with monero crypto stuff

actually you can!

btw I support TinyGo as well

Cindy: just generate elsewhere, small C program linking to monero (or mine)

then import keys

why not mininero?

that has hashToPointCN :D

which is biased hash to point

y'know

9 years old ofc :P

why not that

can i like borrow your code :P

feel free :D

you'll need to make all the cryptographic base if your libraries don't have that, ofc

^ this needs field Elements to work with

isn't elligator a FCMP++ thing?

nope

it was always what it was running hash_to_point

shit i'm dumb

what Carrot / FCMP++ do is define an unbiased hash to point

which runs elligator2 twice

and merges them

see git.gammaspectra.live/P2Pool/consen…ter/monero/crypto/hash.go#L108-L156

likes 108 to 156

y'know i feel like i understood SOME of this :P

definitely will read more about the point generators

well this is hash (blob of data, uniformly sampled) to point

generator is a different term

point.. converters?

<sbt:nope.chat> Wow my last txn got 10 confirms in <5mins never seen that speed before

<ofrnxmr> Havent been spending enough

<sbt:nope.chat> Wdym

see recent blocks :) blocks.p2pool.observer

<kayabanerve:matrix.org> DataHoarder: don't forget bp+ uses biasedhashtopoint for its domain-separation-tag for God knows what reason

oh right

I complained about that

double keccak and that domain tag

not only that but it also does Keccak(Keccak(

Also the transcript does Keccak -> sc_reduce32 -> Keccak -> sc_reduce32

<kayabanerve:matrix.org> Also, DataHoarder, sqrt can be implemented with inverse for Ed25519 for a minor performance benefit. I don't think the theoretical optimization is realizable, but if you are going through the weeds, you may want to try.

such a waste

I did that already

It's on my fork :)

<kayabanerve:matrix.org> Yes, I wrote the function as accepting 32-bytes because every call site did and it was the simplest way to invoke at-time-of-use as points are so literally defined, but that inadvertently introduces the requirement the value is hashed before it's hashed even moving forward...

<kayabanerve:matrix.org> While that requirement isn't currently in consensus, my work on it post-dates the decision of our last hard fork, it do be a quirk

<kayabanerve:matrix.org> Have a link/any insight into the practical performance benefit?

hmm, actually, on Ed25519 FE Sqrt is implemented as pow 2^252-1 with an addchain

err

not exactly, and I noticed a typo on the function naming on original as well

git.gammaspectra.live/P2Pool/edward…19/src/branch/main/field/fe.go#L354 Line 354 is how Sqrt is implemented

Invert is implemented similarly

most of the performances I went for on my libraries was explicitly adding precomputation of point in ScalarMult, and then also adding vartime alternatives for verification

(I have done the same on helioselene one now)

Do you have paper/information on implementing Sqrt from Inverse? As otherwise it's addchains that I have

note most of my edwards25519 implementation comes from github.com/FiloSottile/edwards25519 which is what the Go cryptographer maintained in std (though extracted here with a bunch more things)