BIP34 and BIP66 activated

[domob]

I just noticed that the 75% threshold for soft-fork activation of both BIP34 and BIP66 has been surpassed recently. This is no big deal, but I want to point out the consequences nevertheless:

• Miners can continue (for now) to use the old client.
• However, any user of the old client may be tricked into accepting a specifically-crafted block that Namecoin Core (and with it 75% of the hashrate) considers as invalid. This block will be orphaned soon after, but if you use the old client, you should be aware and potentially wait for a higher number of confirmations for high-value transfers.
• If the hash rate using Namecoin Core drops below 75%, this activation will be reversed temporarily. It is not yet a "point of no return".
• Once we reach 95% of the hash rate (which could happen quickly if a major pool upgrades, for instance), miners with the old client won't be able to produce valid blocks any more. Users of the old client can continue to use it, but are degraded to "SPV security".

EDIT by Jeremy: For anyone stumbling across this thread, we have passed 95%, meaning miners cannot use the old client anymore. End users can still use the old client, but are on semi-SPV security.

[biolizard89]

I just noticed that the 75% threshold for soft-fork activation of both BIP34 and BIP66 has been surpassed recently. This is no big deal, but I want to point out the consequences nevertheless:

• Miners can continue (for now) to use the old client.
• However, any user of the old client may be tricked into accepting a specifically-crafted block that Namecoin Core (and with it 75% of the hashrate) considers as invalid. This block will be orphaned soon after, but if you use the old client, you should be aware and potentially wait for a higher number of confirmations for high-value transfers.
• If the hash rate using Namecoin Core drops below 75%, this activation will be reversed temporarily. It is not yet a "point of no return".
• Once we reach 95% of the hash rate (which could happen quickly if a major pool upgrades, for instance), miners with the old client won't be able to produce valid blocks any more. Users of the old client can continue to use it, but are degraded to "SPV security".

Thanks for the update Daniel. Hopefully the remaining pools upgrade soon.

[phelix]

IMHO we should go forward and completely move to namecore soon.

[flound1129]

Hi,

The messaging here is confusing at best and I don't see much if any information about this anywhere else. What is namecore? Is it the software available at https://github.com/namecoin/namecore, that has code changes as recent as a few days ago and no releases?

At Multipool.US our last 3 namecoin blocks have been orphaned. I am running version 38000 (v0.3.76rc1, compiled from https://github.com/namecoin/namecoin, namecoinq-release branch) . Should I upgrade to namecore? If so, what source should I use? If I should not upgrade, is there a plan to release an updated version of namecoin that conforms to BIP34 and BIP66? I understand that these are causing the issues with orphans as per https://www.reddit.com/r/bitminter/comm ... 66_issues/ .

Why does namecore change the format of getauxblock (specifically, adding a "bits" field, and changing the "target" field to "_target"?) Can I change "_target" back to "target" in the source safely without making changes to my stratum code? Is the "bits" field used for anything on the stratum side?

Are there changes to the RPC call for block submits, or is the call the same as with the old version?

Any insight anyone can provide here would be appreciated.

Thanks,

--Adam

[phelix]

Hi,

The messaging here is confusing at best and I don't see much if any information about this anywhere else. What is namecore? Is it the software available at https://github.com/namecoin/namecore, that has code changes as recent as a few days ago and no releases?

At Multipool.US our last 3 namecoin blocks have been orphaned. I am running version 38000 (v0.3.76rc1, compiled from https://github.com/namecoin/namecoin, namecoinq-release branch) . Should I upgrade to namecore? If so, what source should I use? If I should not upgrade, is there a plan to release an updated version of namecoin that conforms to BIP34 and BIP66? I understand that these are causing the issues with orphans as per https://www.reddit.com/r/bitminter/comm ... 66_issues/ .

Why does namecore change the format of getauxblock (specifically, adding a "bits" field, and changing the "target" field to "_target"?) Can I change "_target" back to "target" in the source safely without making changes to my stratum code? Is the "bits" field used for anything on the stratum side?

Are there changes to the RPC call for block submits, or is the call the same as with the old version?

Any insight anyone can provide here would be appreciated.

Thanks,

--Adam

Please take a look here: https://forum.namecoin.info/viewtopic.php?f=2&t=2354

[cassini]

Please take a look here: https://forum.namecoin.info/viewtopic.php?f=2&t=2354

That OpenSSL consensus thread doesn't answer flound1129's most important questions:

Why does namecore change the format of getauxblock (specifically, adding a "bits" field, and changing the "target" field to "_target"?) Can I change "_target" back to "target" in the source safely without making changes to my stratum code? Is the "bits" field used for anything on the stratum side?

Are there changes to the RPC call for block submits, or is the call the same as with the old version?

I suspect this getauxblock issue is the reason why jedimstr, too, has problems switching from Namecoin-Qt to namecore in his P2Pool mm setup. What is the reason for the changes to getauxblock?

[biolizard89]

Please take a look here: https://forum.namecoin.info/viewtopic.php?f=2&t=2354

That OpenSSL consensus thread doesn't answer flound1129's most important questions:

Why does namecore change the format of getauxblock (specifically, adding a "bits" field, and changing the "target" field to "_target"?) Can I change "_target" back to "target" in the source safely without making changes to my stratum code? Is the "bits" field used for anything on the stratum side?

Are there changes to the RPC call for block submits, or is the call the same as with the old version?

I suspect this getauxblock issue is the reason why jedimstr, too, has problems switching from Namecoin-Qt to namecore in his P2Pool mm setup. What is the reason for the changes to getauxblock?

Daniel would be the one to answer this for sure, but I'm pretty sure Eligius was able to drop in Namecoin Core as a replacement for the old 0.3.80, with the only change needed being changing the merged mining proxy to use "_target" instead of "target". I think P2Pool already merged a change that makes it use _target, so I think the P2Pool problems are from some other issue.

@flound1129 Apologies for the incomplete messaging. We had to deploy Namecoin Core a bit before we intended due to the OpenSSL consensus bug that was disclosed by Bitcoin devs. I'll let Daniel reply to your specific queries, since he's the expert there.

[flound1129]

Daniel would be the one to answer this for sure, but I'm pretty sure Eligius was able to drop in Namecoin Core as a replacement for the old 0.3.80, with the only change needed being changing the merged mining proxy to use "_target" instead of "target". I think P2Pool already merged a change that makes it use _target, so I think the P2Pool problems are from some other issue.

@flound1129 Apologies for the incomplete messaging. We had to deploy Namecoin Core a bit before we intended due to the OpenSSL consensus bug that was disclosed by Bitcoin devs. I'll let Daniel reply to your specific queries, since he's the expert there.

Eligius can probably do that because they're only merge mining Namecoin. Other pools (like Multipool) might mine several merged coins and Namecoin will now need a one-off change in the stratum code. I'd much rather maintain my own namecore fork with "_target" changed to "target" than do that. Is it possible?

This will also break out of the box compatibility with most stratum implementations out there.

[flound1129]

Please take a look here: https://forum.namecoin.info/viewtopic.php?f=2&t=2354

Your link says: "This means that all mining pools MUST upgrade to Namecoin Core or will have all blocks orphaned."

This thread says "Miners can continue (for now) to use the old client.".

If the former is the correct statement, can this thread please be updated to indicated that?

Thanks.

[flound1129]

You may also want to set the title for the Namecore repository to "Namecoin Core", and/or move it into the Namecoin repository, so that it's the first result when searching "Namecoin core".

Currently, the first result for "Namecoin core" is the old repository and the second result (domob1812/namecore) which is the real repository does not give any indication that it is official.