BIP100, BIP 101 and XT nodes status

[Lauda]

Why on Earth would miners in Beijing make all their connections directly to someone halfway around the world. That's not the way the Internet works.

Halfway? I'm actually around 3.5k miles away from them, so much closer than you are. Are you really sure that you should be discussing the technical details of this? Are you really sure that you know what orphans are and how they get created?
Blocks that are too big are going to result with problems and advantages (for some nations, while others will be at a disadvantage).

[Klestin]

Why on Earth would miners in Beijing make all their connections directly to someone halfway around the world. That's not the way the Internet works.

Halfway? I'm actually around 3.5k miles away from them, so much closer than you are. Are you really sure that you should be discussing the technical details of this? Are you really sure that you know what orphans are and how they get created?

Given that you blacked out your location and distance, I'm not sure how I was supposed to know how close you were.  I was obviously speaking of my own connection, which is... halfway around the world from China, and a faster connection to them as well.

Are you really sure you know the difference between mining operations and full nodes?  Are you aware, for instance, that the mining hardware doesn't need to know anything about transactions to do its work?  The entire discussion of China and its great firewall is a red herring.  It doesn't matter in the slightest.  The moment that a China-based mining operation becomes concerned with their latency or bandwidth in regards to transaction volume, they will move the full node outside of China.  Assuming, of course, that it's even there now.

[achow101]

Why on Earth would miners in Beijing make all their connections directly to someone halfway around the world. That's not the way the Internet works.

Halfway? I'm actually around 3.5k miles away from them, so much closer than you are. Are you really sure that you should be discussing the technical details of this? Are you really sure that you know what orphans are and how they get created?

Given that you blacked out your location and distance, I'm not sure how I was supposed to know how close you were.  I was obviously speaking of my own connection, which is... halfway around the world from China, and a faster connection to them as well.

Are you really sure you know the difference between mining operations and full nodes?  Are you aware, for instance, that the mining hardware doesn't need to know anything about transactions to do its work?  The entire discussion of China and its great firewall is a red herring.  It doesn't matter in the slightest.  The moment that a China-based mining operation becomes concerned with their latency or bandwidth in regards to transaction volume, they will move the full node outside of China.  Assuming, of course, that it's even there now.

If they move the full node out of China, then how do they mine inside China? If they are concerned about latency, then there will also be latency from their miners inside China to the full node outside of China.

[Klestin]

If they move the full node out of China, then how do they mine inside China? If they are concerned about latency, then there will also be latency from their miners inside China to the full node outside of China.

Look up pool mining protocols.  The data transferred from the miner to the node is miniscule, and it doesn't get even a single byte bigger if the block size increases.  The whole discussion of block size impact on China-based mining operations is a red herring.

[achow101]

If they move the full node out of China, then how do they mine inside China? If they are concerned about latency, then there will also be latency from their miners inside China to the full node outside of China.

Look up pool mining protocols.  The data transferred from the miner to the node is miniscule, and it doesn't get even a single byte bigger if the block size increases.  The whole discussion of block size impact on China-based mining operations is a red herring.

I don't think it is a red herring. Just look at what the Chinese pools have to say about this.

Also, the data does get larger if the block size increases. Stratun and GBT both send the miners an array of all of the transactions. If there are more transactions, then the array is larger. This is of course more data to send.

[onemorexmr]

If they move the full node out of China, then how do they mine inside China? If they are concerned about latency, then there will also be latency from their miners inside China to the full node outside of China.

Look up pool mining protocols.  The data transferred from the miner to the node is miniscule, and it doesn't get even a single byte bigger if the block size increases.  The whole discussion of block size impact on China-based mining operations is a red herring.

I don't think it is a red herring. Just look at what the Chinese pools have to say about this.

Also, the data does get larger if the block size increases. Stratun and GBT both send the miners an array of all of the transactions. If there are more transactions, then the array is larger. This is of course more data to send.

stratum is able to send transaction but i am not aware of any pool which uses this feature.

[Klestin]

Also, the data does get larger if the block size increases. Stratun and GBT both send the miners an array of all of the transactions. If there are more transactions, then the array is larger. This is of course more data to send.

Stratum and GBT are designed with transparency in mind, so that the pool members can verify that the pool is above board.  Within a mining operation, there is no reason for this double-check.  So, no, there's no need for the data size to increase at all.

[onemorexmr]

Also, the data does get larger if the block size increases. Stratun and GBT both send the miners an array of all of the transactions. If there are more transactions, then the array is larger. This is of course more data to send.

Stratum and GBT are designed with transparency in mind, so that the pool members can verify that the pool is above board.  Within a mining operation, there is no reason for this double-check.  So, no, there's no need for the data size to increase at all.

there is a reason: if pools would submit transactions pool-miners would be able to decide which transactions to include in a block.

but they dont because its to slow....so only the pool can decide which transactions to put in a block.

[achow101]

Also, the data does get larger if the block size increases. Stratun and GBT both send the miners an array of all of the transactions. If there are more transactions, then the array is larger. This is of course more data to send.

Stratum and GBT are designed with transparency in mind, so that the pool members can verify that the pool is above board.  Within a mining operation, there is no reason for this double-check.  So, no, there's no need for the data size to increase at all.

We're talking about Chinese mining pools right? Not all Chinese miners are solo mining. Most of them are in mining pools, which use Stratum and GBT

[Klestin]

We're talking about Chinese mining pools right? Not all Chinese miners are solo mining. Most of them are in mining pools, which use Stratum and GBT

Do you have something that shows that the pools they're using use Stratum/GBT?  

In either event, pool-based mining, which they clearly use, does not require the download of the transaction data.  The block size does not matter.  You can verify this yourself - connect to any of these pools and watch the data transfer - it is a teeny tiny fraction of the full transaction data volume.

Even a dial-up connection can handle that kind of bandwidth.

Red. Herring.

[Lauda]

The entire discussion of China and its great firewall is a red herring.  It doesn't matter in the slightest.  
-snip-

I think that we are done here.
Opinions differ to which proposal is best. However, the opinion of the average forum member(including me) obviously does not matter in the grand scheme of things. BIP100 is the most likely way that we will be going.

[PolarPoint]

Here, you can see the last 24 hours -- default vs. BIP 100 vs. 8MB. https://www.blocktrail.com/BTC/pools?resolution=24h

The latest results has BIP100 leading with 39.31%.
Isn't BIP100 the variable blocksize proposal? Where is the formula for calculating blocksize?
http://gtf.org/garzik/bitcoin/BIP100-blocksizechangeproposal.pdf

[Klestin]

The entire discussion of China and its great firewall is a red herring.  It doesn't matter in the slightest.  
-snip-

I think that we are done here.

I'm afraid you'll have to actually read my entire post to understand why it doesn't matter in the slightest.  Hint: Pooled miners don't download the full transaction data. It doesn't have to cross the great firewall of china, and in fact can be performed over dial-up.

[Mickeyb]

Here, you can see the last 24 hours -- default vs. BIP 100 vs. 8MB. https://www.blocktrail.com/BTC/pools?resolution=24h

The latest results has BIP100 leading with 39.31%.
Isn't BIP100 the variable blocksize proposal? Where is the formula for calculating blocksize?
http://gtf.org/garzik/bitcoin/BIP100-blocksizechangeproposal.pdf

So I see that BIP100 and old core are neck to neck here. BIP101 is getting killed which is good. Which is the percentage that one of these options needs to get in order to win?

[achow101]

We're talking about Chinese mining pools right? Not all Chinese miners are solo mining. Most of them are in mining pools, which use Stratum and GBT

Do you have something that shows that the pools they're using use Stratum/GBT?  

F2pool uses stratum. They have their miners connect to stratum+tcp://stratum.f2pool.com:3333
BTC China uses stratum. They have their miners connect to stratum+tcp://stratum.btcchina.com:3333
AntPool uses stratum. They have their miners connect to stratum.antpool.com:3333
So yes, Chinese miners are using stratum for their pools

In either event, pool-based mining, which they clearly use, does not require the download of the transaction data.  The block size does not matter.  You can verify this yourself - connect to any of these pools and watch the data transfer - it is a teeny tiny fraction of the full transaction data volume.

Even a dial-up connection can handle that kind of bandwidth.

Red. Herring.

Even so, Chinese miners have their full nodes inside China. While you say that "they can just move their full nodes outside of China" that is easier said than done.

[Klestin]

Once again, stratum is EXTREMELY low bandwidth - even dial-up can handle it easily.  The block size doesn't matter.

Even so, Chinese miners have their full nodes inside China. While you say that "they can just move their full nodes outside of China" that is easier said than done.

What are you basing that on?  Chinese companies rent server space all over the world.

[VeritasSapere]

Here, you can see the last 24 hours -- default vs. BIP 100 vs. 8MB. https://www.blocktrail.com/BTC/pools?resolution=24h

The latest results has BIP100 leading with 39.31%.
Isn't BIP100 the variable blocksize proposal? Where is the formula for calculating blocksize?
http://gtf.org/garzik/bitcoin/BIP100-blocksizechangeproposal.pdf

So I see that BIP100 and old core are neck to neck here. BIP101 is getting killed which is good. Which is the percentage that one of these options needs to get in order to win?

There is no mechanisim in place for forking to BIP100. There is also no fuctioning client that has implemented BIP100. So it is not a viable choice at this time. I hope that changes in the future, even though I prefer BIP101.

https://bitcointalk.org/index.php?topic=1164464.msg12267335#msg12267335

[Klestin]

There is no mechanisim in place for forking to BIP100. There is also no fuctioning client that has implemented BIP100. So it is not a viable choice at this time. I hope that changes in the future, even though I prefer BIP101.

Hey now! Settle down there! They are gonna roll out that fork mechanism, just as soon as it's tested.  And they're gonna test it just as soon as it's coded.  And they're gonna code it just as soon as the pseudocode is released.  And they'll release the pseudocode just as soon as the design is finalized. And they'll finalize the design just as soon as they....

[DooMAD]

I'm really not sold on BIP100 (but I can see why the miners obviously approve).  Why would we trust an entity to choose a variable when we can define it algorithmically like we do with everything else?  I'm still adamant the as-yet-un-BIP'd "Proposal 1" is better than BIP100 in every conceivable way.  If BIP101 gets torpedoed, that proposal should be plan B, not BIP100.  What are we going to allow miners to choose next?  The number of leading zeroes to solve the next block?

There is no mechanisim in place for forking to BIP100. There is also no fuctioning client that has implemented BIP100. So it is not a viable choice at this time. I hope that changes in the future, even though I prefer BIP101.

Hey now! Settle down there! They are gonna roll out that fork mechanism, just as soon as it's tested.  And they're gonna test it just as soon as it's coded.  And they're gonna code it just as soon as the pseudocode is released.  And they'll release the pseudocode just as soon as the design is finalized. And they'll finalize the design just as soon as they....

Well played, nicely done.   

[TibanneCat]

https://www.blocktrail.com/BTC/pools?resolution=24h

We are heading toward consensus which is good news
Support for BIP100 has been steadily growing and currently stands at 58.87%
BIP101? 0.71%