Blockchain split of 4 July 2015 - page 23.

Mikestang

legendary

Activity: 1274

Merit: 1000

Quote from: BitUsher on July 04, 2015, 12:44:00 AM

Quote from: ?? on ??

Someone just used the alert key.

"Your node software is out of date and may accept an invalid blockchain fork. Do not trust confirmations"

I have 0.10.1, how can that be right Huh

Latest version is 0.10.2. I don't know if that has anything to do with it or not.

I'm guessing its a blanket statement for all old versions, even though 0.10.0 is technically safe

I'm using 0.10.2 and still got the message, it's not a version specific announcement.

gmaxwell

staff

Activity: 4200

Merit: 8441

Quote from: BitUsher on July 04, 2015, 01:33:09 AM

My concerns exactly and stated more elegantly than I could have. Perhaps there needs to be a change to the protocol to prevent this practice?

It's possible to do, prior proposals were dismissed as paranoia. I expect that even the evidence now won't be enough since this expirenced will be discounted as "sure, they were doing that before, but they won't do it again".

And now, you know why chinese pools thought they could survive with 8MB blocks. Ultimately any proposal to inhibit SPV mining is going to interact in politically complex ways with the block size discussion. Sad

BitUsher

legendary

Activity: 994

Merit: 1034

Quote from: gmaxwell on July 04, 2015, 01:27:04 AM

There was about 4% of the hashrate remaining mining old invalid blocks; these SPV miners would extend them-- effectively boosting the invalid-block mining portion of the hashrate to 50% the moment one of those 4% miners produced an invalid block.

Miners that do not verify, even if they don't include transactions, are a huge amplifier on anyone who mines an invalid or malicious block for whatever reason; and are obviously a grave risk to the system.

More of these forks may still happen-- if these large hashrate miners leave the SPV mining code in place, there still is a few percent mining invalid blocks--- I know for the deployment of the P2SH softfork "50 BTC" continued mining invalid blocks for roughly a month.

My concerns exactly and stated more elegantly than I could have. Perhaps there needs to be a change to the protocol to prevent this practice?

gmaxwell

staff

Activity: 4200

Merit: 8441

Quote from: zvs on July 04, 2015, 01:11:46 AM

from what? all those 1 transaction (the 25btc) blocks?

nobody is going to get scammed with fake transactions, it's obv not intentional.. if it was, they'd be including a lot of their (own) transactions, instead they are including zero

also, re: blockchain has 'thousands of orphan blocks a day', huh? more like 1 or 2 on average

ed: the only 'scamming' from transactions that could possible happen were with the original f'cked up block @ https://blockchain.info/block/0000000000000000009cc829aa25b40b2cd4eb83dd498c12ad0d26d90c439d99

There was about 4% of the hashrate remaining mining old invalid blocks; these SPV miners would extend them-- effectively boosting the invalid-block mining portion of the hashrate to 50% the moment one of those 4% miners produced an invalid block.

Miners that do not verify, even if they don't include transactions, are a huge amplifier on anyone who mines an invalid or malicious block for whatever reason; and are obviously a grave risk to the system.

More of these forks may still happen-- if these large hashrate miners leave the SPV mining code in place, there still is a few percent mining invalid blocks--- I know for the deployment of the P2SH softfork "50 BTC" continued mining invalid blocks for roughly a month.

BitUsher

legendary

Activity: 994

Merit: 1034

Quote from: Pony789 on July 04, 2015, 01:11:29 AM

According to https://getaddr.bitnodes.io/nodes/, only 62% of full nodes are using bitcoin core 0.10.0, 0.10.1 and 0.10.2. Hopefully the alert message will make the remaining users to upgrade.

Except most are using SPV wallets and wont see the alert message.

BTC Nuggets (the trigger in this case) still has 2.2Ph/s and could start this fork again with F2Pool picking up the wrong block and not validating.

Quote from: zvs on July 04, 2015, 01:11:46 AM

ed: the only 'scamming' from transactions that could possible happen were with the original f'cked up block @ https://blockchain.info/block/0000000000000000009cc829aa25b40b2cd4eb83dd498c12ad0d26d90c439d99

It is doubtful that a scammer could have reacted quick enough but during the fork double spending could have occurred, especially with SPV wallets for any transactions spent in BTC Nuggets block that triggered this.

scarsbergholden

hero member

Activity: 686

Merit: 500

whats the deal now in days everyone is calling about forks like this wasn't happening before, why are miners so hard headed when it comes to an update.

zvs

legendary

Activity: 1680

Merit: 1000

https://web.archive.org/web/*/nogleg.com

Quote from: Quantus on July 03, 2015, 11:37:27 PM

So many people are going to get scammed tonight with fake transactions.

from what? all those 1 transaction (the 25btc) blocks?

nobody is going to get scammed with fake transactions, it's obv not intentional.. if it was, they'd be including a lot of their (own) transactions, instead they are including zero

also, re: blockchain has 'thousands of orphan blocks a day', huh? more like 1 or 2 on average

ed: the only 'scamming' from transactions that could possible happen were with the original f'cked up block @ https://blockchain.info/block/0000000000000000009cc829aa25b40b2cd4eb83dd498c12ad0d26d90c439d99

Pony789

hero member

Activity: 882

Merit: 1000

According to https://getaddr.bitnodes.io/nodes/, only 62% of full nodes are using bitcoin core 0.10.0, 0.10.1 and 0.10.2. Hopefully the alert message will make the remaining users to upgrade.

BitUsher

legendary

Activity: 994

Merit: 1034

Another update from Peter Todd:

"The majority of hashing power is now mining only valid blocks. However, SPV wallets are still vulnerable as they do no validation, and ~4% or so of hashing power is still mining invalid blocks. Don't trust txs in SPV wallets w/o >= 2 confirmations right now."

FYI-- no alert message with 0.10.2 now that I upgraded to test this node.

BitUsher

legendary

Activity: 994

Merit: 1034

Quote from: ?? on ??

Someone just used the alert key.

"Your node software is out of date and may accept an invalid blockchain fork. Do not trust confirmations"

I have 0.10.1, how can that be right Huh

Latest version is 0.10.2. I don't know if that has anything to do with it or not.

I'm guessing its a blanket statement for all old versions, even though 0.10.0 is technically safe

BitUsher

legendary

Activity: 994

Merit: 1034

Nice... I just got the warning in core from the Alert Key. May have been Theymos responding to my IM, or Gavin noticing.

"Your node software is out of date and may accept an invalid blockchain fork. Do not trust confirmations. "

The odd thing is that I received this message on 0.10.0rc1 which is supposed to be safe In this instance. I'm going to update and see if the alert is removed.

edit **** warning was just changed to --

" Warning: This version is obsolete, upgrade required!"

OROBTC

legendary

Activity: 2912

Merit: 1852

...

BitAmigos

Well, this has been perhaps my most exciting Friday night looking at BTC transactions. But, maybe the excitement is over, for now.

I have seen unflattering references to AntPool landing a lot of blocks with just their win as the only transaction. This kind of thing why I (as a non-tekkie) have no interest in trying to mine BTC, too complicated.

* * *

OK, back to my Lee Child book...

BitUsher

legendary

Activity: 994

Merit: 1034

Quote from: ajareselde on July 04, 2015, 12:26:44 AM

Quote from: BitUsher on July 04, 2015, 12:24:28 AM

Current status: F2Pool still broken; Antpool fixed (but no promise they won't intentionally re-break in the future).

How is it fixed when they just had another bad one? https://blockchain.info/block/00000000000000000b9d7006b0a893302e02becbc2faf78e79e000bb51721db0
Since thhere are no transactions in these blocks, it should be easier to fix the problem.

cheers

That is actually on the correct chain... Take a look. Antpool appear to have a safeguard to not include transactions under certain conditions as well.

ajareselde

legendary

Activity: 1722

Merit: 1000

Satoshi is rolling in his grave. #bitcoin

Quote from: BitUsher on July 04, 2015, 12:24:28 AM

Current status: F2Pool still broken; Antpool fixed (but no promise they won't intentionally re-break in the future).

How is it fixed when they just had another bad one? https://blockchain.info/block/00000000000000000b9d7006b0a893302e02becbc2faf78e79e000bb51721db0
Since thhere are no transactions in these blocks, it should be easier to fix the problem.

cheers

BitUsher

legendary

Activity: 994

Merit: 1034

Current status: F2Pool still broken; Antpool fixed (but no promise they won't intentionally re-break in the future).

jonald_fyookball

legendary

Activity: 1302

Merit: 1004

Core dev leaves me neg feedback #abuse #political

Quote from: OROBTC on July 04, 2015, 12:11:36 AM

...

AntPool just put up another "bad one", selfish bastids... Block 363741. Whee... (?)

Penalty box for miners who do that? But with everything so de-centralized, how would THAT work?

people will start leaving the pool if they keep up those shenanigans.

OROBTC

legendary

Activity: 2912

Merit: 1852

...

AntPool just put up another "bad one", selfish bastids... Block 363741. Whee... (?)

Penalty box for miners who do that? But with everything so de-centralized, how would THAT work?

EDIT: Thanks, BitUsher, that explains something for me.

BitUsher

legendary

Activity: 994

Merit: 1034

Quote from: jonald_fyookball on July 03, 2015, 11:59:07 PM

Quote from: Quantus on July 03, 2015, 11:58:11 PM

If 1 MB blocks are already to big for mining farms to validate properly won't 8mb blocks just slow down the network even more?

It wasn't the size that was the issue here. The pool wasn't keeping updated with all the BIPs.

You are both right. One Mining pool which mined an invalid block was picked up by other pools who do not verify as they are "SPV mining" . The reason f2Pool and Antpool among others SPV mine instead of fully validate is for a slight~1% edge in latency which does indicate that larger blocks do have an impact upon the economics of mining.

ajareselde

legendary

Activity: 1722

Merit: 1000

Satoshi is rolling in his grave. #bitcoin

Both blockchain and blocktrail are now reporting identical blocks from block 363737, and transactions seam to be passing through.
I'm interested to know what will happen to the transactions from few block before that?

cheers

Shogen

legendary

Activity: 966

Merit: 1001

Quote from: coinableS on July 03, 2015, 11:55:38 PM

Is there something us users can do to help this issue? If I launched my full node and started validating transactions would this help, not help this issue? This seems serious.

If you are using bitcoin core older than 0.9 or older, you should upgrade to 0.10 or newer.
If you have contacts with miners still mining v2 blocks or not rejecting v2 blocks, you could ask them to upgrade as well.

Quote from: OROBTC on July 03, 2015, 11:56:06 PM

...

And WHAT has happened to all the transactions that occurred during that time frame? Do they all get recorded in new blocks? Block 363737 only has some 166 trx listed.

Maybe other tools are better than blockchain.info?

Is there a way to see recently orphaned blocks?

You could use blocktrail.com or btc.blockr.io. They are correctly rejecting v2 blocks AFAIK.

Topic: Blockchain split of 4 July 2015 - page 23. (Read 45594 times)