Title: Testnet will likely fork to BIP101 Post by: worhiper_-_ on November 09, 2015, 05:03:42 AM It seems like there's an active effort to make bitcoin's testnet fork to BIP101. As seen on reddit (https://www.reddit.com/r/Bitcoin/comments/3s0uii/bip_101_has_activated_on_the_testnet_need_your/):
Quote We have activated BIP 101 on the testnet. The hashpower there after activation is now less than the non-BIP 101 hashpower. This makes for an interesting test. If you would like to help gather data, please run the BIP 101 client of your choice with -testnet and keep the debug.log file to share for data analysis later. The more nodes the better! Transaction spam is welcome and encouraged :) What's interesting is that the hashrate that's causing seems to be coming from a single participant. https://www.reddit.com/r/Bitcoin/comments/3s0uii/bip_101_has_activated_on_the_testnet_need_your/cwtkyka What are your thoughts? Title: Re: Testnet will likely fork to BIP101 Post by: notbatman on November 09, 2015, 05:21:51 AM Really, how likely is it that they'll be able to achieve 75% of the total hashing power on the testnet?
Title: Re: Testnet will likely fork to BIP101 Post by: mezzomix on November 09, 2015, 07:28:21 AM It's not a problem. When the first large block is created all other participants will reject this block.
Title: Re: Testnet will likely fork to BIP101 Post by: forevernoob on November 09, 2015, 07:43:27 AM How does this affect Bitcoin?
Is it possible to create a new testnet if they fork it? Title: Re: Testnet will likely fork to BIP101 Post by: shorena on November 09, 2015, 07:45:05 AM It seems like there's an active effort to make bitcoin's testnet fork to BIP101. As seen on reddit (https://www.reddit.com/r/Bitcoin/comments/3s0uii/bip_101_has_activated_on_the_testnet_need_your/): Quote We have activated BIP 101 on the testnet. The hashpower there after activation is now less than the non-BIP 101 hashpower. This makes for an interesting test. If you would like to help gather data, please run the BIP 101 client of your choice with -testnet and keep the debug.log file to share for data analysis later. The more nodes the better! Transaction spam is welcome and encouraged :) What's interesting is that the hashrate that's causing seems to be coming from a single participant. https://www.reddit.com/r/Bitcoin/comments/3s0uii/bip_101_has_activated_on_the_testnet_need_your/cwtkyka What are your thoughts? Great, thats what testnet is for. Title: Re: Testnet will likely fork to BIP101 Post by: coinpr0n on November 09, 2015, 07:53:25 AM It seems like there's an active effort to make bitcoin's testnet fork to BIP101. As seen on reddit (https://www.reddit.com/r/Bitcoin/comments/3s0uii/bip_101_has_activated_on_the_testnet_need_your/): Quote We have activated BIP 101 on the testnet. The hashpower there after activation is now less than the non-BIP 101 hashpower. This makes for an interesting test. If you would like to help gather data, please run the BIP 101 client of your choice with -testnet and keep the debug.log file to share for data analysis later. The more nodes the better! Transaction spam is welcome and encouraged :) What's interesting is that the hashrate that's causing seems to be coming from a single participant. https://www.reddit.com/r/Bitcoin/comments/3s0uii/bip_101_has_activated_on_the_testnet_need_your/cwtkyka What are your thoughts? Great, thats what testnet is for. Exactly. But this should have done a long time ago. Testing should be paramount for such a change to the protocol. Title: Re: Testnet will likely fork to BIP101 Post by: worhiper_-_ on November 09, 2015, 12:44:38 PM We already see the overwhelming majority (~92%) of the hashrate coming from one source (address: mqkpEhkPbbtnmqqy1Dobuvq9Wnaytc458k), bigger blocks will be allowed @ 2015-11-09 15:31:00 GMT which is less than 3 hours away as of posting this.
The current backlog is more than 300k transactions according to https://www.blocktrail.com/tBTC Would you consider this a proper test, given that almost all of the testnet's hashrate is coming from a single source? Should we assume that the owner of the hashrate is honest? Title: Re: Testnet will likely fork to BIP101 Post by: manselr on November 09, 2015, 01:32:26 PM We already see the overwhelming majority (~92%) of the hashrate coming from one source (address: mqkpEhkPbbtnmqqy1Dobuvq9Wnaytc458k), bigger blocks will be allowed @ 2015-11-09 15:31:00 GMT which is less than 3 hours away as of posting this. The current backlog is more than 300k transactions according to https://www.blocktrail.com/tBTC Would you consider this a proper test, given that almost all of the testnet's hashrate is coming from a single source? Should we assume that the owner of the hashrate is honest? Obviously paid by someone with an agenda to be sure that blocks are raised so they can centralize nodes and control Bitcoin. Anyone going for big blocks is just missing the point or wants to control Bitcoin. Title: Re: Testnet will likely fork to BIP101 Post by: worhiper_-_ on November 09, 2015, 01:35:53 PM We already see the overwhelming majority (~92%) of the hashrate coming from one source (address: mqkpEhkPbbtnmqqy1Dobuvq9Wnaytc458k), bigger blocks will be allowed @ 2015-11-09 15:31:00 GMT which is less than 3 hours away as of posting this. The current backlog is more than 300k transactions according to https://www.blocktrail.com/tBTC Would you consider this a proper test, given that almost all of the testnet's hashrate is coming from a single source? Should we assume that the owner of the hashrate is honest? Obviously paid by someone with an agenda to be sure that blocks are raised so they can centralize nodes and control Bitcoin. Anyone going for big blocks is just missing the point or wants to control Bitcoin. Testnet hashrate is 6.3 Th/s (https://chain.so/testnet/btc), so it could be anyone. Such a hashrate could be coming from just a few machines. Title: Re: Testnet will likely fork to BIP101 Post by: bambou on November 09, 2015, 01:43:24 PM We already see the overwhelming majority (~92%) of the hashrate coming from one source (address: mqkpEhkPbbtnmqqy1Dobuvq9Wnaytc458k), bigger blocks will be allowed @ 2015-11-09 15:31:00 GMT which is less than 3 hours away as of posting this. The current backlog is more than 300k transactions according to https://www.blocktrail.com/tBTC Would you consider this a proper test, given that almost all of the testnet's hashrate is coming from a single source? Should we assume that the owner of the hashrate is honest? Obviously paid by someone with an agenda to be sure that blocks are raised so they can centralize nodes and control Bitcoin. Anyone going for big blocks is just missing the point or wants to control Bitcoin. Testnet hashrate is 6.3 Th/s (https://chain.so/testnet/btc), so it could be anyone. Such a hashrate could be coming from just a few machines. Obviously, consensus and interest seems overwhelming for this.. not :P Title: Re: Testnet will likely fork to BIP101 Post by: kolloh on November 09, 2015, 01:58:46 PM Interesting. I think testnet isn't really used to much though. But the testnet becomes much less useful if it doesn't mirror the normal block structure. It would certainly be interesting to see how the fork affects it though.
Title: Re: Testnet will likely fork to BIP101 Post by: worhiper_-_ on November 09, 2015, 02:15:28 PM How does this affect Bitcoin? Is it possible to create a new testnet if they fork it? It doesn't affect bitcoin at all, the testnet is separate and as the name indicates, for testing purposes only. It's like a replica network but with No reason to create a new testnet though, this fork could be be temporary, there's no profit from throwing significant amounts of hash to a network with useless coins. In the event that two blockchains remain active at the same time services using the testnet will function normally under the non-fork blockchain. Title: Re: Testnet will likely fork to BIP101 Post by: unamis76 on November 09, 2015, 02:28:47 PM Good thing that there are indeed people testing this on the testnet. It was a big and valid argument from people who don't want to change blocksize. How will the network react to this? It will also be interesting to see the behaviour of the clients when the fork deploys.
Despite this, the conditions for testing are indeed not ideal, but I guess it will give us a general idea of what's to come. Title: Re: Testnet will likely fork to BIP101 Post by: Mickeyb on November 09, 2015, 02:41:30 PM It seems like there's an active effort to make bitcoin's testnet fork to BIP101. As seen on reddit (https://www.reddit.com/r/Bitcoin/comments/3s0uii/bip_101_has_activated_on_the_testnet_need_your/): Quote We have activated BIP 101 on the testnet. The hashpower there after activation is now less than the non-BIP 101 hashpower. This makes for an interesting test. If you would like to help gather data, please run the BIP 101 client of your choice with -testnet and keep the debug.log file to share for data analysis later. The more nodes the better! Transaction spam is welcome and encouraged :) What's interesting is that the hashrate that's causing seems to be coming from a single participant. https://www.reddit.com/r/Bitcoin/comments/3s0uii/bip_101_has_activated_on_the_testnet_need_your/cwtkyka What are your thoughts? Great, thats what testnet is for. Exactly. But this should have done a long time ago. Testing should be paramount for such a change to the protocol. Well better ever than never! I agree as well, this is what testnet serves for. BIP 101 is very radical change proposal so let's first see what will it give in the near world performance. Title: Re: Testnet will likely fork to BIP101 Post by: Kprawn on November 09, 2015, 03:16:22 PM Is this not what was supposed to happen with the Testnet? You go there, test your concept and see if it has any merit and go from there? I guess a real world scenario would be
a better result, but a active Blockchain should not be used as a Testnet. It would be less expensive to prove the merit of a change on a Testnet, but most would contest that it's not a real world result on a live environment. :P Title: Re: Testnet will likely fork to BIP101 Post by: shorena on November 09, 2015, 09:05:27 PM I tried to follow the blocks on blocktrail[1] durring the day. From what I saw the following happend
#1 some time around ~1400-1600 CET a fork happened and the last main block was suddently 12 hours old #2 some time between #1 and ~2030 CET the difficulty reset to 1. #3 right now (since ~2000 CET) a very high number of blocks are found due to the reset difficulty most of them are v4 some are BIP101. Code: 2015-11-09 20:36:29 UpdateTip: new best=000000000077942bdaa940e21f6440028b4e03d97a78e426ad5b92ed2e3d800a height=584525 log2_work=66.690577 tx=6640517 date=2015-11-09 22:26:43 progress=1.000003 cache=0.1MiB(167tx) [1] https://www.blocktrail.com/tBTC/blocks/1 Title: Re: Testnet will likely fork to BIP101 Post by: worhiper_-_ on November 10, 2015, 12:33:57 AM Is there any blockexplorer that reports on the fork? I'd assume that the blockexplorer would have to be running a testnet node compatible with the bip101 fork in order to be able to spot the blocks instead of seing them as invalid.
Title: Re: Testnet will likely fork to BIP101 Post by: saturn643 on November 10, 2015, 02:01:50 AM Is there any blockexplorer that reports on the fork? I'd assume that the blockexplorer would have to be running a testnet node compatible with the bip101 fork in order to be able to spot the blocks instead of seing them as invalid. Insight would probably work since it just queries the rpc server on Bitcoin Core. Bitcoin core would obviously need the bip101 patch in order for it to actually work with the fork.I am interested in seeing the behavior of Bitcoin Core running pre-fork rules and seeing how the transactions work. I want to see what technical effects having two working forked blockchains can have since the hashrate of the testnet is so low, it is completely possible to run two simultaneous chains. Also, has the fork actually happened there yet? Title: Re: Testnet will likely fork to BIP101 Post by: European Central Bank on November 10, 2015, 02:22:42 AM Why aren't there multiple testnets? It seems like a strange idea to only be able to run one experiment at a time. I can appreciate it still requires miners and nodes but I would've expected they'd allocate resources to different concepts.
Title: Re: Testnet will likely fork to BIP101 Post by: worhiper_-_ on November 10, 2015, 03:05:15 AM Why aren't there multiple testnets? It seems like a strange idea to only be able to run one experiment at a time. I can appreciate it still requires miners and nodes but I would've expected they'd allocate resources to different concepts. Good question. I assume it's mostly due to lack of interest. In order to be able to test things on a network replicating bitcoin's one this network would need to have a decent number of participants. Starting an entirely new testnet out of the blue to test something temporarily likely wouldn't attract a decent amount of participants willing to put effort and resources into building it and maintaining for the tests. So the existing test net was used. Title: Re: Testnet will likely fork to BIP101 Post by: YarkoL on November 10, 2015, 07:54:42 AM Why aren't there multiple testnets? It seems like a strange idea to only be able to run one experiment at a time. I can appreciate it still requires miners and nodes but I would've expected they'd allocate resources to different concepts. They are (or actually, he is - for time being), see jtoomins post on Reddit https://www.reddit.com/r/Bitcoin/comments/3s0uii/bip_101_has_activated_on_the_testnet_need_your/cwtkyka Title: Re: Testnet will likely fork to BIP101 Post by: alani123 on November 10, 2015, 07:57:37 AM Why aren't there multiple testnets? It seems like a strange idea to only be able to run one experiment at a time. I can appreciate it still requires miners and nodes but I would've expected they'd allocate resources to different concepts. They are, see jtoomins post on Reddit https://www.reddit.com/r/Bitcoin/comments/3s0uii/bip_101_has_activated_on_the_testnet_need_your/cwtkyka This fork seems to be taking place on the one existing testnet, I think the last sentence in that post is nothing more that a hypothesis. Title: Re: Testnet will likely fork to BIP101 Post by: worhiper_-_ on November 10, 2015, 07:59:51 AM http://xtnodes.com/#testnet now shows blocks in the fork that highlights blocks that are larger than the previous block size cap 1 Mb.
Title: Re: Testnet will likely fork to BIP101 Post by: YarkoL on November 10, 2015, 08:11:11 AM This fork seems to be taking place on the one existing testnet, I think the last sentence in that post is nothing more that a hypothesis. Because if you're going to test forking in somewhat realistic scenario, you can only do it on "official" testnet. Performance on the other hand can be tried out on a private testnet. Title: Re: Testnet will likely fork to BIP101 Post by: spin on November 10, 2015, 08:45:08 AM Please note that this doesn't mean you need to run a BIP101 node to use the testnet. Regular nodes will work fine and will just ignore the non-bitcoin blocks.
Why aren't there multiple testnets? It seems like a strange idea to only be able to run one experiment at a time. I can appreciate it still requires miners and nodes but I would've expected they'd allocate resources to different concepts. Good question. I assume it's mostly due to lack of interest. In order to be able to test things on a network replicating bitcoin's one this network would need to have a decent number of participants. Starting an entirely new testnet out of the blue to test something temporarily likely wouldn't attract a decent amount of participants willing to put effort and resources into building it and maintaining for the tests. So the existing test net was used. There is functionality to run your own private simulated testnet. To test things out. |