Amoveo ♥🧿
13 March 2018
06:56
Deleted Account
when is block fork?
MF
07:05
Mr Flintstone
tentatively 9000 but idk if that is still plan Zack
JM
07:12
J M
What does fork mean for miners?
OK
07:20
O K
Just an update for miners right now
07:20
Zack If you continue to push changes I'll keep testing
07:20
@Mandelhoff How are things going?
MH
07:21
Mandel Hoff
Pretty good, Windows server IIS install locked up on me, otherwise, looks promising. Just more downtime.
OK
07:22
O K
Cool, thanks for the update
Z
07:31
Zack
In reply to this message
Yes, we are now planning on height 9000. I got the go ahead from Mandel and sy.
MH
07:43
Mandel Hoff
Ok, I think the pool is up again! Wow, I thought it was totally lost for a while there.
Z
07:43
Zack
How many gigahashes were needed to mine a block in each period:
31
1827
80421
339894
4238682

In the 4 retargeting events, the difficulty increased by these ratios:
59x
44x
4.2x
12.5x
OK
07:43
O K
Good job @Mandelhoff 👍🤘
Z
07:43
Zack
In reply to this message
good you fixed it
MH
07:44
Mandel Hoff
I'm sorry for all the down time. Happy Hashing!
G
07:44
Gonzalo
well done!!
O
07:44
OldPaul AmaZix
great job!
JS
07:44
Jon Snow
Sticker
Not included, change data exporting settings to download.
👍, 39.5 KB
BB
07:46
Brian Brian
you da man mandel
M
07:46
Mike
In reply to this message
Thanks Mandel. You’ve been invaluable. Too bad I’ve only got one graphics card able to mine or I’d donate something substantial. If you’re ever in Washington DC I’ll buy you a drink 👍🏻
MF
08:08
Mr Flintstone
happy 1060
08:08
8060*
AS
08:21
Aizen Sou
Half of network left, we might see 30min block soon
I
08:27
Iridescence
Mmm tasty 30 minute blocks
08:27
Means roughly 1000 hours to retarget
Z
08:30
Zack
Even 3 hour long blocks should be fine at this stage. Slow confirmation times don't matter for the way that people are currently using Amoveo.
OK
08:30
O K
We were pretty well on target before the problems earlier, we should be okay as people get back onlien
M
08:31
Mike
If there was a public AMD miner it’d pick back up probably
OK
08:32
O K
Someone named Tallak
OK
08:32
O K
Deleted Account 09.03.2018 03:18:13
Im working on a radeon miner that does 210 Mh/s on a rx570. Its still a work in progress and will be open sourced once it works and has some minimal testing...
Z
08:32
Zack
I was working on one too. it is still on my github, unfinished.
M
08:33
Mike
Anyone who made one would presumably not release it
MF
08:33
Mr Flintstone
I’d be surprised if block times weren’t under 10min tomorrow
08:33
provided a blockchain exists and is live
Deleted joined group by link from Group
08:35
Deleted Account
Hi All - I'm relatively new to mining and just downloaded the Cuda miner for windows. Anyone able to DM to assist with configuring the miner?
08:35
(NVIDIA)
OK
08:35
O K
I have some time Michael, feel free to DM me if you'd rather take it out of the chat
MF
09:10
Mr Flintstone
if anyone feels like they have too much veo exposure I’m happy to lighten the load. You can pm me
M
09:10
Minieep21
Anyone mining with P100? What hash rate?
MH
09:11
Mandel Hoff
My cuda miner has been reported to run like this on a P100:
Tesla P100: 1920 Mh/s - Suggested BlockSize: 192, Numblocks: 168
M
09:14
Minieep21
Yea I've tested different settings
Best I can get is 1650 Mh/s on Linux
G
09:15
Gonzalo
wow much blocks, so fast
09:15
😁
09:17
5 blocks in 13 minutes
MF
09:17
Mr Flintstone
wow
N
09:19
NM$L
wow
OK
09:28
O K
lmao
10:08
Deleted Account
Hi, a bit late but I added support for original amoveo-mining-pool.
Now you can mine on Zack's pool by specifying <PoolType> option.
https://github.com/tumblecatweed/AmoveoMinerGpuCudaLinux
OK
10:09
O K
Awesome!
G
10:09
Gonzalo
great!
N
10:11
NM$L
excellent
Z
10:12
Zack
In reply to this message
Great. thanks for the help catweed :)
N
10:14
NM$L
the diff will increase or not?
10:23
Deleted Account
5 blocks 5 veo right？
G
10:24
Gonzalo
yes
10:25
Deleted Account
wow so fast
10:25
Sticker
Not included, change data exporting settings to download.
👍, 34.6 KB
N
10:26
NM$L
what price do you predict at EOY？
G
10:27
Gonzalo
who knows!
10:28
between 0 and 500?
10:29
Deleted Account
In reply to this message
waht is EOY?
D
10:30
D
In reply to this message
End of year
10:30
Deleted Account
thanks
JS
10:35
Jon Snow
The more realistic question is if this chain will survive at EOY.
10:37
Deleted Account
has the fork ahppened already?
10:44
Deleted Account
yet determined I think. maybe on height 9000.
10:44
Deleted Account
can i using multiple computers in parallel, using cpu to provide hash power
10:45
Deleted Account
cpu mining is obsolete now...
10:45
too much total hashpower
10:45
Deleted Account
yes so to parallel multiple pc
10:45
Deleted Account
How many cores? 1000?
10:46
Deleted Account
😂
10:46
4 cores
10:46
Deleted Account
sad but truth
10:46
only first few days were the cpu era
10:47
Deleted Account
sad
10:48
Deleted Account
If you do not own gpu, still you can mine on AWS gpu instance. It's a bit of gamble though
10:49
looks like the number of synced node is increased. Now 11.
10:55
Deleted Account
ok thanks
10:57
Deleted Account
what is current hashrate and block reward
10:57
is it still using same algorithm?
10:59
Deleted Account
Same algo, 4.3TH/s, block reward 1 veo
http://159.65.120.84:8080/explorer.html
P
11:08
Pan
keys:change_password(''old passwd'', ''new_passwd'').

doesn't work!
11:09
Deleted Account
hi all - not sure what i'm doing wrong
11:09
windows - NVidia miner
11:09
think I installed all dependencies
11:09
C:\Users\Mike>AmoveoMinerGpuCuda.exe BGN1llT3FAQxtl+zEwEdJedD5daI9EHyGiy6DmbOWyrBtXIPFRAscjzoNhEFeCAwk1Mhdvcu8zrttwmn8FUI9dQ= 0 'AmoveoMinerGpuCuda.exe' is not recognized as an internal or external command, operable program or batch file.
P
11:09
Pan
return errors:
keys diedok
start key
11:09
Deleted Account
typed that into cmd prompt
M
11:10
Mike
If you put the exe first and address second it should work as a batch file
Z
11:10
Zack
In reply to this message
thanks for sharing this error report.
11:10
Deleted Account
Run as admin
11:11
Deleted Account
C:\Users\Mike>AmoveoMinerGpuCuda.exe Access is denied.
11:11
says can't run on my pc
11:11
hmmm
M
11:11
Mike
In reply to this message
I have a file AMOVEO and I unpacked the miner 1.0.0.7 there. Just gotta open notepad and make the batch file
11:11
Deleted Account
AmoveoMinerGpuCuda.exe <Base64AmoveoAddress> <CudaDeviceId> <BlockSize> <NumBlocks> <RandomSeed> <SuffixMax> <PoolUrl>
M
11:12
Mike
All the batch file is is “miner.exe address=“
11:12
Deleted Account
U can do AmoveoMinerGpuCuda.exe <Base64AmoveoAddress> and it will defualt
11:14
Deleted Account
just created a batch file - won't open... hmm
M
11:15
Mike
You’ve got the CUDA 9.1, newest Nvidia drivers and Visual Studio 2017?
11:18
Deleted Account
looks s like it's working now
11:18
probably just need to tweak settings since takes up all of cpu
11:18
thanks all!
11:18
do I need a different bat file for each gpu
11:18
or all will run with the one?
M
11:18
Mike
That I wouldn’t know
11:18
It cripples my cpu yeah
11:19
Deleted Account
anyone test settings w/ 1070ti?
MF
11:19
Mr Flintstone
@zack getting block height 800 from your light wallet
11:19
Is it syncing?
Z
11:20
Zack
I just pushed the update to the master branch of the Amoveo full node.
To update you need to delete your dependencies.

You need the update before height 9000.
11:20
In reply to this message
yes, I just updated and am resyncing.
MF
11:20
Mr Flintstone
In reply to this message
cool, thanks
11:25
Deleted Account
Anyone know what hash rate should be on 1070ti?
11:28
Also - does anyone know if a different bat file is needed for each gpu?
P
11:31
Pan
Zack,
after installed full node amoveo,
'api:acount().' returns
empty
'api:pubkey().' returns
<<''B ...... =''>>
this two seem to be conflict?
How can I find my private key?
MF
11:35
Mr Flintstone
In reply to this message
I think so. Not a miner so not 100% sure
11:36
I know for certain miners you need to open a new terminal for each device id
P
11:37
Pan
Still puzzled!
I tried 'api:account(Pub).' with replacing Pub by public key.
Then error again:
''exception error: no function clause matching .... ''
Z
11:38
Zack
In reply to this message
The private key should stay encrypted. It doesn't let you take it out.
11:39
Deleted Account
thanks
11:39
just need to get it stable now
11:39
keeps crashing
11:39
and getting a ton of duplicate nonce errors
P
11:40
Pan
In reply to this message
Then how can I move my wallet in full node to other node? or how can I backup it for safe?
Thanks!
Z
11:42
Zack
In reply to this message
save the encrypted private key file. in _build/prod/rel/amoveo_core/keys/keys.db
11:43
In reply to this message
the command works for me.
(amoveo_core_prod@ubuntu-s-2vcpu-4gb-fra1-01)4> api:account(keys:pubkey()).
{acc,145424163,795,     <<4,245,141,14,218,146,197,92,121,207,82,203,28,167,247,246,75,58,51,75,145,48,133,152,84,...>>,
     1,  <<7,106,39,199,158,90,206,42,61,71,249,221,46,131,228,255,110,168,135,43,60,34,24,...>>}
P
11:44
Pan
In reply to this message
OK👍
J
11:44
Jim
In reply to this message
Need to set a different random seed for each GPU instance
11:46
Deleted Account
great - do I need to fill out all values for AmoveoMinerGpuCuda.exe <Base64AmoveoAddress> 0 64 96 0 (one for each GPU) <SuffixMax - what's good to use for this if experiencing lots of lag with default?> <PoolUrl>
11:47
better put: does anyone have an example bat with all mods filled out?
11:47
thank you!
11:50
tried 50k and still tons of lag
J
11:51
Jim
./AmoveoMinerGpuCudaLinux <address> 0 1024 64 randomseed1 1000
./AmoveoMinerGpuCudaLinux <address> 1 1024 64 randomseed2 1000
./AmoveoMinerGpuCudaLinux <address> 2 1024 64 randomseed3 1000
./AmoveoMinerGpuCudaLinux <address> 3 1024 64 randomseed4 1000
11:52
change for your setup
11:54
Deleted Account
Awesome thank you!!!
11:57
yep - way better now - awesome!
P
12:05
Pan
In reply to this message
Sorry, I used address not purely pubkey.
After fixed this, the command still returns:
''empty''

and also, ''keys:keypair().'' returns:
''none''
12:05
Deleted Account
geez - seems like only one GPU is mining
12:05
anyone know why only one GPU is running with different seeds for each gpu setup?
Deleted joined group by link from Group
12:09
Deleted Account
I have four cmd prompts open that seem to be working
12:10
but definitely only one gpu running....
N
12:11
NM$L
set the different deviceid
12:12
In reply to this message
how do you check
Z
12:13
Zack
In reply to this message
yes, it should be empty.
i didnt know "keys:keypair()." was a valid command. it should probably be deleted. Why did you do that? is it in the docs somewhere?
12:15
In reply to this message
oh yes. keypair is only for test mode. it is supposed to return "none" if you are connecting to the testnet.
N
12:16
NM$L
Zack when can I see all my address transcation
12:16
I lost 1 veo
Z
12:17
Zack
In reply to this message
you can see a few recent txs here http://amoveopool.com/transaction
12:17
In reply to this message
so what is the pubkey?
12:17
Deleted Account
In reply to this message
msi afterburner
12:17
gpu usage at 0
N
12:18
NM$L
In reply to this message
pm you
12:21
Deleted Account
anyone have any advice on how to get all my gpus running? sucks to only have one...
12:35
Deleted Account
why is it so many seconds per block lol
JM
12:36
J M
nvmd
N
12:38
NM$L
I lost 1 veo...
12:41
Deleted Account
where can find the total supply and current supply?
N
12:43
NM$L
who can help me check my address transcation history?
M
12:43
Mike
In reply to this message
Total supply isn’t something that is determined at the moment. It depends on governance
12:47
Deleted Account
is it always find a block 10 min and 1 veo reward a block?
M
13:20
Mike
In reply to this message
No, it’s rather complicated. There’s voting on block size. Check out the GitHub, there’s 4 or 5 hours of great reading. Zach has done a great job.
13:54
Deleted Account
Is there a veo discord?
G
14:10
Greg
In reply to this message
looks like $39 to me https://discord.gg/a52szJw
14:56
Deleted Account
In reply to this message
i lost 1 too
S
14:56
Sy
How do you loose veo except by running a pool?
14:57
Deleted Account
i didn't save my private key and pubkey
14:57
and reinstall my PC
S
15:09
Sy
Oh.... 😅
15:13
Uhh...node still synced :D
G
15:29
Gonzalo
Zack Zack one node lost sync, I've updated + prod-clean, now headers are 0 and sync:start says nothing to sync
15:29
you said we have to clean dependencies...
S
15:29
Sy
hmm i updated and didnt do clean...seems to be working fine
G
15:30
Gonzalo
I was out of sync, thats wgy I cleaned
15:30
why
S
15:30
Sy
ah k
15:30
lost your peers maybe?
G
15:30
Gonzalo
no, peers are there
15:31
How do you uninstall dependencies? anyone helps this poor windows guy? 😂
15:32
In reply to this message
^^^
N
15:44
NM$L
In reply to this message
I don't know why I lost..
15:44
It's strange.
G
15:46
Gonzalo
In reply to this message
Zack asked you pubkey before, so he can trace it
N
15:49
NM$L
In reply to this message
He said he can't
S
15:56
Sy
running a pool?
Z
16:02
Zack
In reply to this message
Save your keys.db file first.
Then rm -rf _build
G
16:27
Gonzalo
great, thanks!
16:28
In reply to this message
yes
S
16:34
Sy
if you are running the original pool then you might have found an orphaned block but neither the pool nor the node check for that, you pay anyway
G
16:35
Gonzalo
syncing 👍
16:35
thanks will check
S
16:38
Sy
its kinda hard to find out without a real block explorer that shows blockfinds aswell for your address but it goes the other way too, at least with the old node your pool told the node to pay up and it sometimes just didnt happen
G
16:38
Gonzalo
synced 😁
S
16:38
Sy
its up and down for my pool, had -3 and -4 after a good nights sleep but the overall is plus 😝
P
16:39
Pan
Zack,
does the following two miner support the coming hard fork??
which one products more hashrate?

https://github.com/decryptoed/amoveo-cuda-miner

https://github.com/zack-bitcoin/AmoveoMinerGpuCudaLinux
S
16:39
Sy
they are the same, first one is windows, second linux
G
16:41
Gonzalo
In reply to this message
I guess that if you run your own node + pool it doesn't matter. Veos will be either in your pool or your node
S
16:45
Sy
how can veos be in my pool?
16:45
the pool doesnt do much than giving out work and submitting potential blocks to the pool and telling him at the same time to pay the finder
16:45
to the node*
G
16:45
Gonzalo
if your pool dont pay your node, then your pool is keeping the veos
S
16:45
Sy
the pool never holds any veo
16:46
na...
P|
16:46
Paul N. | BAND (🅑) - I'll never DM you
In reply to this message
lol
S
16:46
Sy
yes paul?
P|
16:47
Paul N. | BAND (🅑) - I'll never DM you
I'm curious as well. What's your pool's private key?
S
16:47
Sy
your pool got none
16:47
no pool does ever, every pool out there is connected to a node / wallet
16:47
all a pool does it gets the hash from your node, sends it out to the worker
G
16:47
Gonzalo
not sure but I think that the pool gets the block reward and at the same time pays the node. That's why you need to have a positive balance in your pool before you start.
S
16:47
Sy
and submits a nonce good enough to be a block back to the node
16:47
nope
G
16:48
Gonzalo
and that's why you can loose money as well running a pool
S
16:48
Sy
a normal pool then adds up each submittet share and pays every user a fraction
16:48
sorry but no, read up on it
16:48
thats now how any pool works no matter the coin
16:48
you wouldnt need a node running if this were the case
16:49
but in general pools check back with the node if the block got accepted, wait x confirms and THEN tell the node to pay the finder but this isnt the case here
16:49
and you dont need positive balance unless you set your pool payout to 100% and pay tx fees on top which...nobody does
G
16:49
Gonzalo
yes youre right, I was explaining it wrong
16:50
when I say pool I mean your "pool+node"
P|
16:50
Paul N. | BAND (🅑) - I'll never DM you
I think the assumption here is that the *pool* always come with a *node* which holds a key (and veos mined through it)
S
16:50
Sy
paul...thats no assumption, its a technical fact
G
16:51
Gonzalo
yes
S
16:52
Sy
so back to topic, if your pool got a nonce submitted that qualifies for a block it sends it to the node and tells the node to create a api:spend for the finder which the node does, it submits the block to the network and instantly pays the finder
16:52
if that block is orphaned it still pays the finder thus you are -1 VEO
G
16:53
Gonzalo
so If you run a pool+node on a computer, and a miner on another, if your miner does not get paid, your money will be at your "pool+node", that's what I was trying to say
S
16:53
Sy
it rarely happens with a 10 minute blocktime coin but we were way below that so...it did happen a few times
16:53
yep but it should never happen that way around
16:53
since you are basicly cheating your miner out of their reward
G
16:53
Gonzalo
In reply to this message
right
16:53
In reply to this message
right
S
16:54
Sy
it will be interesting to check once a block explorer is around how often a pool found a block and didnt send out any....
G
16:55
Gonzalo
I guess Zack wrote a really basic pool, just for demonstration purposes. Its not a shared pool, it pays without verification, etc.
S
16:55
Sy
because right now with the original pool software and explorer there is no way to check
16:55
yep, the whole pool code is 129 lines with at least 10% wasted on output ^^
16:55
im not blaming zack, i knew that from the beginning, im just stating how it is so ppl dont get confused
G
16:56
Gonzalo
In reply to this message
👍 it's good to know
N
17:11
NM$L
17:12
why not valid?
G
17:12
Gonzalo
wait and refresh
17:12
sometimes happens
Z
17:23
Zack
In reply to this message
I think none have been updated. It is a small change.
17:24
In reply to this message
Sy is correct.
17:25
In reply to this message
Gonzalo is correct too. But he is calling the full node + mining pool software a "pool"
G
17:26
Gonzalo
yes, hope I did not confuse anybody
Z
17:26
Zack
In reply to this message
Correct
17:26
In reply to this message
Correct
S
17:27
Sy
yeah we were talking roughly about the same, not paul tho 😁
17:27
and afaik no miner has been updated yet
17:43
Deleted Account
horrible hash power
17:43
is he here
Z
17:56
Zack
In reply to this message
I updated my c miner as an example. It is in the partial-hash-fix branch
T
18:07
TownAlwaysGet
Is there something wrong with mining? It seems inaccurate details...
P|
18:07
Paul N. | BAND (🅑) - I'll never DM you
In reply to this message
lol
Z
18:19
Zack
In reply to this message
Mandel s pool is measuring hashrate inaccurately
M
18:46
Marc
Hi! I did a tx via light wallet at http://159.65.120.84 yesterday but it isn't processed yet. Should I wait or try again?
Z
18:56
Zack
In reply to this message
What is your pubkey? which pubkey did you send funds to? Did you use "create_account" or "spend"? How much did you send?
M
18:59
Marc
pm'd
S
20:07
Sy
until what block was the reward 1.5?
Z
20:12
Zack
was able to fix the hash2integer part of this https://github.com/zack-bitcoin/AmoveoMinerGpuCudaLinux/tree/partial-hash-fix

But I am a little lost on how to fix the calculation of the bytes that should be hashed.
20:13
In reply to this message
it was 1.5 for about 2000 blocks I think. I can't remember exactly when.
N
20:14
NM$L
what's next diff adjust?
20:15
when and how?
G
20:15
Gonzalo
@Iridescence was coding a block explorer, we will see there
20:16
In reply to this message
every 2000 blocks, so next will be at 10000 I suppose
N
20:16
NM$L
how is the diff change?
20:16
Gonzalo
G
20:17
Gonzalo
next diff I dont know
N
20:18
NM$L
In reply to this message
will be more difficult or not?
Z
20:20
Zack
In reply to this message
It adjusts to try and keep the block time at about 10 minutes.
G
20:20
Gonzalo
you can calc an estimation, take the last say 100 blocks, sum the time and divide by 100. If its less than 10 minutes per block you can estimate that yes, it will be more difficult
Z
20:20
Zack
In reply to this message
I tried to fix the mining function now too.
I need to compile and test it for the changes after the hard fork.
20:25
I increased the reward for mining in my pool to 97%
N
20:26
NM$L
what's your pool original reward?
Z
20:27
Zack
It started at 70%, and has been increasing.
N
20:27
NM$L
what's your pool diffferent with mandel's?
Z
20:35
Zack
Both pools are now claiming to charge a 3% fee.
20:44
I launched this testnet that uses the new pow function: http://159.89.106.253:3010/explorer.html?en

I also launched a mining pool connected to that node. on port 8085
The difficulty is very low.
This is an opportunity to make sure your miners will work after the hard fork at 9000.
N
20:51
NM$L
In reply to this message
set pool url = http://159.89.106.253:8085 right?
Z
20:51
Zack
In reply to this message
that is a testnet. any coins you mine are worthless.
N
20:52
NM$L
error:getwork:500 sleep and retry
20:52
In reply to this message
I know worthless.
Z
20:53
Zack
In reply to this message
which software did you use? I don't think anyone has written gpu software for this yet.
G
20:53
Gonzalo
@windance how many GPU have you got?
N
20:54
NM$L
In reply to this message
mandel's software
20:54
haha
20:54
my fault
G
20:54
Gonzalo
😁
N
20:54
NM$L
In reply to this message
many many..
G
20:55
Gonzalo
many many, ohhh many many tra la ra la laaaa, oh many many 😂
21:34
Deleted Account
hi there
21:35
how do i create key pairs offline?
21:35
create them using secp256k1 and then?
21:39
more precisely, the pairs i generate using openssl/secp256k1 does not look like amoveo addresses
MF
21:41
Mr Flintstone
You can use the light wallet
21:41
offline
21:42
and input your own entropy
21:42
do you have any issues with this method?
S
21:56
Sy
the pubkey is base64 encoded
22:06
Deleted Account
In reply to this message
no but i was curious about the format
22:07
In reply to this message
after encoding it can i use it right away? do i need to setup a light node?
S
22:08
Sy
You can easily test your key generator by putting the privkey into a text file and upload it to one of the lightwallets

If it displays your oubkey base64 encoded then it's valid
22:09
AmoVeo Mining Pool
http://amoveopool.com/Wallet
N invited Dan Notestein
22:13
Deleted Account
In reply to this message
thank you
S
23:44
Sy
my node is off for now, node kept spamming that the pool found a block, clean restart doesnt sync anymore since api:height(). returns 0 and im out of town for the next days
OK
23:46
O K
bummer
MF
23:47
Mr Flintstone
if you want to share, how much hash rate were you contributing?
OK
23:48
O K
Are any of your people switching to another node
14 March 2018
G
00:02
Gonzalo
In reply to this message
you can 2 two things:
00:04
1) reinstall cleaning dependecies as Zack said: 
Save your keys.db file first.
Then rm -rf _build
00:04
2) point your miners to my pool at 2%: 51.15.75.100:8085 😁
00:06
Deleted Account
Hey - is there anything specific we need to do for the fork?
S
00:09
Sy
I was doing around 70gh I think and yes, some are still mining
P
00:17
Pan
What is the price of veo now?
00:18
In reply to this message
I my mind, you are not the owner of amoveopool.com, am I right?
OK
00:21
O K
He is not
P
00:21
Pan
In reply to this message
👌😆
00:22
Deleted Account
what is current hashrate?
OK
00:22
O K
block:hashrate_estimate().
00:23
I believe it's also on the explorer if you choose get height
MF
00:38
Mr Flintstone
In reply to this message
Getting block height 8129 on this
00:39
See 8153 on zacks
00:39
In reply to this message
see 10 T
00:39
Block times about 7 minutes
OK
00:40
O K
In reply to this message
Thanks for pointing that out, I am resyncing
00:43
actually I can't seem to get any headers now
M
00:45
Marc
is there a way to get block:hashes_per_block(), etc. via command line? i would like to have a munin plugin to graph it.
MF
01:22
Mr Flintstone
still at 8161?
02:31
Deleted Account
Given 0.3 GH/s on a 1080TI and a price on about 1000 dollar per card, the total value of the current net hashpower is worth about 45 mill dollars. Kinda crazy.
P|
02:32
Paul N. | BAND (🅑) - I'll never DM you
In reply to this message
It's actually closer to 2.2 - 3 GH/s on 1080 Ti
I
02:32
Iridescence
1080Ti should be able to get more than 0.3 GH/s
Z
02:32
Zack
In reply to this message
I think some people can get more hash power than that out of that card
02:32
Deleted Account
alright, so about 4.5 mill dollars then
Z
02:33
Zack
And most of the cards are probably rented, not owned
02:33
Deleted Account
still crazy for a couple of weeks old currency
I
02:33
Iridescence
It's crypto
G
02:33
Gonzalo
Im running 2 1080ti at the same time, one gets 1,8, the other 2,2
I
02:34
Iridescence
Somebody got 3.5 GH/s on 1080Ti using my optimized miner
02:34
Deleted Account
Alright, have lost track of the hasing power of the 1080TI cards the last couple of days
02:34
now it is 48 GH/s per day for 1 veo
MF
02:40
Mr Flintstone
anyone have a rough idea how much it costs to mine 1 veo now?
P|
02:41
Paul N. | BAND (🅑) - I'll never DM you
In reply to this message
Close to $100 for rented hardware
M
02:41
Mike
In reply to this message
No way
02:41
Probably like $6
P|
02:42
Paul N. | BAND (🅑) - I'll never DM you
In reply to this message
How do you come up with that?
M
02:42
Mike
I get like .55 VEO a day with a 980
P|
02:43
Paul N. | BAND (🅑) - I'll never DM you
In reply to this message
What's the GH/s for the 980?
M
02:44
Mike
In reply to this message
Around 2 GH I think
02:44
It’s changed so much with Mandel’s updates so I’m not so sure
MF
02:44
Mr Flintstone
2/10000 times 144 equals .03
02:44
that is not 0.55 per day
02:45
these are rough numbers, but we can get a sense for order of magnitude
02:45
I guess the 10000 should be like 5000
02:45
But still
02:46
and maybe you comp for the fact that blocks are twice as fast, so this would be like 0.12 veo a day?
02:46
not sure how accurate my math is
Z
02:49
Zack
11 terahashes per second. At this rate we will reach 9000 in no time. I hope everyone will be ready.
02:50
How much to rent 1 gh/s?
02:54
I think 50 gh/s is 1 Veo per day right now.
OK
02:54
O K
Sounds too high
M
02:55
Mike
I mined .25 at least over night with one card
02:55
Deleted Account
sounds too low
Z
02:56
Zack
If it is 4329 terahashes per block, then it must be 50 gh/s to find 1 block per day.

So I guess my estimate of 4329 th per block must be wrong.
02:56
In reply to this message
This was after block 8000?
M
02:57
Mike
In reply to this message
I’ll check when I get home
Z
02:58
Zack
We hit 8000 less then 24 hours ago.
02:58
I think
MF
02:59
Mr Flintstone
we hit 8000 this morning? Like 8am or do ET?
02:59
not 100% sure
02:59
or so*
02:59
or maybe that was yesterday
Z
02:59
Zack
It was 12.5x increase.
So I think your profitability makes sense before 8000
OK
03:00
O K
I'm only, allegedly, at like 4 GH/s ... and I'm definitely not at 10 days per Veo
M
03:00
Mike
Makes sense
Z
03:00
Zack
You only needed 4 gh/s to make 1 Veo per day before
03:01
Deleted Account
I think it is 48 GH/s for 1 veo per day now
M
03:01
Mike
Damn. That makes it at least $50 to mine a VEO
03:02
Deleted Account
amoveopool is showing wrong numbers?
MF
03:02
Mr Flintstone
anyone have a sense of how much 1 gh/s is to rent on AWS or some other provider?
03:02
Deleted Account
3 GH/s is about 1 dollar an hour, cheapest I found on google, but maybe someone has better numbers
03:03
Deleted Account
thats my cost
MF
03:04
Mr Flintstone
so 16 dollars per hour to maintain 48 gh/s
03:04
that’s like 400 dollars per day? Am I missing something
03:04
Deleted Account
for 1 veo uhhh per day
03:04
Deleted Account
hmm, that sounds wrong. I estimated 140 dollars per veo in mining cost like 10 minutes ago
03:05
oh well, I am confusing myself
OK
03:05
O K
I am 99% confident I have mined at least 0.25 Veo since block 8000 on 4GH/s
03:06
Deleted Account
difficulty update didnt happen? :P
MF
03:06
Mr Flintstone
any easy way to check the time stamp on 8000?
Z
03:06
Zack
I could be estimating the hashes per block incorrectly. The math is a little complicated.
OK
03:07
O K
My memory matches yours Flinstone, I was having my morning coffee yesterday when it happened
03:08
I'm thinking like 28hours
MF
03:08
Mr Flintstone
ok, call it a day
03:08
so 16 gh/s for one veo per day?
OK
03:09
O K
In reply to this message
Yesterday about 8 EST
MF
03:09
Mr Flintstone
Which is like 130 usd for renting
03:09
Per veo
03:10
this number is probably higher
Z
03:10
Zack
Maybe the era of rented gpu is over?
Or someone could be finding better deals by renting in bulk from mining pools
03:11
Deleted Account
or Veo price will increase accordingly
JS
03:11
Jon Snow
We need ASIC
G
03:19
Gonzalo
I'll mine tonight to 1 new pubkey, so I can do the calcs, will tell you tomorrow
M
03:20
Mike
In reply to this message
Renting directly from miners is cheap, just difficult. No one trusts projects they’ve never heard of
OK
03:24
O K
Does TH/block = GH/s (avg) * Seconds/block ?
03:25
Deleted Account
sounds good
03:27
Deleted Account
im really confused
03:27
maybe hash rate only 913.5
MF
03:33
Mr Flintstone
In reply to this message
I would think so
03:34
this is what zacks page does, it looks like
Z
03:35
Zack
In reply to this message
No. That isn't how you cancel units.
When seconds / seconds cancels, you are left with gh/ block
MF
03:35
Mr Flintstone
I think he was just ignoring the multiplication by 10**3
OK
03:36
O K
Well right, I mean GH to TH is trivial I should have left off the G and T
MF
04:02
Mr Flintstone
is the difficulty of a block stored in the block header?
Z
04:28
Zack
In reply to this message
Yes.
MF
04:30
Mr Flintstone
In reply to this message
can’t we convert this to the number that the hash needs to be less than to be considered valid? then we can easily figure out how many hashes per block
04:30
or is this less straightforward
Z
04:34
Zack
Mandel found the mistake in my hash per block function. I will fix it soon, then we will know the actual hashrate.
MF
04:35
Mr Flintstone
a moment of truth approaches
Z
04:35
Zack
In reply to this message
Right. That is the challenge.
MF
04:35
Mr Flintstone
I assume this transformation is actually embedded in the node code
04:35
or the chain would never be able to validate a block
04:36
the node would never be able to*
Z
04:36
Zack
When we check if the difficulty is valid, it is stored in a sort of scientific notation which is optimized to be easily calculated by looking at the hash of the header.
MF
04:37
Mr Flintstone
o I c
Levi Fishman joined group by link from Group
CARS 10 joined group by link from Group
Z
04:50
Zack
http://159.65.120.84:8080/explorer.html
There, now it estimates correctly.
OK
04:52
O K
Based on the new numbers what is the theoretical hashrate needed for 1/day?
Z
04:53
Zack
Here is the gigahashes per block at all the difficulties so far:
37, 786, 22517, 88740, 301496.

So the difficulty has changed by these ratios:
21.2, 28.6, 3.9, 3.4
CB
04:54
CeloStarter Bandits
does the hardfork requietr any change to the miners who mine against mandelhoff?
JS
04:54
Jon Snow
So Mike is not wrong on the cost? 6$ each?
CB
04:54
CeloStarter Bandits
what will change? exactly?
Z
04:54
Zack
301 th/b = X th / s * (seconds per day)
-> 3.5 gh/s to find 1 amoveo per day.
M
04:55
Minieep21
What sounds better?
VeoMarket
AmoveoMarket
?
Z
04:55
Zack
In reply to this message
Yes. Mandel says he will release an upgrade before height 9000.
CB
04:56
CeloStarter Bandits
and we can still mine with GPU
04:56
?
OK
04:56
O K
Totally
Z
04:57
Zack
In reply to this message
If Mandel releases an update in time, and you install the update, then you will be able to mine with GPU.
MF
05:00
Mr Flintstone
if anyone wants to sell me their veo pm me
J
05:00
Jim
In reply to this message
Excellent, this is much more in line with my “gut instinct”
Z
05:00
Zack
In reply to this message
Maybe you can re-make your chart now that we have accurate data?
J
05:01
Jim
Yes I will, and make it live as well. I wrote some php for this, now to make it cache so that it doesn’t hammer my server for every lookup
OK
05:01
O K
very cool
J
05:02
Jim
Is there another curl statement to get difficulty / hashes per block?
M
05:02
Mike
In reply to this message
I wasn’t accounting for the difficulty adjustment
M
05:03
Marc
yeah. i would like to have that, too. website is js so curl and wget dont work.
MF
05:04
Mr Flintstone
In reply to this message
so I guess like 20 dollars or so per veo to mine
AS
05:07
Aizen Sou
In reply to this message
No way. How did u do ur math ?
S
05:08
Sy
In reply to this message
That doesn't sound correct tbh...
05:09
It feels way less
05:09
I mean what you get 😁
MF
05:12
Mr Flintstone
In reply to this message
it cost mike 6 dollars per day per veo before the diff adjustment
Z
05:12
Zack
In reply to this message
it only changes every 2000 blocks.
You can download the entire block in curl, and then take the difficulty out of the block, or I could add another command if you think it would be useful.
05:15
In reply to this message
curl is a standard way of writing it so that everyone in every language can understand.
You can use my library of http requests in javascript as an example if you haven't done this before.
Here is the library: https://github.com/zack-bitcoin/amoveo/blob/master/apps/amoveo_http/priv/external_web/rpc.js
here is using the library to access the external port of the node: https://github.com/zack-bitcoin/amoveo/blob/master/apps/amoveo_http/priv/external_web/create_account_tx.js#L27
it uses callbacks so that the api is asynchronous.
M
05:18
Marc
thanks for your reply. i will try that. i want to code a munin plugin for the stats. because munin is great. 😉
Z
05:22
Zack
In reply to this message
How much less?
The mining pool has about 4 seconds delay, the miner has about 10 seconds delay, and their is a delay for block propagation.
All that together might be slowing you down.
S
05:25
Sy
Are we talking about new or old diff?
M
05:28
Mike
Extrapolating the past 4 hours where I’ve mined .0251, should get just about .15 VEO a day on a 980. 6.5 days to get a full one. I’ll give a 980 an optimistic $2 return per day on NiceHash. $13 per VEO with new difficulty
05:41
Deleted Account
is amoveopool down?
JM
05:41
J M
no
05:42
Deleted Account
ok
05:43
the total supply is 120-180 per day？
MF
05:51
Mr Flintstone
In reply to this message
currently 1.2 veo per block. Block time is targeted to be 10 minutes
05:51
but is currently about 5:30
OK
05:51
O K
1.2?
05:51
When did it go up?
MF
05:52
Mr Flintstone
0.2 dev reward I believe
Z
05:52
Zack
He is including the developer reward.
OK
05:52
O K
Oh I see
05:52
👍
05:52
Cool
07:05
This should auto-update provide google sheets keeps working and my node is up
M
07:07
Marc
cool. big up!
M
07:09
Minieep21
Wow that difficult spike is insane
07:09
Thanks for this :)
G
07:09
Gonzalo
looks great!!
Z
07:09
Zack
In reply to this message
very cool
J
07:15
Jim
Code for this project (php): https://github.com/jimhsu/amoveo-stats-scraper - very basic stuff
MF
07:19
Mr Flintstone
thanks Jim
07:20
it looks like the axis is using the old hash rate? the one based on the wrong hash/block number
07:21
the right y axis
J
07:34
Jim
Yep, let me update my node
M
07:38
Minieep21
I now own AmoveoMine.com
Gonna try and setup a pool over the next few weeks :)
G
07:39
Gonzalo
👍👍
07:56
Deleted Account
im guessing the algo didnt change yet?
07:57
hm only 570 gh/ s now
J
07:58
Jim
Yeah, actually a moderate drop on the graph. Remember though that amoveopool was down for a while yesterday
07:58
Deleted Account
xo fonsued whats going on righ tnow
07:58
but its the same gpu mining algo right lol
07:58
it was 3 th/s now only 570 gh/s so confused
J
07:59
Jim
Yes, the previous calculation was wrong
07:59
Deleted Account
oh waht do you mean
07:59
what was the previous hashrate then?
08:00
total net hash was like 3 th/s , and now it say 570 gh/s
08:01
Deleted Account
ok so the net hashrate was wrong before
J
08:02
Jim
Yes
N
10:23
NM$L
sell some veo. pm me
10:28
10:28
Is there some imprecision about how diff is converted to a human readable value?
10:30
10:31
I'm guessing the human readable version is "probably wrong" instead of the implementation, but I wonder if we can get a more accurate prediction. In any case, I don't think diff 12287 is actually 128 times more difficult than 12288
10:36
10:36
Old vs new. Note, exponential trendline is the same. I guess that's close to the "actual difficulty", no?
10:38
What I don't know is what the y-axis intercept for the trendline should be
J
10:54
Jim
10:54
It should be more like this, no? Unless I'm not understanding the code right
10:54
That, is,
=POWER(2,A)*POWER(2,B/256) / (10^12). Or this is simply power(binary difficulty / 256) if your language supports floating point division which erlang does not "by default"

I think, I don't know erlang
10:56
Deleted Account
da algo is the same? so confused
J
10:56
Jim
I'm confused and will wait for Zack to respond
10:57
The algo is the same, the human readable version is where the problem is
10:58
Deleted Account
what do you mean human readable
10:58
such confuse, much wow
J
10:58
Jim
The stuff that says gigahashes/sec or terahashes per block is human readable. The "actual" diff is binary (for instance, the "real" diff right now is "12305")
10:59
The function diff2hashes converts that binary number into gigahashes or terahashes
11:00
The mining works on the real, or binary difficulty
MF
11:05
Mr Flintstone
maybe this is a stupid question but is there another meaning of binary I am unfamiliar with?
11:06
like how can 12305 be interpreted as binary
J
11:07
Jim
It has something to do with bits, like any mining algo, this looks for the number of bits equal to zero, or something
11:07
I'm not sure how the 12305 translates into that equation though
11:12
In reply to this message
Hm, that doesn't work for erlang (division w/ only integers). I actually don't know how to do this for integers.
11:19
Deleted Account
anyone have use the other miner?
MF
11:33
Mr Flintstone
is there some kind of Taylor series used
11:33
if it’s int only
11:34
idk
11:41
Much easier would be to just make that particular part of the calc floating point, it's just human output stuff anyways and precision is not critical
MF
11:47
Mr Flintstone
yeah
12:32
Deleted Account
anyone that know basic erl how to do something?
P
13:13
Pan
I am selling some veo in OpenBazaar platform!
😄
13:14
Anyone wants to try one time?
address is

ob://QmP7p8MHNicxBjDjVbrwq8JTCkr4grkDWTeVXx9VW8GH7o/store/amoveo
13:23
Deleted Account
Zack Is testnet pool working? amoveo-c-miner partial-has-fix version seems to find a block but not accepted. Pool URL is -define(Peer, "http://159.89.106.253:8085/")
1> miner:start().
Started mining.
Found a block. 1
read nonce n is 2
server gave confusing response
13:26
I'm also trying to update AmoveoMinerGpuCuaLinux for new pow. Not sure yet it's working correctly...
https://github.com/tumblecatweed/AmoveoMinerGpuCudaLinux/commit/d666b57150151894e4adc0a7ec28d0729ccb9360
13:28
anyway I'm learning a lot about sha256 :)
S
13:37
Sy
Found a block doesn't really mean you actually found one, the original miner isn't so clear about that
13:44
Deleted Account
thx, I should write test locally
Z
14:58
Zack
In reply to this message
Yes. Hopefully this hard fork fixes it.
14:59
In reply to this message
Right now it is 128 times more difficult. This is part of the reason we are doing the hard fork.
15:01
In reply to this message
Erlang has floating point. Blockchains cannot use floating point because it is not deterministic.
15:14
In reply to this message
Yes, you are right. There is still a mistake in how it is being estimated in human readable form.

This is the formula being used to see if it is difficult enough for mining https://github.com/BumblebeeBat/pink_crypto/blob/master/src/pow.erl#L61
15:15
In reply to this message
Power only takes positive integers. Power(2, b/256) does not work because b/256 is not an integer.
15:17
In reply to this message
2 bytes. A is in range 0-255. B is in range 0-255. Total value can be up to 65,000 or so.
15:21
In reply to this message
No.
This is called "scientific notation".
It is a way to express big numbers in less space. You ignore some insignificant digits.
15:21
In reply to this message
There is no floating point in blockchains.
15:23
In reply to this message
Hmm. That should have worked. I'll look into it.
15:23
Deleted Account
Is the "fix" as simple as updating diff2hashes to use that pair2int formula from pink_crypto? I tried that out and diff2hashes does give a consistent increase in hash count with increasing difficulty
Z
15:27
Zack
In reply to this message
Maybe.
This does make it increase smoothly.
Z
15:53
Zack
Yes, I think That Guy is right.
16:01
ok, I pushed the fix from That Guy.
16:01
Deleted Account
Awesome, thanks for the quick fix 😀
Z
16:20
Zack
ok, I fixed the issue with the test version of the code so that we can practice making blocks with the new pow algorithm.

I can use the partial-hash-fix version of the c-miner to make blocks, so it is working now.
Baudoin Collard joined group by link from Group
Baudoin Collard invited Bastien
17:07
Deleted Account
Zack Thanks. I confirmed the miner can produce blocks on test net. It's in fix-pow branch.
https://github.com/tumblecatweed/AmoveoMinerGpuCudaLinux/tree/fix-pow
Z
17:07
Zack
In reply to this message
Awesome. Good job catweed.
N
17:17
NM$L
N
NM$L 14.03.2018 10:23:17
sell some veo. pm me
Z
17:17
Zack
In reply to this message
You are selling Veo? What price?
P
17:24
Pan
In reply to this message
😄I want sell little Veo, can you be moderator with some reward fees?
Z
17:25
Zack
In reply to this message
Yes
P
17:26
Pan
In reply to this message
I thank price among 50 to 100 usd each should be all resonable
17:27
We need an exchange now!😊
17:28
I think the trade of Veo will help the network hashratr to be more stable!
17:30
I just mined little with one i5 cpu in early days😂😂
17:30
when 1 veo cost about 1~2days
Z
17:33
Zack
In reply to this message
BCjdlkTKyFh7BBx4grLUGFJCedmzo4e0XT1KJtbSwq5vCJHrPltHATB+maZ+Pncjnfvt9CsCcI9Rn1vO+fPLIV4=
I can hold the Veo for security, but first you both need to tell me the bitcoin or eth address that will be receiving funds, and you both need to tell me how many veo and how many eth or bitcoin are being traded.
Wait for me to say "ready" before you send any money. I need to check and confirm that you are both intending to trade the same amount of tokens.
For now I charge 0.2 Veo fee to help make the exchange.
18:08
Deleted Account
There's one thing I'm not sure about.
https://github.com/tumblecatweed/AmoveoMinerGpuCudaLinux/blob/1a469b310e09e303ff368d0e8fec34360808f0d8/main.cu#L407
When I printed the nonce while debugging, The final result, nonceSolution is composed of 15 bytes (d_nonce) + 2 bytes (nonceSuffix) + 6 bytes (currentBlockIdx).
But I think it should be d_nonce + currentBlockIdx + nonceSuffix because they are sha256_updated in this order and copied inside sha256_kernel:
memcpy(out_nonce, &currentBlockIdx, 6);
memcpy(out_nonce + 6, &nonceSuffix, 2);

So the last two is reversed somehow. @Mandelhoff Please take a look ifyou have time.
MH
18:38
Mandel Hoff
The code looks right. Not sure why printnonce would look like it's swapped. I have nearly the same changes.
18:39
I have a new pow pool setup for testing. Diff is so low that I don't trust it and speeds are impossible to test, so I'm mining up to 2k for more testing.
18:41
Deleted Account
Thanks! I'm relieved a bit.
Deleted joined group by link from Group
19:33
Deleted Account
Noob question, I just started mining today, and I'm only seeing mining activity on one card, running Windows 10 64bit, and have 4 GTX 1070Ti's
M
19:42
Minieep21
In reply to this message
Mandel's miner?
You'll need to run separate instances for each card
OK
20:07
O K
In reply to this message
Awesome !!!!
20:07
Deleted Account
Got it! Thank you!
J
20:31
Jim
In reply to this message
20:31
Yep, looks good
20:32
Sheet also updated
20:35
Deleted Account
So just curious, I'm running 4 1070Ti's (should be getting a 5th here soon) what should I see daily (give or take)?
Z
20:35
Zack
In reply to this message
Great. Thank you Jim Hsu
20:37
It looks like 3.5 gh/s is 1 Veo per day.
How much does that cost?
20:38
Deleted Account
My gh/s rating is all over the board lol, I think my average is about 9-10Gh/s
OK
20:38
O K
If one V100 is $3 per hour and Iris miner gets...
20:39
6.3 GH/s
20:40
$72 per day,
20:41
1.8 veo
20:41
Deleted Account
Thank you :)
OK
20:41
O K
$40?
Z
20:42
Zack
So 1 gh/s costs $12per day by your estimate, or $8 per day by previous estimate.
So it is around $28-42 per veo.
OK
20:42
O K
Okay, additionally Iris miner is closed source and he charges a thousand dollars
J
20:42
Jim
Self-mining costs (electricity, depreciation) should be about 4-6x cheaper than Amazon, so ~10$ for a miner with existing rig
MF
21:00
Mr Flintstone
there has been a premium in the otc market unsurprisingly
M
21:20
Minieep21
I got a Google Cloud P100 currently mining at 1.63 GH/s. Cost of roughly $1.5 per hour
Z
21:21
Zack
In reply to this message
So you are paying $80 per veo?
M
21:27
Minieep21
Got $300 free credit. Had to pay $35 to use the P100.
21:29
I won't be paying after the free credit expires. Too expensive.
P
21:29
Pan
My friend begin mining veo by more than 1 hundred GTX 750TI GPU, which had about 32GH/S, but only get ~2 veo per day?
Why so bad1?
21:31
21:32
Deleted Account
@Mandelhoff I realized that little endian byte-order caused the swap when we memcpy to unsigned char array.
Also confirmed that this does not affect the final nonce by dumping and comparing ctx.data.
So it's OK I think.
Z
21:33
Zack
In reply to this message
Are you mining valid blocks on the test net?
21:34
Deleted Account
Zack Yes, It confirmed and appear on explorer.
Z
21:34
Zack
Then it is fine
P
21:40
Pan
In reply to this message
In this practice, 10Gh/s only products ~1 veo per day!
OK
21:41
O K
Pan, what miner is he using?
P
21:53
Pan
windows GPU miner from amoveopool.com
OK
21:54
O K
From github.com/mandelhoff? Are you sure you are using 1.0.0.7 not 1.0.0.0?
P
22:00
Pan
and, the parameter RandomSeed of each mining worker is set from 1 to 1xx ...
OK
22:04
O K
Try setting fewer variables, only .exe address gpu#
22:04
With one, do you get better hashrate?
AS
22:10
Aizen Sou
Zack 159.65.120.84 shows block height 8293 while amoveopool shows block 8305
22:10
Fork or ur node doesn't sync ?
P
22:14
Pan
In reply to this message
Yes, 1.0.0.7 version.
See above picture for hashrate of each GPU, seems OK.
22:15
In reply to this message
This picture shows hashrate, which seems OK
OK
22:16
O K
Are you sure they are all using the same address?
P
22:16
Pan
In reply to this message
Yes
22:22
Deleted Account
22:22
This is what I should be seeing right?
OK
22:22
O K
Looks good depending what card you have
22:23
Deleted Account
1070ti's
OK
22:24
O K
My 1070 doubles that
22:26
Deleted Account
Hmm, perhaps I've got something wrong?
OK
22:27
O K
My first guess would be you're not using the most recent release, second guess would be some wrong setting. I would set only the bare minimum settings and see if the defaults work better or worse than your configuration
22:27
Deleted Account
C:\Users\Rig1\Desktop\AMOVEO\AmoveoMinerGpuCuda.exe [wallet] 0 64 96 randomseed0 1000
OK
22:28
O K
Yeah, I don't use suffixmax1000
22:28
Deleted Account
So just take the 1000 out?
OK
22:28
O K
Try with one
22:29
You could probably also do up to 128 128
MH
22:29
Mandel Hoff
In reply to this message
Sounds great. Good work.
22:29
Deleted Account
Oh also, is there anyway to get it to stop pegging my cpu?
22:29
lol
OK
22:29
O K
Yeah... the low suffixmax I think...
22:30
Deleted Account
Ah damn
22:30
22:30
Test on bottom left
OK
22:30
O K
looks about double to me
22:31
Deleted Account
:D I appreciate it!
22:32
Looks like it may have frozen
OK
22:32
O K
perhaps there is a middle ground for you that would give you better hashrate, but not make your PC unusuable
22:33
Deleted Account
I'll try out some different things and see what happens
Z
22:34
Zack
In reply to this message
thanks for warning me. This means my node crashed.
22:34
Deleted Account
@Mandelhoff 👍
Deleted joined group by link from Group
M
22:59
Marc
is amoveopool pplns?
MH
23:03
Mandel Hoff
No, it is currently the Proportional model.
M
23:04
Marc
ah. oki. thanks for your reply.
23:25
Deleted Account
What's preventing people from bribing amoveo holders to vote for false outcome?
Z
23:25
Zack
In reply to this message
Amoveo doesn't use voting.
23:26
Deleted Account
I've had a discussion with a friend and he was asking,probably i need to understand it more.
So let's say that there is a bet that UFO landed in Paris in 2018
23:26
And people are putting their coins on "true" side?
23:27
"If you are uncomfortable with the outcome of an oracle, simply move to a fork of the blockchain that you think is honest. "
Z
23:28
Zack
First we make an oracle. Then we make a market that is connected to the oracle.
Until 2019, people will bet in the market only. They will not bet in the oracle.

Once 2019 has arrived, then people will bet in the oracle, because they know if the UFO arrived or not.
23:29
the market is for gambling on what might happen. The oracle is for making bets on what has already happened.
23:30
Deleted Account
How markets on side channel will work exactly?
https://github.com/zack-bitcoin/amoveo/blob/master/docs/design/state_channel_without_off_chain_market.md

Let's proceed with this example
23:30
People will be trading on probability of such an event?
Z
23:30
Zack
the smart contract is programmed so that the market has to publish the current market price every so many blocks.
All open trades which can match at that price are matched at that price.
If the market fails to publish in sufficient time, then it loses every bet.
Since I used a turing complete smart contract to program the market, anyone can write new market contracts with different rules, and we don't have to change Amoveo's source code or do a hard fork.
23:31
yes, the market lets people trade on whether or not an event will happen.
23:33
Deleted Account
So market is a person?
Z
23:33
Zack
it is a piece of software that some person is running.
23:34
everyone who wants to trade needs a channel with the market.
23:35
Deleted Account
Ok, so how this specific example can work?
With UFO - let's say i assume that such event have 10% probability
23:35
So i am creating some buy/sell order? What is priced?
Z
23:35
Zack
if the market says >10%, you should bet false. and if the market says <10%, you should be true.
23:36
i don't understand the question.
23:36
Deleted Account
Ok, so i am betting true (current state of market =1%)
So I am betting on true outcome and I get money from people that voted otherwise?
23:36
If oracle will give "true" answer?
Z
23:37
Zack
no one is voting.
If you bet correctly, then you will win money from people who bet wrong.
23:37
If the price is at 1%, then you could make your money 100x bigger.
23:39
Deleted Account
It's possible to withdraw money from bet before closing date?
Z
23:39
Zack
you can bet in the opposite direction, and then your 2 bets cancel out so you get the Veo back.
23:39
Deleted Account
"maintains it's lead for a long enough amount of time"
23:39
Long enough have precise definition?
Z
23:40
Zack
that is a minor implementation detail.
What is more important is that the community can do a soft fork to force the result of the oracle.
23:41
Deleted Account
Ok, such thing happened with Ethereum/Ethereum Classic
23:41
People will usually just follow "default" chain
23:41
And for some it may not matter if outcome of bet is blantantly false
Z
23:41
Zack
It isn't such a good strategy to follow a chain that lies sometimes. I would prefer a chain that is honest every time.
MF
23:42
Mr Flintstone
the ethereum / ethereum classic split is not black and white
23:42
like a false oracle outcome is
OK
23:42
O K
It was also not a soft fork
23:43
Deleted Account
Different forks may be traded independently?
Z
23:43
Zack
governance decisions aren't black/white either.

This is very experimental software. We may have to change rapidly to keep it alive, or it might fail.

Do not invest too much.
23:46
Deleted Account
Let's say that some people believe that one person never become a president and they are having fork that's true 99% of time except for this specific particular info
Z
23:46
Zack
the market says percentages. The oracle only says true/false/bad_question.
23:47
the market isn't on any fork. it is in the lightning network.
23:48
Deleted Account
I am trying to think about failure modes - right now I'm investing into understanding and maybe owning like $1000 worth of coins for testing and betting.

So people may fork the network on some question if they believe that outcome of oracle is not the same as reality - and somehow they more coins on forked chain? Right?
23:48
Deleted Account
Anything going on with amoveopool? Haven't gotten a p/o in 2 hours
Z
23:49
Zack
In reply to this message
they make bets in the oracle, and on the fork that they think is honest, they will have more tokens. Since they won the bets.
23:49
Deleted Account
In case of network fork the outcome is reversed? Bets are overriden?
Z
23:49
Zack
In reply to this message
it found a block 3 minutes ago, so I think it is fine.
23:50
In reply to this message
what do you mean by "network fork"?
23:50
Deleted Account
Of course as soon as I said something i paid out
23:51
Deleted Account
" Once enough honest individuals participate in the defense, then the oracle will catch the attention of the miners."
23:51
Deleted Account
Has difficulty gone up a lot as of late? A friend told me about this coin and has the exact same setup I do, and said he was getting about 3 coins per day
23:51
Deleted Account
" but every time an attack happens, the miners can participate in the defense, and double all their money"
Z
23:52
Zack
In reply to this message
I think it went up 2.5x about 2 days ago
23:52
In reply to this message
correct.
23:52
In reply to this message
correct.
23:54
Deleted Account
How exactly miners are doubling their money?
Z
23:55
Zack
by making honest bets in the oracle.
15 March 2018
OK
00:06
O K
Something that would help me with basic concepts is more hypothetical situations like @kacperwikiel was trying to lay out, except thought out by Zack
00:06
especially edge cases
00:07
In addition to"Users can bet", "whichever outcome" you could do both outcomes, If users bet A, and reality is A, if users bet !A and reality is A etc
00:09
Starting with the simplest case, everyone bets the same, and everyone was correct according the oracle
00:09
building up to the most complex distaster scenario
Z
00:10
Zack
I think Kacper was asking help because of a language barrier.
The oracle docs do talk about the possibility that people bet against reality in the oracle.
OK
00:13
O K
I was talking about what would help me
00:14
Can I get a show of hands for people that have read the docs and feel they have a comfortable understanding of how hypothetical cases would play out, and who would benefit, who has incentives to act how, etc?
MH
00:17
Mandel Hoff
I understand most of the happy-path stuff. I have not thought through almost any of the "attack" scenarios.
00:21
Deleted Account
Any amoveo sellers in here? My 2 1050tis are not making much progress :)
Z
00:24
Zack
In reply to this message
I am happy to answer any specific questions you have in the docs. But when you say "hypothetical cases", I have no idea what you want to know about.
OK
00:24
O K
In reply to this message
I know, being a part of this community has really taught me that I'm not as clear as I try to be sometimes
MH
00:25
Mandel Hoff
The document Zack linked a while ago about predictive markets in general is probably a must 1st-Read. It's a little academic, but it does walk through several example scenarios that are very helpful. I can't recall what the url for that was.
00:27
Found it on Zack's twitter posts:
http://bitcoinhivemind.com/papers/3_PM_Applications.pdf
00:27
In reply to this message
Oh right. That is one of my favorites
MH
00:27
Mandel Hoff
👍 Must read.
Z
00:27
Zack
Amoveo is my take on truth coin, nearly all the documentation for truth coin applies for us too.
00:28
Blockchain prediction markets are Paul Sztorcs idea.
OK
00:28
O K
I am going to read it, and if it does what I'm talking about with hypotheticals I will point it out
MH
00:49
Mandel Hoff
Do you all measure 1KH/s as 1000h/s or 1024h/s? I'm so used to measuring in byes that this "feels" very non-obvious to me.
Z
00:50
Zack
In reply to this message
A kilometer is 1000 meters.

I think the difference is if you capitalize the letters?
But I can't remember which is which.
MF
00:50
Mr Flintstone
@zack how has node stability been recently?
MH
00:51
Mandel Hoff
👍 Great node stability. I only had to restart once in the last 48 hours.
MF
00:51
Mr Flintstone
good to hear!
Z
00:52
Zack
Slower block times are forgiving, and the code is slowly improving
MH
00:52
Mandel Hoff
Yes, two nights in a row it stayed alive. It's like the first few times a new born baby sleeps through the night.
Z
00:52
Zack
Haha
00:54
Deleted Account
Yup, i'm not a native speaker and sometimes i need more verbose info to understand.
Z
00:55
Zack
In reply to this message
That is fine. You are welcome to ask these types of questions here.
OK
00:55
O K
Your english is great @kacperwikiel, my suggestions had nothing to do with you specifically
Z
00:55
Zack
In reply to this message
Maybe you can help to translate some stuff into other languages
00:57
Deleted Account
Sure, i can do polish translation also i can do nice latex whitepaper from existing one
Z
00:57
Zack
In reply to this message
Awesome :)
MH
01:10
Mandel Hoff
I've tested my pow miner changes. It's hard to get reliable numbers, but it looks like it's much faster h/s after the pow change. I'm seeing roughly 230mh/s -> 320mh/s on my 750Ti. That's doing "invalid work", so don't hold me to those speed boosts. The pow test net has such a low diff that I don't get good numbers due to solutions being found too fast. Just wanted to pre-warn for a very likely h/s boost when pow change happens.
Z
01:14
Zack
In reply to this message
Sounds good.
This is more evidence that an attacker could have been exploiting us and mining faster than Mandel s miner.
D
01:19
D
In reply to this message
How so? The new PoW hashes fewer bytes for the nonce (23 instead of 32) so a speedup of ~1/3 should be expected imo
01:20
Lots of room for improvement in the current implementation of the cuda kernel. I don't see any vectors being used and converting to int each time to check if it's below the target is very wasteful
Z
01:21
Zack
In reply to this message
That isn't how Sha256 works.
It hashes in blocks of about 55 bytes.
The old version of the code had 2 blocks, but you only had to do the first block every 64000 times or so. Then you could manipulate the nonce in the second block.
01:22
The new version only has 1 block. So it is simpler, but the ideal solution is the same amount of work.
MH
01:22
Mandel Hoff
My pool is still planning on stopping operation. I'm handing over the binaries for OK to operate, and I will help with launch and occasional maintenance. The miner tool Url argument can be used after transition happens. I think the Url argument works in 1.0.0.7 of the mining tool, but let me know if it doesn't. I did test it against a second pool I stood up with an IP address and no domain today.
I
01:23
Iridescence
Are you keeping the domain?
MH
01:24
Mandel Hoff
I can look in to transferring the domain, but I have not yet done so.
I
01:24
Iridescence
It's a nice domain =k
01:24
=)
MH
01:26
Mandel Hoff
If the tax code ever comes out with very clear direction on miner payouts not being contractors and subject to information reporting, then I'd certainly be happy to continue operating.
D
01:26
D
In reply to this message
I'm aware of how SHA256 works. Please note I referenced the size of the nonce. It's common knowledge that if you put the nonce at the end then anything before can be "pre-hashed". Mandel is likely initializing the context with the first 32 bytes and only hashing the rest of the 23 byte nonce
01:27
If you want to "fix" your PoW then put the nonce at the front so the miner has to perform a computation on the entire buffer each round
I
01:28
Iridescence
Oh god please don't change the pow again lol
D
01:29
D
If you do at least choose something more able to be mined by the common man >_<
I
01:29
Iridescence
How is one form of pow more mineable by the common man over another form of pow?
D
01:30
D
Memory hard algos offer a more level playing field
P|
01:30
Paul N. | BAND (🅑) - I'll never DM you
In reply to this message
not gonna find it here ;)
D
01:30
D
Monero is forking because an ASIC is rumored to have come out
I
01:31
Iridescence
Zack has explicitly stated that he is supports Asics for Amoveo
D
01:31
D
I understand
01:31
imo he will lose community support before anyone approaches that level of investment
MH
01:32
Mandel Hoff
In reply to this message
The new pow is exactly one transform buffer, so it doesn't matter if the nonce is at the front or back. The memcpy would be slightly different, but the transform cost is exactly the same - 1x64 byte transform.
01:37
All I see is the nonce size being reduced to 23 bytes
I
01:37
Iridescence
Also difficulty isn't included in the block anymore
D
01:37
D
Yes sorry that as well
Tv
01:39
Tarrence van As
What is the new PoW algo?
D
01:40
D
Old PoW: sha256(<block hash (32 byte)>+<difficulty (2 byte)>+<nonce (32 byte)>)
New PoW: sha256(<block hash (32 byte)>+<nonce (23 byte)>)
01:41
Same shared context trick applies
01:41
Deleted Account
BKyLqCvF5vt7KBiPLId7xlW64Bb/uV+icPnvn5mgEKncqlCKiaXUTkRcZv7iYK3GKxSspNh0leIhc4YA+DJAY/s=
01:41
who tf is that guy or lady lol
01:41
payouts out the ass
D
01:41
D
In reply to this message
It's amoveopool.com
01:41
Deleted Account
Ah so it uses that to distribute?
I
01:41
Iridescence
That's @Mandelhoff pool
D
01:42
D
If that pool goes down the whole amoveo network does
01:42
Deleted Account
Okay gotcha, I was like "what kind of monster rig must this person have?"
01:45
Deleted Account
just got a 569340992ERROR: GetWork: 500 Sleep and retry...
01:45
any idea why?
MH
01:50
Mandel Hoff
Maybe a fluke where the node didn't give the pool work for a brief moment. Pool seems up.
MF
02:00
Mr Flintstone
In reply to this message
where is diff stored?
Z
02:01
Zack
In reply to this message
In the block and in the header.
MF
02:01
Mr Flintstone
sorry, where is it stored after pow change
02:01
just header?
Z
02:02
Zack
In the block, and in the header.

We no longer store it in the digest of the header which we are doing pow on.
MF
02:21
Mr Flintstone
got it thanks
OK
03:33
O K
lol c ya
M
03:34
Mike
Lol
04:02
Deleted Account
we need asics here!
Z
04:22
Zack
should we make a bounty for a verilog or VHDL miner?
JM
04:39
J M
are there any live markets?
04:40
Deleted Account
May i miss something? i have read 3,5ghs per veo day. I´ve been 4 hours with 7x1060gtx = 2,5Ghs (not so much with Mandel´s miner but this is other issue) and only get 0.03veo (almost 20ghs per veo day) . is normal such a difference?
Z
04:42
Zack
In reply to this message
Maybe our math is still wrong, or maybe your miner isn't doing as much as it could.
04:42
20/7 is pretty close to 3.5
Maybe all 7 of your miners are doing the same work?
M
04:43
Mike
In reply to this message
You miner should be doing significantly better
MH
04:43
Mandel Hoff
What address? I can check. Confirmed block credit has a delay >1 hour, so you probably have a bunch of credit coming when blocks get confirmed.
OK
04:43
O K
.118 veo would be based on 2.5 GH/s for 4 hours I think
04:45
Deleted Account
currently getting 4,2 GH/s on a V100, is this expected hashrate?
OK
04:45
O K
I have some variance in the short term too, give it a day
04:45
With what miner Ivar?
04:46
Deleted Account
tumblecatweed's miner for linux
OK
04:46
O K
Sounds about right
04:46
Deleted Account
In reply to this message
BPFuy9IF7xKGyYwmedCNuY+epxKnzdK/UoUCmpasD366l3sttKyERyR2daHlaxc8IrzVEajSGwI3YL9VZrhbMsg=
MH
04:46
Mandel Hoff
Someone reported my miner on windows Tesla V100: 3900 Mh/s - Suggested BlockSize: 1024, Numblocks: 64, so that sounds about right for linux.
04:47
Deleted Account
currently using 128/2048, but not sure what is best configuration
MH
04:47
Mandel Hoff
In reply to this message
You have 0.26 credited already and more coming - not 0.03. Seems about right.
04:47
Deleted Account
In reply to this message
i had 0.227 from the first days.
04:48
In reply to this message
i had 0.227 from the first days.
04:49
i have the 7 gpu´s with the same address, is it matter?
MH
04:50
Mandel Hoff
If you aren't getting "Duplicate Nonce" errors, then it's fine. If you do get that - use different seed strings for each miner.
04:51
Deleted Account
somtimes log Info: "ERROR: Duplicate nonce submitted this work item" . and sometimes log "nfo: Not valid share. Difficulty too low." but mainly appears good logs
04:53
seed string is RandomSeed, right? i will try
MH
04:53
Mandel Hoff
You can fix the duplicate nonce errors with different seed arguments per miner you kick off. The diff too low is a result of the bhash changing
04:53
In reply to this message
Yes
04:53
Deleted Account
In reply to this message
Tnaks
MH
04:59
Mandel Hoff
The pool database shows you started 2 hours and 44 minutes ago too, which is different than 4 hours. About half your credit is waiting to be applied via the confirmed block delay too.
04:59
Deleted Account
In reply to this message
Itis working and also ghs increased till 3ghs in total.
MH
04:59
Mandel Hoff
👍
05:00
Deleted Account
In reply to this message
No, i have started at 17:20 (then i went to the swining pool with children ;))
M
05:01
Mike
Lol
MF
05:01
Mr Flintstone
are we still at 8632?
05:01
Or does someone else have a better block height?
05:01
8362*
05:01
Deleted Account
In reply to this message
It is also strange the ghs stimations in pool, it shows much more ghs than i have (25ghs...)
Z
05:02
Zack
looks like we are still at 8632.
MH
05:04
Mandel Hoff
In reply to this message
Yes, pool hash rates are not very accurate. No smoothing is done like a pplns system would do. It only knows the difficulty of the partial solutions you submit, and you get a proportional credit towards that partial solution.
05:05
Deleted Account
In reply to this message
ok, great to know. thanks. I will let all night and see how it is working. Thanks again.
05:21
Deleted Account
@Mandelhoff What about the total hashes, are they accurate?

I'm trying to figure get a better sense of the rates I'm getting on different machines / with different settings, but am seeing weird results such as: for 2 machines in the same period of time the amount of hashes they completed does not directly correlate with the amount of coins I was awarded.

Could this be a lag issue? Any other ideas?
05:24
Also the hash rate (either the displayed one or the one inferred from total hashes rate of change) doesn't seem to match the reported hash rate in terminal
MH
05:31
Mandel Hoff
The partial solutions you submit have a difficulty. The difficulty of that partial solution converts to X hashes to accomplish, just like a block difficulty. You are credited with X hashes. When the pool finds a block, you get the sum of all your credited hashes divided by the sum of all other miners hashes (since the last block found by the pool) and that's the fraction of the 1.0 veo you are credited in Unpaid Balance (after that block is confirmed).
05:33
Your credited veo balance definitely can vary across blocks based on luck of partial solutions submitted.
D
05:34
D
Zack the output from hash2int is non-uniform as the hash's *raw value gets lower. Is this expected? I have the function copied out on its own here with the results https://ideone.com/MEDelq
Z
05:35
Zack
In reply to this message
It sounds like you could be making a mistake.

Set a limit of work, and give credit to above that limit. Give them all the same amount of credit for each solution above the limit.

Do not scale the reward with the difficulty of that particular solution.
05:37
In reply to this message
It is impossible that the output of Sha256 would start with more than 247 bits of 0s.
05:37
So it doesn't matter how the function acts in that range
D
05:37
D
For any range it isn't uniform. That's a cop out
Z
05:38
Zack
What is the problem? I don't understand.
D
05:40
D
The difficulty output of hash2int does not follow in a linear fashion. I asked if that was expected behavior, because the name of the function implies that you are looking to convert and not mix around the value of a hash
Z
05:42
Zack
you are interpreting the hash value as an integer.
It is correct behaviour that the output is like a saw. There is nothing wrong here.
D
05:42
D
Is it expected behavior? The base 2 numbers make me think you were trying to do the equivalent of bit shifting
Z
05:44
Zack
The scientific notation is split into a mantisa and a value. The value is the byte starting with the first nonzero bit. The mantissa is the number of leading zeros.
S
05:45
Sy
In reply to this message
Why is a more advanced miner an attack? Irid is selling one, I would t consider him an attacker on the net...
Z
05:46
Zack
In reply to this message
He is not malicious. this is a type of failure mode that damages the network to iris benefit.
S
05:47
Sy
He will probably still hash faster than Mandel because he is most likely doing something fundamental different
Z
05:47
Zack
It is increasing the startup cost for mining. Which increases the profit margin.
05:48
In reply to this message
Perfection is impossible, that doesn't stop me from trying to be better than I am now.
S
05:49
Sy
I don't think it will take that long for a public miner to become as fast as closed ones, it isn't a that complicated algo after all
OK
05:49
O K
someday the amd gods will hear our prayers
05:50
🙌
S
05:50
Sy
Use open CL, problem solved...
OK
05:51
O K
someday the opencl gods will hear our prayers
05:51
🙌
S
05:51
Sy
😁
I
05:52
Iridescence
I don't see how having a closed source miner that performs better than open source ones damages the network
Z
05:53
Zack
In reply to this message
@Iridescence
I
05:54
Iridescence
I don't see how a profit margin damages the network
Z
05:54
Zack
Plus everyone using it has bigger profit margins
05:54
The profit margin is economically analogous to counterfeiting.
If you get free tokens, it makes our tokens worth less.
I
05:55
Iridescence
😂
Z
05:56
Zack
Healthy miners increase profit by increasing scale, not by improving software.

It isn't @Iridescence fault. We shouldn't blame him.
He is just a symptom of the underlying problem.
I
05:57
Iridescence
Somebody else is just as welcome to come up with an open source miner that is just as if not more performant
05:57
But as I said in the very beginning
05:57
What are the incentives
05:57
Think about it =)
Z
05:58
Zack
And the fact that they haven't is probably my fault.
I could have hired someone before launching.
I
05:58
Iridescence
Mandel is incentivized to improve his miner because it mines to his closed source and very awesome pool
Z
05:59
Zack
If someone owns a bunch of tokens, then it is in their interest to make the best miner available to everyone.
I
05:59
Iridescence
I can nearly guarantee you that if anybody in this chat developed tech that made them have majority hashrate of the network, they would not open source it.
05:59
At least not initially
05:59
Expecting pure altruism in crypto is folly
Z
06:00
Zack
Like I said, it isn't iris fault, and we shouldn't blame him. If it is anyone's fault that this is happening, it is mine.
M
06:01
Mike
In reply to this message
It is each our own faults for not having the drive to make it on our own
I
06:01
Iridescence
Right now Mandel's pool is mining s large majority of blocks. Is that an aggreable situation to you? (It is to me, but I wanted to get your thoughts)
Z
06:02
Zack
It isn't good because sometimes his node crashes and then the network is frozen. It would be better if there were at least 2 big mining pools.
I
06:03
Iridescence
Fair enough
06:03
When I have some time I'll see how easy it is to make a shared pool from your code
06:04
Hopefully you won't change the pow algo again XD
OK
06:05
O K
We are looking into splitting the mining pool also for some redundancy
Z
06:06
Zack
Oh yeah, if you just ran 2 of them, that would be a big improvement.
s
07:20
scott_l | DeFi Pulse
In reply to this message
ahhh yes. false equivalency season has begun
07:21
iri isnt destorying the block over a miner he is selling for $900. that is pedantic
JS
08:47
Jon Snow
Just did a deal of buying 34 VEO. Much appreciated Zack’s help on being a great escrow! Very responsible
08:47
Thanks Zack !
M
08:49
Marc
thanks for esrow service Zack
08:49
big up!
08:51
we must have been annoying, lol. thanks for your patience.
N
08:51
NM$L
In reply to this message
what price
Z
08:53
Zack
In reply to this message
I think you bought 37?
08:54
In reply to this message
0.00572972972 bitcoin per Veo.
About $46
JS
08:55
Jon Snow
Right, sorry 😬, 37, I remembered incorrectly
M
09:07
Marc
ja. 37.:D
BG5EHI joined group by link from Group
P
10:07
Pan
10:08
Zack,
found 2 abnormal:
1, alway show 'talk_helper fail_connect 1'
Z
10:10
Zack
In reply to this message
It is running normal.
P
10:11
Pan
2, when fail to connect to a well synced node after 'sync:start().', I have to do the second 'sync:start()'
Z
10:11
Zack
In reply to this message
Yes. That is normal. It is explained in the sync.md document.
P
10:13
Pan
I wish the code could automatic check and change to another peer to sync
Z
10:14
Zack
Yes. Me too.
I have tried several times. This code is tricky because it is easy to waste bandwidth if something goes wrong.
P
10:19
Pan
Ok, can be deal later, you are so busy!👍🍺
N
10:23
NM$L
btc crashing.
M
10:31
Minieep21
BTC bubble has popped.
Now is when the real innovation starts.
Amoveo will grow during the bear and explode during the bull
M
10:33
Mike
In reply to this message
Popped?
M
10:34
Minieep21
Yes. BTC going back to $3k over the next 6 months.
MF
10:39
Mr Flintstone
still 6x in a year lol
N
10:41
NM$L
In reply to this message
why do you think so
10:41
5k is the bottom
MF
10:41
Mr Flintstone
who knows what the bottom is
10:42
I would be very surprised if we saw <2k sustained
10:42
anything else is fair game imo
10:42
not gonna sell tho
N
10:46
NM$L
It's bloody
M
10:46
Minieep21
In reply to this message
People will keep making bottoms until everyone has their guts spilled from their longs.
10:46
I agree there will be many bounces.
10:46
But it's done, it's a classic euphoric bubble. Just like the past 3 times.
10:53
Learn to short and you'll make a killing. Bull climbs stairs, Bear jumps out of windows ;)
MF
11:31
Mr Flintstone
block time has been a healthy 7 minutes for a while now
12:10
Deleted Account
what's the price now？
JS
12:10
Jon Snow
You wanna sell or buy
13:39
Deleted Account
In reply to this message
Zack, why would this be a mistake? I believe most pools work this way. Is there something different about Amoveo we're not considering?
Deleted joined group by link from Group
Baudoin Collard invited gpo
Z
17:25
Zack
In reply to this message
It sounds like he implemented pow wrong.
You don't want to pay at every difficulty, otherwise people would submit more attempts and spam you with bandwidth.
17:25
Deleted Account
ah, well it is above a limit, which grows (or shrinks) based on your submission rate
17:26
but if you find a higher difficulty share, it is worth proportionally more of the reward - I thought that is what you suggested should be avoided?
Z
17:30
Zack
In reply to this message
Yes. I am saying to avoid this. Otherwise people will spam you wih too many solutions.
17:31
In reply to this message
This rewards sybil attacks. I would sign up my laptop as 500 different miners, that way my cpu would get many times more rewards.
17:32
Deleted Account
I thougth that would be the case if the difficulty were disregarded?
17:32
if you reward any difficulty (above some limit), then your attack scenario makes sense
17:32
if it is proportional to difficulty, then you are not penalised for sending less shares (and sending more lower difficulty ones is equivalent but possibly wasteful)
Z
17:33
Zack
In reply to this message
For example, it takes 2 hours to make a 2-hour proof. But in the process of making it, I will have probably also created 2 1-hour proofs and 4 half hour proofs and 8 15 minute proofs
17:34
16 7 minute proofs. 32 4 minute proofs
17:35
Deleted Account
hrm, interesting, that does make sense
Z
17:35
Zack
If I send in all these proofs, I would get paid 6x more than just sending the 1 high difficulty proof in.
17:35
Deleted Account
the pool will stop accepting lower difficulty shares if you submit too fast though, but as you point out that does not prevent a Sybil attack?
Z
17:35
Zack
But Mandel would have to deal with 65 messages instead of 1.
17:36
Deleted Account
maybe I misunderstand how it works
Z
17:36
Zack
In reply to this message
Correct. If we use a sybil attack then the server doesn't know it all came from one person.
17:36
Deleted Account
but I thought most pools worked this way
Z
17:36
Zack
I submit wah proof as a different person.
17:36
In reply to this message
No they don't.
Every proof above the limit needs to have the same value.
17:37
You can give each worker a different limit, but all the proofs above that limit must have the same value.
17:37
Deleted Account
it is commonly said that 10 low difficulty shares are equal to 1 high difficulty share, you're saying that must not be the case, or if it is, it's a flaw that could be exploited?
Z
17:38
Zack
In reply to this message
Probably you just don't understand what they mean.

It is safe to give different limits to each worker, and to pay more to workers with a higher limit.
17:40
The amount you pay needs to be based on the limit. Not by how far you exceed the limit.
17:40
Deleted Account
so the payment is proportional to the limit, not the share
17:53
right, I think this problem is avoided in stratum-based pools because of the job ID
Z
18:10
Zack
Rolling your own crypto is very risky.
It is a bad idea to try and invent a new POW for your mining pool.
18:14
Deleted Account
yes, thanks for explaining, that is a subtle one!
18:14
hopefully Mandel's pool doesn't have this problem
18:14
hard to know with closed source though!
20:23
Deleted Account
1050ti mining 4 days get 0.6 veo
20:23
too hard to mine
20:34
Deleted Account
10$ per day not bad?
21:05
Deleted Account
How many 1050ti's are you running?
23:32
Deleted Account
Added hardfork annoucement. 493 blocks to go...
https://github.com/tumblecatweed/AmoveoMinerGpuCudaLinux
23:33
announcement
16 March 2018
MF
01:14
Mr Flintstone
what would the confidence interval for hash rate be if you sample n blocks? apparently I need to brush up on stats
Z
01:33
Zack
In reply to this message
I think it is something like hashrate*(1+-1/sqrt(N*hashrate))
02:13
+-1.96sqrt(period/N) is easiest if period*N is large
MF
02:17
Mr Flintstone
thanks!!
S
03:37
Sy
whats the reason of api:height(). 0 although ive just traded peers with zacks node and the list is full
Z
03:38
Zack
In reply to this message
That means you don't have headers.
S
03:39
Sy
shouldnt i get them automaticly?
03:40
sync:start doesnt do anything without headers...
03:40
ah get headers...nvm
03:40
annoying tbh ^^ it should do that at startup vs my nodelist
Z
03:41
Zack
It tries to do this at startup, but sometimes your peer doesn't respond.
S
03:42
Sy
** exception error: undefined function pow:check_pow/3
03:42
in function headers:absorb/2 (/root/coins/amoveo/_build/prod/lib/amoveo_core/src/consensus/chain/headers.erl, line
J
03:42
Jim
Can it be set to retry after a sane time interval until it works? Would make managing a node less manual process
Z
03:43
Zack
In reply to this message
It is failing to sync headers? I haven't changed the POW.erl file today or yesterday.
03:45
In reply to this message
Sounds like a good upgrade.
S
03:45
Sy
I did a git pull before start
Z
03:48
Zack
In reply to this message
I am building fresh from github to try and recreate the bug
03:52
In reply to this message
it synced all the headers normally.
03:52
In reply to this message
now it is syncing blocks normally.
S
03:52
Sy
Well Mine broke in-between somewhere
Z
03:53
Zack
oh, I know what happened.
S
03:53
Sy
Good 😁
Z
03:53
Zack 
rm -rf _build/prod/lib/pink_hash
make prod-restart
03:54
the upgrade for the hard fork changed pow.erl
If you try running the new amoveo full node using old dependencies, then it fails.
03:54
dependencies don't automatically get reinstalled when they change.
03:54
git pull doesn't update them.
S
03:56
Sy
Ah k
S
04:39
Sy
===> Verifying dependencies...
===> Skipping pink_hash (
04:39
thats okay?
Z
04:40
Zack
pink_hash is a dependency for 2 dependencies of Amoveo. Rebar is smart enough to skip downloading it a second time.
S
04:41
Sy
error ir still the same tho, how do i tell rebar to update deps?
05:05
Deleted Account
what's current price of veo? might be willing to part with a few...
05:16
Deleted Account
I've heard roughly 65-70 per
05:16
possibly more