Author

Topic: [DVC]DevCoin - Official Thread - Moderated - page 188. (Read 1058949 times)

legendary
Activity: 2044
Merit: 1005
February 10, 2014, 06:22:48 PM
to help understand the code path, what is the setup of the pool that is testing it? i see it is p2pool but is it mining bitcoin as the primary with devcoin as a merge mine coin? or is this initial test mining with devcoin as the primary? if the latter, has getblocktemplate (which i assume p2pool is using) modified from the original bitcoin one at all? If the former, it will be using the getaux... paths.

The P2Pool is mining bitcoins, and merge mining namecoin, i0coin, ixcoin and devcoin. I don't think you can merge mine bitcoin yet, the others you can do because the alt coins have code that allows it, whereas the bitcoin just ignores what they add to the bitcoin blockchain (or something, I don't understand it too well yet).


There was a small change to template which matches bitcoin but wasnt in devcoin. When creatinh a block I set txout value to -fee in the template.. This could be the cause.

Why is it in bitcoin as -fee? I assumed it was something to do with the client checking to ensure sufficient balance before sending coins.

https://github.com/sidhujag/devcoin/search?q=-nFees&ref=cmdform

But this is in original commit and he said that one connects.. he should see rpccalls in the pool log to template code?

At the very least see the msg "running devcoin miner"?

The problem with the P2Pool is it doesn't create a log for successful attempts...it just says "Got new merged mining work!" but not from which daemon.

can I make that change locally on the master git and see if it still causes an error?

Hey I pushed a change, can you see if it still crashes?
legendary
Activity: 1988
Merit: 1007
February 10, 2014, 06:21:33 PM
When I made the 48 share bounty for the new client, I didn't know there were so many differences between the 0.3 codebase and the 0.8 codebase. Now I see that there's many things to do do, so I suggest boosting the bounty from 48 shares to 96 shares. The 48 shares have already been sent, so this would be a boost of 48 share in round 33. Any objections, or should something be changed?


48 shares isn't probably going to be as much next round as it would have been previously, in terms of the number of devcoins.  Maybe increase the number?

I'd say wait. People may drop off now that the rewards are lower (as we've seen in the past, as rewards drop, people stop. As rewards get bigger, people jump back on). Don't make any assumptions yet.
legendary
Activity: 2044
Merit: 1005
February 10, 2014, 06:20:42 PM
that simply means we start fresh from 0.8.6 and add in devcoin again and not worry about backwards compatibility. We only have 1 major pool to consider and this new one, so its just a matter of setting  a date/time in the future and saying we are moving to a new codebase...
there is more than 1 pool to consider. i'm assuming you mean ghash.io but there are a bunch of other large miners as well. discus fish for example. eligius mines but doesn't distribute to miners. there are also many smaller pools and solo people mining devcoin.

Quote
f we do follow this path, this would mean we follow bitcoins difficulty retargeting and all other specs, except the main devcoin differences like auxpow/reciever share system, to keep consistent and be able to reuse the tools of bitcoin, by adding in merged-mining capability similar to what I did with the android wallet... however the wallet took longer because of the difficulty algorithm which the code didn't have in mind to check previous set of blocks to average. Small things like this add up and will cause problems one day.
this would completely change the coin. it wouldn't be devcoin anymore. you'd be moving from a coin with no upper limit to a 21 million limit. from 50,000 reward to 25 reward that halves. so many changes it would be bad for those that have supported the coin believing such fundamentals wouldn't change.

No I meant keep everything else except core devcoin features, the reciever share system, block reward and merged-mining were the main differences.
legendary
Activity: 1008
Merit: 1005
February 10, 2014, 06:12:38 PM
When I made the 48 share bounty for the new client, I didn't know there were so many differences between the 0.3 codebase and the 0.8 codebase. Now I see that there's many things to do do, so I suggest boosting the bounty from 48 shares to 96 shares. The 48 shares have already been sent, so this would be a boost of 48 share in round 33. Any objections, or should something be changed?


48 shares isn't probably going to be as much next round as it would have been previously, in terms of the number of devcoins.  Maybe increase the number?
legendary
Activity: 1008
Merit: 1005
February 10, 2014, 06:10:01 PM
The Devcoin Bounty page has been updated with the cryptocurrency convention speaker bounty, the devcoin payment processor bounty, the first mined block with sidhujag's daemon bounty, and the Blisterpool testing bounty.  Please visit the site and verify that the aforementioned bounties are all accurate.
newbie
Activity: 51
Merit: 0
February 10, 2014, 05:44:01 PM
that simply means we start fresh from 0.8.6 and add in devcoin again and not worry about backwards compatibility. We only have 1 major pool to consider and this new one, so its just a matter of setting  a date/time in the future and saying we are moving to a new codebase...
there is more than 1 pool to consider. i'm assuming you mean ghash.io but there are a bunch of other large miners as well. discus fish for example. eligius mines but doesn't distribute to miners. there are also many smaller pools and solo people mining devcoin.

Quote
f we do follow this path, this would mean we follow bitcoins difficulty retargeting and all other specs, except the main devcoin differences like auxpow/reciever share system, to keep consistent and be able to reuse the tools of bitcoin, by adding in merged-mining capability similar to what I did with the android wallet... however the wallet took longer because of the difficulty algorithm which the code didn't have in mind to check previous set of blocks to average. Small things like this add up and will cause problems one day.
this would completely change the coin. it wouldn't be devcoin anymore. you'd be moving from a coin with no upper limit to a 21 million limit. from 50,000 reward to 25 reward that halves. so many changes it would be bad for those that have supported the coin believing such fundamentals wouldn't change.
newbie
Activity: 51
Merit: 0
February 10, 2014, 05:39:39 PM
Seems like it may be a problem with a txout entry in one of the blocks the
node has... Almost like its better to delete the data directory and let it sync again to ensure you have all valid blocks until you turn on the pool.. but you already did this you said.. so Ill try it on my end.
might be a good idea to add a bunch of printing debug there for now so it's easier to see what is going on in the logs.
legendary
Activity: 1420
Merit: 1010
February 10, 2014, 04:14:40 PM
I'm getting this error a lot right now. Unfortunately that means I have to hold off on signing up new writers. If you messaged me over the weekend, thanks in advance for your patience, and I'll give it a shot tomorrow.

Yep many error 508 here.
Same here, i'll sign u up once site is back up

Fuzzybear
legendary
Activity: 1806
Merit: 1029
February 10, 2014, 03:53:39 PM
I'm getting this error a lot right now. Unfortunately that means I have to hold off on signing up new writers. If you messaged me over the weekend, thanks in advance for your patience, and I'll give it a shot tomorrow.

Yep many error 508 here.
legendary
Activity: 2044
Merit: 1005
February 10, 2014, 03:33:12 PM
When I made the 48 share bounty for the new client, I didn't know there were so many differences between the 0.3 codebase and the 0.8 codebase. Now I see that there's many things to do do, so I suggest boosting the bounty from 48 shares to 96 shares. The 48 shares have already been sent, so this would be a boost of 48 share in round 33. Any objections, or should something be changed?

No objection but further reasoning is as follows:

I tried to keep it consistent but there were manual merges that were code reviewed and passed people's judgements but the code is complicated because its not "that" well designed so its hard to spot problems in the large scope of things by simply looking at the code and considering all options. I think I may have ideas why Hunterbunter is not able to connect and its simply a connection issue and doesn't make sense because its not even getting to the submit work state so its not even testing merged-mining other than setting it up so workers can start getting work... im not sure why its calling CreateNewBlock right on connect, whcih is called from rpcmining, but its something that I will figure out and know the answer soon (just that it takes longer than it should)

As far as the rpcmining, nothing was changed between 0.8.5 and devcoin, but there are probably many changes in the underlying structure because the base changed drastically... I don't see why it would be a problem with new client to new client interaction but to keep backwards compatibility with merged-mining in mind is a tougher task, and its hard to debug. If someone was familiar with the code they could help by getting the unit tests up and going, so we can actually run tests against the code without running it live. It would speed up development and I just didn't get around to it. Instead of doing that I think I'm good at finding issues in code by simple problem solving which is what I'll do in this case. It may be a small fix that will resolve the client issues and we will be fine after that, since syncing and sending/recieving doesn't seem to have any issues between client versions. Even the android client is happy now. But I do think if we want to roll forward it makes sense to keep up-to-date with bitcoin source and doing that means we wouldn't have to reimplement devcoin manually from a rebase everytime... that simply means we start fresh from 0.8.6 and add in devcoin again and not worry about backwards compatibility. We only have 1 major pool to consider and this new one, so its just a matter of setting  a date/time in the future and saying we are moving to a new codebase...

If bitcoin has found to have a major flaw, then so does ours.. .we should leverage the development effort of bitcoin and not try to redevelop the wheel so its a more efficient path to achieve the same thing.

If we do follow this path, this would mean we follow bitcoins difficulty retargeting and all other specs, except the main devcoin differences like auxpow/reciever share system, to keep consistent and be able to reuse the tools of bitcoin, by adding in merged-mining capability similar to what I did with the android wallet... however the wallet took longer because of the difficulty algorithm which the code didn't have in mind to check previous set of blocks to average. Small things like this add up and will cause problems one day.

BTW for piece of mind I did merge in the version 1 block stuff in AcceptBlock already:

Code:
// Devcoin currently doesn't enforce 2 blocks, since merged mining
// produces v1 blocks and normal mining should produce v2 blocks.
#if 0
        // Reject block.nVersion=1 blocks when 95% (75% on testnet) of the network has upgraded:
        if ((nVersion&0xff) < 2)
        {
            if ((!fTestNet && CBlockIndex::IsSuperMajority(2, pindexPrev, 950, 1000)) ||
                (fTestNet && CBlockIndex::IsSuperMajority(2, pindexPrev, 75, 100)))
            {
                return state.Invalid(error("AcceptBlock() : rejected nVersion=1 block"));
            }
        }
        // Enforce block.nVersion=2 rule that the coinbase starts with serialized block height
        if ((nVersion&0xff) >= 2)
        {
            // if 750 of the last 1,000 blocks are version 2 or greater (51/100 if testnet):
            if ((!fTestNet && CBlockIndex::IsSuperMajority(2, pindexPrev, 750, 1000)) ||
                (fTestNet && CBlockIndex::IsSuperMajority(2, pindexPrev, 51, 100)))
            {
                CScript expect = CScript() << nHeight;
                if (vtx[0].vin[0].scriptSig.size() < expect.size() ||
                    !std::equal(expect.begin(), expect.end(), vtx[0].vin[0].scriptSig.begin()))
                    return state.DoS(100, error("AcceptBlock() : block height mismatch in coinbase"));
            }
        }
#endif
hero member
Activity: 935
Merit: 1015
February 10, 2014, 03:01:38 PM
When I made the 48 share bounty for the new client, I didn't know there were so many differences between the 0.3 codebase and the 0.8 codebase. Now I see that there's many things to do do, so I suggest boosting the bounty from 48 shares to 96 shares. The 48 shares have already been sent, so this would be a boost of 48 share in round 33. Any objections, or should something be changed?
hero member
Activity: 935
Merit: 1015
February 10, 2014, 02:56:55 PM
has the new devcoin client added any of the hard or soft forking options that the new bitcoin activates? what has it done with regards to:

* version 2 blocks. i believe devcoin currently uses version 1 coinbases. when version 2 coinbases were first used in bitcoin the client had code that would reject version 1 blocks when a majority of version 2 blocks existed in the last X blocks. this was later removed once the supermajority was met and i believe the current bitcoin code now rejects version 1 blocks.

* is p2sh activated? p2sh transactions are unsafe if the majority of miners don't run p2sh enabled clients.

* there was a fork during the 0.8 version of the client due to the change to leveldb and differing locking behavior to bdb in the older client. this caused a chainfork when the old client rejected blocks that the new client created due to size constraints. a patch was created for old clients to enable them to have better locking constraints to accept the unusual blocks from the new client. this issue will exist in devcoin if miners run old and new clients. it will be possible for new client operators to force a chain split.

i'm sure there are other differences. were fee policies changed?

Devcoin has a thin block chain so there are no size constraints. There's no need to reject version 1 blocks in devcoin. If there is code in the devcoin 0.8.x codebase to reject version one blocks, that should be commented out.

We don't need p2sh transactions, so that could be commented out also. Maybe in the future we could enable multisig, but it's not necessary now.
sr. member
Activity: 368
Merit: 250
February 10, 2014, 02:19:33 PM
Yep many error 508 here.
hero member
Activity: 720
Merit: 500
February 10, 2014, 01:32:37 PM
So, I'm a little late to the party, I just heard about this 'coin', but I like the concept as a I do a lot of open source work.  I am currently working on an open source bitcoin blockchain parser and analysis tool.  https://code.google.com/p/blockchain/  What's the process I go through to become an officially recognized 'devcoin' open source project?
http://devtome.com/doku.php?id=earn_devcoins_by_developing
Looks like devtome's having issues atm so otherwise msg hunterbunter
legendary
Activity: 1008
Merit: 1005
February 10, 2014, 01:27:45 PM
edit 2: The massive buy walls just disappeared - not sure what's going on  Huh

The buy walls were all one person I think...

Also is cryptsy OK now? I stopped using them.
legendary
Activity: 2044
Merit: 1005
February 10, 2014, 01:25:29 PM
Hunterbunter,

Looking at the code the CreateNewBlock gets called but txout nvalue can be negative either through a rounding error or if say the reciever file is messed up. It goes through all reciever entries to calculate value for each address. Does the receiver value exist? The latest one in the data dir? if its synced up you should have last rounds receiver file.

Id like to know the latest reciever file you have and that if its valid. I can try to replicate the scenario on my end then by doing local mining through the merged mining proxy with the same reciever file to see if it throws the same error...

Seems like it may be a problem with a txout entry in one of the blocks the
node has... Almost like its better to delete the data directory and let it sync again to ensure you have all valid blocks until you turn on the pool.. but you already did this you said.. so Ill try it on my end.

Also is it possible to turn off the other coins amd just have devcoin node running? Maybe some other coins are usingthe same rpc port and its a case of mistaken identity.. check that all the conf files have different rpcports defined.
member
Activity: 82
Merit: 10
February 10, 2014, 12:41:20 PM
So, I'm a little late to the party, I just heard about this 'coin', but I like the concept as a I do a lot of open source work.  I am currently working on an open source bitcoin blockchain parser and analysis tool.  https://code.google.com/p/blockchain/  What's the process I go through to become an officially recognized 'devcoin' open source project?
legendary
Activity: 2044
Merit: 1005
February 10, 2014, 12:27:50 PM
Is anyone planning to speak at the crypto convention in NY in April on behalf of DVC?

I hope someone does and gives a glimpse of whats coming up!
hero member
Activity: 742
Merit: 500
Its as easy as 0, 1, 1, 2, 3
February 10, 2014, 11:30:28 AM
Is anyone planning to speak at the crypto convention in NY in April on behalf of DVC?
full member
Activity: 166
Merit: 100
February 10, 2014, 11:19:15 AM
Hey, since I wasn't included in payment round 32 for devtome, what happens to 24,000 words I already posted this round? My last article was posted about 1 day before the deadline. Do I have a fresh 50k limit for this round or will the 24k overflow in my word limit for round 33 and I only have 26k left?

Another question. Smeagol said that he included me in the list for a 6 share financial writing bounty. Does this stand or will it too be delayed until round 33?

Thanks.

Re the questions on my report, I will link a few examples here once I start it up. Taking a small break at the moment.
Jump to: