Okay, got a new version -
This should improve stabilty, but there will still be some glitches,
https://docs.google.com/file/d/0B-5Ax5kejTpMWGkxaFoxT19SWVk/edit?usp=sharing
It should also do improve multithreaded performance a good deal. The hash rate won't improve, but it cuts down on redundant hashing, so you should see a more blocks for multi-core mining.
Topic: [ANN] MemoryCoin - page 32. (Read 100338 times)
Today I have mined 4 blocks already (while in the past week I only mined 2)....could it be that I am on a forked chain? If so, what could I do to get on the right chain again and how would I save my mined MEG?
Thanks!!!
Thanks!!!
Dose any one have an idea how I could check if I am still on the right chain?
My miners are at block 1207 right now.
You can also use the blockchainexplorer posted a few post above this to see at what block your chain should be.
EDIT: Let's make it easy, here's the link: http://memorycoin.21stcenturymoneytalk.org/
Today I have mined 4 blocks already (while in the past week I only mined 2)....could it be that I am on a forked chain? If so, what could I do to get on the right chain again and how would I save my mined MEG?
Thanks!!!
Thanks!!!
Dose any one have an idea how I could check if I am still on the right chain?
Today I have mined 4 blocks already (while in the past week I only mined 2)....could it be that I am on a forked chain? If so, what could I do to get on the right chain again and how would I save my mined MEG?
Thanks!!!
Thanks!!!
Man this is buggy as shit. Crashes constantly and the blocks are slow as hell. On Linux and Windows.
If you have to hardfork you should probably include all the things that need can be done at a hard fork at the same time.
Agreed - I've set up a thread about this here -
http://21stcenturymoneytalk.org/index.php/topic,29.0.html
One problem that is very easy to reproduce on all machines I have, is deleting the blockchain and then starting the miner with the StartWithMining.bat command. On every machine it will crash when loading the next grant-block into the chain (40, 60, 80, etc...)
I don't know if this problem is related but it might be, so fixing this might also solve the other problem.
I don't know if this problem is related but it might be, so fixing this might also solve the other problem.
Thanks! It's very helpful to have a reproducible error. I'm almost sure now it is to do with multithreading. Hope to have a fix very soon.
looks to me like the dev is making some LARGE coin here !
or maybe i don't get what is going on ?
a response would have been nice guys (anyone)
i think i asked a fair and yet very important question.. and i figured it would slip by ignored..
Fair enough. For the first few award blocks there were no mature coins and no-one understood the voting so there were no votes or vote winners. I thought if fair for the code to award those coins to me as the first grants as the developer. The first voted-for grant was awarded in Block 140, and by block 460, there were 5 grant award address, and only voted-for grants were being awarded. The total amount awarded was 1560. To put that in perspective . . at current difficulty, mining cost is currently around 2 cents per coin on a Amazon server. There can't be any more awards to that address as there are more than five grant award addresses in the system now.
well that sucks having it say i have coins getting block after block only to find out later i get nothing for my efforts 
and what is up with this line i see in my log 500 times ?
looks to me like the dev is making some LARGE coin here !
or maybe i don't get what is going on ?
and what is up with this line i see in my log 500 times ?
Quote
No Candidates with support remaining. Grant awarded to developer MVTEoEoww5Zz1SJ5DLcbdrxGwSY8VgFZDk
looks to me like the dev is making some LARGE coin here !
or maybe i don't get what is going on ?
a response would have been nice guys (anyone)
i think i asked a fair and yet very important question.. and i figured it would slip by ignored..
concerns me this is a "Spots" style self modded topic too lol
don't get your hopes up. FreeTrade rarely responds to critical posts. And as far as the self moderated topic, i think its an insult to compare this guy to Spots. this guy here knows what he's doing i've seen the code.
well that sucks having it say i have coins getting block after block only to find out later i get nothing for my efforts
and what is up with this line i see in my log 500 times ?
looks to me like the dev is making some LARGE coin here !
or maybe i don't get what is going on ?
and what is up with this line i see in my log 500 times ?
Quote
No Candidates with support remaining. Grant awarded to developer MVTEoEoww5Zz1SJ5DLcbdrxGwSY8VgFZDk
looks to me like the dev is making some LARGE coin here !
or maybe i don't get what is going on ?
a response would have been nice guys (anyone)
i think i asked a fair and yet very important question.. and i figured it would slip by ignored..
concerns me this is a "Spots" style self modded topic too lol
If you have to hardfork you should probably include all the things that need can be done at a hard fork at the same time.
This is a bit embarrassing. I left in a line of code that causes subsidy reduction in 1 month rather than 1 week.
I've uploaded a new client that fixes the issue, but it's a hard fork and it happens at Block 1680. At this time the network will split into anyone running the old code and anyone running the new.
So you need to download the new client as soon as possible and stop running the old one. Here it is -
https://docs.google.com/file/d/0B-5Ax5kejTpMOTNBYU8tckthMEU/edit?usp=sharing
If you're compiling from source - the offending line is 116 in chainparams.h. There's a rather ironic line nearby.
remove -
nSubsidyHalvingInterval = 7305;
The correct value is already set above at
nSubsidyHalvingInterval = 1680;
It might occur to you to stick with the old client and 1 month subsidy reduction - but rapid minting is an important part of this coin, and without it, it becomes just another shitcoin, so I wouldn't want any part of that.
I've uploaded a new client that fixes the issue, but it's a hard fork and it happens at Block 1680. At this time the network will split into anyone running the old code and anyone running the new.
So you need to download the new client as soon as possible and stop running the old one. Here it is -
https://docs.google.com/file/d/0B-5Ax5kejTpMOTNBYU8tckthMEU/edit?usp=sharing
If you're compiling from source - the offending line is 116 in chainparams.h. There's a rather ironic line nearby.
remove -
nSubsidyHalvingInterval = 7305;
The correct value is already set above at
nSubsidyHalvingInterval = 1680;
It might occur to you to stick with the old client and 1 month subsidy reduction - but rapid minting is an important part of this coin, and without it, it becomes just another shitcoin, so I wouldn't want any part of that.
love this coin, will love it even more if it takes off, ive managed to mine 10% of all the blocks anyone got any ideas what these puppies will launch at ? or grow to? looks like they will be hard to mine in a few months.
anyone got any ideas what these puppies will launch at ? or grow to? looks like they will be hard to mine in a few months. OK, good to hear you have identified the problem and are solving it.
Looking forward to the fix.
Looking forward to the fix.
Having some difficultly finding this.
Can you try deleting hashlookup.txt and the blocks and chainstate folders and restarting. It'll need to re-download and index the blockchain again, but might resolve the problem for you.
I think the issue is related to the multithreading as it happens more often on my 6-core CPU than it does on my 4-core.
Yesterday I put the miner on a simple 2-core machine as well and it hasn't crashed on that machine yet.
Actually the miner doesn't crash (on the 6 and 4 core) but gets stuck on a block so it keeps mining and wasting energy but not to find the correct block.
Getting stuck on a block happens more often when a grant-block is coming, but it can happen on other blocks as well.
One problem that is very easy to reproduce on all machines I have, is deleting the blockchain and then starting the miner with the StartWithMining.bat command. On every machine it will crash when loading the next grant-block into the chain (40, 60, 80, etc...)
I don't know if this problem is related but it might be, so fixing this might also solve the other problem.
21st Century Money just announced a Block Viewer - more details here,
http://21stcenturymoneytalk.org/index.php/topic,28.0.html
Remember to vote for 2CMT to support continued efforts - voting address is
MVTEoEomFfe7WuSEepMLsorgYzCQGBVw1z
http://21stcenturymoneytalk.org/index.php/topic,28.0.html
Remember to vote for 2CMT to support continued efforts - voting address is
MVTEoEomFfe7WuSEepMLsorgYzCQGBVw1z
Thanks - I added code to fix this, but I've just reviewed it and looks like I messed it up.
Problem is on line 1385
https://github.com/memorycoin/memorycoin/blob/fasterstarup/src/main.cpp
What's the ramifications of fixing this? Hard fork required?
Problem is on line 1385
https://github.com/memorycoin/memorycoin/blob/fasterstarup/src/main.cpp
What's the ramifications of fixing this? Hard fork required?
Ah, yeah.
Yeah, a hard fork to now fix it.
What's the ramifications of fixing this? Hard fork required?
I think the network just needs more miners.
Still reading through some of the old posts about this.
Some interesting finds:
https://bitcointalksearch.org/topic/m.554877
https://bitcointalksearch.org/topic/m.517370
(And then read the previous 20 or so posts above this one, for the backstory).
Not sure if a hard fork will be necessary.
That said, this reminds me of another potential issue. Bitcoin itself never fixed the time warp bug, so since this was forked from bitcoin, the time warp patch may have been overlooked to add in. Code to exploit it already is in the hands of those who like to do such things, so yeah, they would use it on a new coin, and have in the past.
So if/when a hard fork is ever done the fix for that should also go in ( https://github.com/twobits/MultiCoin-2bit/commit/98deadb47968229d71d537324a139f15edb0c3f9 ).
Thanks - I added code to fix this, but I've just reviewed it and looks like I messed it up.
Problem is on line 1385
https://github.com/memorycoin/memorycoin/blob/fasterstarup/src/main.cpp
What's the ramifications of fixing this? Hard fork required?
I'll add that after spending a few hours looking into it, I don't think it's caused by your code.
IMO, it's possible it could happen by chance, due to a smaller network and peers that are not running NTP.
IMO, it's possible it could happen by chance, due to a smaller network and peers that are not running NTP.
Hmm, thanks - I'll add that I'm not seeing the issue on my local machine here, running multicore at full throttle. Maybe because it is well connected?
That sounds pretty sophisticated for such a new coin - could someone really be taking all that effort before it is even listed on an exchange?
I'll add that after spending a few hours looking into it, I don't think it's caused by your code.
IMO, it's possible it could happen by chance, due to a smaller network and peers that are not running NTP.
Most likely due to the smaller network with limited peers accepting incoming connections. The two builds we uploaded yesterday ( http://21stcenturymoneytalk.org/index.php/topic,23.0.html ) both have UPnP enable which should help some, as well as the node running on the block viewer ( http://memorycoin.21stcenturymoneytalk.org/ ) being a seed node that can accept incoming connection and be long running will help
That said, this reminds me of another potential issue. Bitcoin itself never fixed the time warp bug, so since this was forked from bitcoin, the time warp patch may have been overlooked to add in. Code to exploit it already is in the hands of those who like to do such things, so yeah, they would use it on a new coin, and have in the past.
So if/when a hard fork is ever done the fix for that should also go in ( https://github.com/twobits/MultiCoin-2bit/commit/98deadb47968229d71d537324a139f15edb0c3f9 ).
Jump to: