block.version=1 blocks will all be orphaned soon

jl2012

legendary

Activity: 1792

Merit: 1087

Quote from: caveden on March 27, 2013, 04:37:29 AM

I think I was sleeping while all this happened.

Could somebody please explain me one thing.... The whole point of this "version 2 blocks" was just to include the height in the coinbase? I mean... can't the height just be calculated? Couldn't local databases add the height on their data, in an extra field of some kind? Need it be on the protocol?

I fail to see why going through all this hassle just to add the block height to coinbase...

It's to avoid identical coinbase transactions: https://en.bitcoin.it/wiki/BIP_0034

caveden

legendary

Activity: 1106

Merit: 1004

I think I was sleeping while all this happened.

Could somebody please explain me one thing.... The whole point of this "version 2 blocks" was just to include the height in the coinbase? I mean... can't the height just be calculated? Couldn't local databases add the height on their data, in an extra field of some kind? Need it be on the protocol?

I fail to see why going through all this hassle just to add the block height to coinbase...

spunit262

newbie

Activity: 18

Merit: 0

Quote from: Nancarrow on March 26, 2013, 07:22:10 PM

Quote from: evilpete on March 25, 2013, 06:49:58 AM

A very large portion of v2 hashing is offline (btcguild - roughly 30-40%; for nearly 3 hours now) and a couple of v1 miners got lucky in a row.

The rest of the v2 network overtook them and the network reorganized.

Excuse the tinfoil hat, but, notwithstanding Eleuthria's correction of your percentages, did we just see an attempt to sabotage Bitcoin? I know Eleuthria has had the occasional DDOS before, but this particular one at this particular time? I sure hope the dev team and a few pool operators are busy combing through transaction logs with forensic attention to detail...

Given all the hubbub about his pool getting close to 51%, it could have been an extremely miss guided attempt to save Bitcoin.

Nancarrow

hero member

Activity: 492

Merit: 500

Quote from: evilpete on March 25, 2013, 06:49:58 AM

A very large portion of v2 hashing is offline (btcguild - roughly 30-40%; for nearly 3 hours now) and a couple of v1 miners got lucky in a row.

The rest of the v2 network overtook them and the network reorganized.

Excuse the tinfoil hat, but, notwithstanding Eleuthria's correction of your percentages, did we just see an attempt to sabotage Bitcoin? I know Eleuthria has had the occasional DDOS before, but this particular one at this particular time? I sure hope the dev team and a few pool operators are busy combing through transaction logs with forensic attention to detail...

eleuthria

legendary

Activity: 1750

Merit: 1007

Quote from: evilpete on March 25, 2013, 06:49:58 AM

A very large portion of v2 hashing is offline (btcguild - roughly 30-40%; for nearly 3 hours now) and a couple of v1 miners got lucky in a row.

The rest of the v2 network overtook them and the network reorganized.

Only about 3 TH/s of BTC Guild was offline. The DDoS was only affecting getwork and a few people using the very first versions of Stratum which couldn't connect properly without being redirected from getwork.

jgarzik

legendary

Activity: 1596

Merit: 1091

Quote from: jl2012 on March 25, 2013, 04:48:16 AM

What happens to the few coinbase transactions with duplicate transaction id? Are the coins irrecoverably lost?

Yes. The newer coinbase stomps the older coinbase, making the older coinbase unspendable.

evilpete

member

Activity: 77

Merit: 10

A very large portion of v2 hashing is offline (btcguild - roughly 30-40%; for nearly 3 hours now) and a couple of v1 miners got lucky in a row.

The rest of the v2 network overtook them and the network reorganized.

jl2012

legendary

Activity: 1792

Merit: 1087

Quote from: jl2012 on March 25, 2013, 05:20:47 AM

Quote from: solex on March 25, 2013, 04:59:12 AM

http://blockchain.info/block-index/368023/0000000000000124fefda3e554e72cd08aad4b9caaa6e4c3221ae93238180c0c

latest block 227938 shown as v1, so I would expect this to be orphaned, although blockchain has it shown as main chain still...

edit: blockexplorer shows a different block, v2
http://blockexplorer.com/block/00000000000001f205a713a6e23dd2f8c3eed42ffbd765bdfe54d353757ee18a

The latest 2 blocks (227938, 227939) are both version 1. What's going on??? A fork??

Something strange happened..... I'm running 0.8.1 and only have 00000000000001f205a713a6e23dd2f8c3eed42ffbd765bdfe54d353757ee18a (227938, v2)

However, blockchain.info has no record of 00000000000001f205a713a6e23dd2f8c3eed42ffbd765bdfe54d353757ee18a, and have 0000000000000124fefda3e554e72cd08aad4b9caaa6e4c3221ae93238180c0c (227938, v1) and 000000000000025fcc58a882db6c3b47ea90935b522c08d279e415b9529de41d (227939, v1) instead

jl2012

legendary

Activity: 1792

Merit: 1087

Quote from: solex on March 25, 2013, 04:59:12 AM

http://blockchain.info/block-index/368023/0000000000000124fefda3e554e72cd08aad4b9caaa6e4c3221ae93238180c0c

latest block 227938 shown as v1, so I would expect this to be orphaned, although blockchain has it shown as main chain still...

edit: blockexplorer shows a different block, v2
http://blockexplorer.com/block/00000000000001f205a713a6e23dd2f8c3eed42ffbd765bdfe54d353757ee18a

The latest 2 blocks (227938, 227939) are both version 1. What's going on??? A fork??

solex

legendary

Activity: 1078

Merit: 1002

100 satoshis -> ISO code

http://blockchain.info/block-index/368023/0000000000000124fefda3e554e72cd08aad4b9caaa6e4c3221ae93238180c0c

latest block 227938 shown as v1, so I would expect this to be orphaned, although blockchain has it shown as main chain still...

edit: blockexplorer shows a different block, v2
http://blockexplorer.com/block/00000000000001f205a713a6e23dd2f8c3eed42ffbd765bdfe54d353757ee18a

jl2012

legendary

Activity: 1792

Merit: 1087

What happens to the few coinbase transactions with duplicate transaction id? Are the coins irrecoverably lost?

jgarzik

legendary

Activity: 1596

Merit: 1091

This event is now past.

keystroke

hero member

Activity: 900

Merit: 1014

advocate of a cryptographic attack on the globe

Quote from: solex on March 24, 2013, 04:08:08 PM

The reason all this is important is not just that some miners will be forced to upgrade or give up Bitcoin when the 95% is reached. What's important is that this is a proof of concept for implementation of changes in the "wish list" and "hard fork" baskets.

The block version mechanism for a majority consensus on accepting such changes is essential if Bitcoin is to succeed by capturing a significant chunk of the fiat and payments market share. News from the EU this week shows the dark alternative of trusting fiat systems forever.

Yes, exactly. +1
Just got to 94%.

Edit:
11 blocks to go. Block 227930 should be it.

solex

legendary

Activity: 1078

Merit: 1002

100 satoshis -> ISO code

The reason all this is important is not just that some miners will be forced to upgrade or give up Bitcoin when the 95% is reached. What's important is that this is a proof of concept for implementation of changes in the "wish list" and "hard fork" baskets.

The block version mechanism for a majority consensus on accepting such changes is essential if Bitcoin is to succeed by capturing a significant chunk of the fiat and payments market share. News from the EU this week shows the dark alternative of trusting fiat systems forever.

evilpete

member

Activity: 77

Merit: 10

Who? http://blockorigin.pfoe.be/top.php implies that coinlabs and btc warp haven't updated. But you don't have to be a pool to mine, it could be somebody with a gpu that's solo mining and getting lucky, a group of friends, somebody keeping their company's PC's busy overnight.. who knows?

FWIW, in age order:
50th: 226951
49th: 226954
48th: 226960
47th: 226971
46th: 226974

Assuming no more v1's, then the "last 1000" will have 50 x v1's and hence 95% at 227951.

One more v1 block pushes it to 227954. Two more -> 227960. Then 971, and 974.

At the current 855, that's 96 to go, best case.

keystroke

hero member

Activity: 900

Merit: 1014

advocate of a cryptographic attack on the globe

Quote from: evilpete on March 24, 2013, 02:13:59 PM

A fresh v1 turned up at 227835, and that caused the finish line to jump forward.

The 50th oldest v1 is currently at 226951, so the cutover now should be 227951. As of writing, the block number is 227851 so that's 100 to go.

Who might these unknowns be who haven't upgraded? Solo miners? Botnets still possible?

evilpete

member

Activity: 77

Merit: 10

A fresh v1 turned up at 227835, and that caused the finish line to jump forward.

The 50th oldest v1 is currently at 226951, so the cutover now should be 227951. As of writing, the block number is 227851 so that's 100 to go.

keystroke

hero member

Activity: 900

Merit: 1014

advocate of a cryptographic attack on the globe

Quote from: evilpete on March 24, 2013, 03:01:20 AM

FYI, as of 227743, counting v1 blocks backwards puts the 50th version=1 block at about 226930.

That means, at 227930 (assuming no more v=1 blocks), there will be 950 of the last 1000 and a 95% supermajority.

That's roughly 185 blocks from now. 24-30 hours. It should be all over in time for breakfast in the US on Monday morning.

Of course, there's probably going to be 5-10 more v1 blocks so I'm guestimating 227940.

Edit: 24-30 hours is wishful thinking given the spike in hash rate. Probably < 24 hours now.

Thanks for doing this analysis!

Just before 1pm EST: We are currently at 62 out of the latest 1000 blocks on version 1 (6.20%) and 938 out of the latest 1000 blocks on version 2 (93.80%)

evilpete

member

Activity: 77

Merit: 10

FYI, as of 227743, counting v1 blocks backwards puts the 50th version=1 block at about 226930.

That means, at 227930 (assuming no more v=1 blocks), there will be 950 of the last 1000 and a 95% supermajority.

That's roughly 185 blocks from now. 24-30 hours. It should be all over in time for breakfast in the US on Monday morning.

Of course, there's probably going to be 5-10 more v1 blocks so I'm guestimating 227940.

Edit: 24-30 hours is wishful thinking given the spike in hash rate. Probably < 24 hours now.

Graet

vip

Activity: 980

Merit: 1001

Quote from: antimattercrusader on March 24, 2013, 12:57:32 AM

Quote from: DBG on March 18, 2013, 07:10:40 PM

So long and thanks for all the bitcoin Deepbit?

I am missing something, is deepbit not compatible with the updates?

Please fill me in!

wishful thinking
http://blockorigin.pfoe.be/top.php
Deepbit updated to v2
coinlabs and btcwarp are the only vulnerable pools now

Topic: block.version=1 blocks will all be orphaned soon (Read 7211 times)