wait... does the pool miner not print out it's hashrate?
ok, there it is
ok, going to take a break for now
if anyone wants to try the hybrid miner it's in tster/hybrid-miner for now
INFO: Mining rate: 130666.289/sec that's with my 64gb memory desktop
memfield_precache_gb=52
From the pool side the hash rate is a wild estimate since it can only infer from shares
And it is 20 second rate period
So really nonsense
@Fireduck Pool diff increases if there are more than 3 shares per minute?
Holy shit someone looked at my code
Yeah
@Protovist are you the silent pool operator?
Yes, I'm working on adding some UI.
cool. what UI are you looking to do?
Cool
I was thinking of adding a get status Rpc to MrPlow so explorer can get info and show it
Just the basic miner/pool stats to start. Seeing if I can get rocksdb-node to read from plowdb.
You are better off making a wrapping class that starts a node and MrPlow in the same process.
Otherwise you'll have strange rocks read issues if something else it writting to it
Hmm, OK, I'll keep that in mind.
If you want a review if whatever you are planning I'm happy to help but of course you can do whatever
Thanks. I'd like to just get something simple working ASAP, but we'll see how it goes...
How does the pool determine which address is used for the reward in the coinbase tx?
It doesn't seem to be the configured pool_address.
Heh, that is the entirety of the magic
It passes a weighted map of addresses and share counts to the node to build the coinbase
Which directly pays the miners
So easy, no pool wallet and payment processing
The shares are managed by ShareManager
So is the address shown in the block explorer the highest weight?
I think the explorer just shows the first one
That isn't fancy
Having the payments in the coinbase directly is very slick..
The snowday pool always seems to have the same address shown.
Probably same first address each time
yeah, it is a treemap inside somewhere so ends up being alphabetic
so it is that miner with the low address
[snowblossomcoin/snowblossom] Issue closed by fireduck64
So...we have an issue
I meant to make block reward half every 4 years. I apparently wrote 2 years in comments but actually implemented 1 year.
https://github.com/snowblossomcoin/snowblossom/blob/master/lib/src/PowUtil.java#L178 ``` public static long getBlockReward(NetworkParams params, int block_height) ```
So we could roll with it, or suggest a protocol change for sometime in the next year
If we make a protocol change, I highly suggest we do it sooner than later
If you're going to change it, a continuous reduction rather than a cliff would be nice.
Easier and more painless to do it now while everyone is more more capable of upgrading their systems and fewer people affected.
@Protovist that is an interesting idea
@Protovist What doy ou mean cliff exactly?
there will be no cliff if we make the change quickly enough.
I think he means rather than a big jump from 50 to 25 we do a slow scaling
like one block is 49.5, the next is 49.48
Yes, precisely.
AH, yeah, I like that idea.
I prefer a formula.
Formula( Height ) => Current in Existence.
if you public docs promise X and your code has a bug and delivers Y, I'm not sure I consider a vote appropriate for change. I would say it is proper to fix the code to match the original promise. Any nodes which haven't updated by the halving block will then end up forking. Likely no one currently in the community would still be running it, but if they are oh well, then there are 2 snowblossoms.
I agree that makes sense
but at this point, the safe and respectful thing to do is to make it a vote
that's not hittable from here
ipv6 only of course
Transaction: afc78aa2b9feeb5c38d0b9107a5c61ae662749079e41249fa8f462c88532c163 size: 123 Coinbase - height:3617 remark:pool test 4 Motions Approved: 1 3 Motions Rejected: 2 4 Output: snowtest:emcpfwr7m5u6zmg3n4wmnz3lf4lz9h5ecarth54a 49.500000 Output: snowtest:yt60p2097jyf9prp2qd5v3p8cxl7673u2n7pwrfh 0.500000 Fee: 0.000000
Why are you only allowing ipv6
just my internal network stuff on testnet
I only have one ipv4 address, so most of my test and dev systems are ipv6 only from the outside
Is that just a obscurity measure to limit access?
Just curious because I haven't seen that before
not at all
on my home network, every VM I spin up has an ipv6 address that is reachable from the outside automatically
so mostly I just use those
so if I am testing something, it is likely to be ipv6 only
and I tease anyone who can't access it
Lol I understand
is there a pool operator quickstart somewhere or up to me to figure it out?
@Tyler Boone when you do memory mapping of snow fields for the miner, if you can have the maximum per thread chunk match the maximum NUMA block on a symmetric multiprocessing system that'd help as well
@Fireduck FYI shackleton still does not build for me on the 1.0.6 tag - seems like some build config issue as it's trying to write to a temp file which already exists
``` ERROR: /Users/rotonen/koodaus/git/snowblossom/BUILD:38:1: Creating runfiles tree bazel-out/darwin-fastbuild/bin/Shackleton.runfiles failed (Exit 1) _bin/build-runfiles (args bazel-out/darwin-fastbuild/bin/Shackleton.runfiles_manifest bazel-out/darwin-fastbuild/bin/Shackleton.runfiles): mkdir '__main__/shackleton': File exists [17] ```
for heterogenous systems it's a touch more complicated to figure out, but there's ample literature out there, snowblossom is sorta the same problem as any modern physics experiment is dealing with
and figured this out from the source code as well
the Pool
about, but i'm still not sure as to what the double for the pool fee means
`1.0 == 100%` or `1.0 == 1%`
pff reading comprehension, never mind
@Rotonen Reward * 0.01 => 1%
1.0 is 100%
i suppose yours is the only currently public pool as it's in the docs and in the pool list
namely now i also have one at `http://snowplough.kekku.li:23380` and unsure on what's appropriate for advertising
@Rotonen Feel free to mention it in general. I can put it : https://github.com/snowblossomcoin/snowblossom/wiki/9.-Mining-Pools if you want.
your HTTPS seems broken
it does belong in the wiki, but yours will get most of the users as it's in the default config snippet section
so rather have a pointer in that section to a pool list?
ultimately there'd be a dynamic one with health statuses etc.
@Rotonen I believe in being fair and those are valid points. I will add health statuses and remove the default config
rather than health, continent or so would probably be useful
@Rotonen continent? ...
man, I barely recognize the word.
agreed
miners usually like to use a pool near them, for various reasons, which is not very important for snowblossom for now, but that'll cater to their expectations
@Rotonen How would you feel about me creating another url above yours for eastern hemisphere?
legitimately asking. I'm considering how I should enact policy on it for people putting pools up and wanting to group them together.
indifferent, people who actually care just care of shortest-path ballpark
so just group by country?
That works.
I would've figured they'd group by some other affiliation
boils down to speed of light + crap at national borders
@Rotonen What are your pool fees? (including potential donation to fireduck)
0.5%, 0 to the duck
playing dirty I see
@Rotonen Do you want any comment/remark on your info? ``` #SNOWPLOUGH #fee 0.5% pool_host=http://snowplough.kekku.li:23380 ```
none, and dunno if the port needs to be there if it's the default one
that's default, I'll make it pretty
I avoided putting any pool in the windows release config
Yeah, I only had it there because it was the only one in existence and a bunch of people wanted to jump in.
@Rotonen Sweet. https://github.com/snowblossomcoin/snowblossom/wiki/9.-Mining-Pools and I fixed it: https://github.com/snowblossomcoin/snowblossom/wiki/7.-Configs#poolminer Let me know if you have any other ideas/thoughts on fairness, even small ones, be petty.
That's not sarcasm, be petty. I prefer to identify things now rather than have people feel slighted later.
[snowblossomcoin/snowblossom] Issue opened by fireduck64
@Rotonen OH, better yet, I can set that entire codeblock to whatever you want in those 3 lines. :) Do you want to be named as pool operator?
no, simple minimalism is my thing
secondly
As a matter of policy, should we require pool operators to be known/public? that's the sort of thing I prefer not to decide for everyone.
Up to them I guess
Well, my concern is fraud.
If you're going to represent people, be known to them, sort of thing.
I think people should be suspicious of unknown pool operators but I don't think we should have any policy about it
sounds fair.
agreed
all my data is public via RIPE, for one
@Fireduck I like your comment on Pool failover. I was thinking about suggesting that myself given earlier discussion.
if anyone actually wants to argue for non-anonymity as an inclusion criterion for pool operators, i'd like to see all listed pool operators chip in for a neutral 3rd party law firm arrangement to actually ID individuals across nationstates, but somehow that seems a bit heavyhanded
I was just slightly concerned about people complaining that we advertised someone that might scam them
(not you, but future postings)
I guess if someone well know claimed to run a pool we should verify that it is them
Known account on Reddit it Twitter or gpg
But other than that..meh
@Rotonen Btw, my standard is, you can post it in the main channels on occasion, just don't spam. Discord, slack, etc. Maybe we should make an announcement channel.
@Clueless I think fraud is less of a concern with all the funds being paid out in the coinbase transaction.
I don't want to gatekeeper the community. just compartmentalizing my roles.
A dishonest pool operator could still alter the transactions, I suppose.
I mean down the line when snow is worth $1000 apiece.
`INFO: Error in send work: io.grpc.StatusRuntimeException: CANCELLED: call already cancelled`
pool thing, seems ominous
Just a broken connection
Not a problem
specifically, there's hundreds of entries/constant
just flooding logs, looks concerning
Hum. Might need exception isolation for the notify loop
Might be running out of connections
But I am pretty busy today
Any error from your miners?
I'm no longer mining personally, until I get more ram
hashrate seems lower than I'd expect.
but it's been fluctuating enough it's hard to tell
72,872,405 h/s is the average of the last 3 values in the last 5 minutes.
Ok. Shouldn't get those errors more than occasionally
I'll try to take a look
You take all the time you need
someone else really needs to set up a pool
I guess I can do it
We have two
oh sweet
There's 2 public, and I think 1 private
nice, no fireduck donation on it
pending 1-2 more private
screw that guy
lol, right?
I like that there are secret private pools
there's a bit of concern about people objecting to raising the supply
I recommend early as possible to avoid polarization and potential splits
you are asking early adopters to voluntarily dilute their blossoms by 4x
of course, I think without doing it snowblossom is destined to fail
Exactly
I think the early adopters are smart enough to understand that.
also, would voting work based on current miners, or amounts of snow?
Miners
hm.
Well miners may be inclined to allow the expansion so they can mine more
We need someone to write up the proposal
Give it a number
Write up proposal on wiki?
Can you also link me the line number for reference?
Not at the moment, PowUtil getBlockReward or something
thanks will lookup
another perspective is that we could just add it in. As that's what's originally advertised. That's honestly a fix.
Yeah, true
But open is good, I think
how long does the vote take?
depending on the length, someone could spend a couple thousand to rent enough hashing power to win the vote
easily, that's my concern.
I think it's fair to stick to the intended design. Someone trying to exploit it should probably be ignored.
doesn't sound like fireduck is as concerned about this problem as I am, which is fine
it's not like they were mislead
he's busy
no, he doesn't consider the reduced halving time as big a deal as I do
different priorities. it's fine
@Tyler Boone I'm planning to make my pool public as soon as I get some stats working.
@Fireduck The time it takes for difficulty to adjust is somewhat concerning. There's a very real potential for large mining pools to push up the difficulty and then abandon it once it's no longer profitable. This happened recently with VEO.
It's less of a concern since this isn't a GPU PoW, but still possible...
It adjusts pretty fast. Hash power dropped hard yesterday and it is already steady
Oh. I guess I hadn't looked at the logs to check the rate.
Anyone here familiar with Angular?
What for? (but no)
Was trying to use it for the UI. :slightly_smiling_face:
Perhaps a configuration option to dictate the amount of recent history to cache.
@FxxNIHAO has joined the channel
Jello
hi
how are you doing?
Alright
@Clueless I think those messages you were seeing about miners breaking connections is due to a bunch of miners connected by without the snowfield
so they keep thinking they are stalled and restarting the connection
This is definitely a lot faster adjusting than veo. Is every 2000 blocks isn't it? It would be best to be instantly variable. Similar to what bismuth switched to.
It actually shifts between blocks as well
Diff drops slightly each new get block template