Mempool is now up to 25.5 MB with 22,200 transactions waiting.

[Dissonance]

I suspect the main "problem" is that people use fixed fees instead of good estimates (like e.g. via core, electrum or "cointap" provides). At least some of the problematic TX come from services that are most likely based on API calls with fixed fees of 10k satoshi per start kbyte.

As these transactions age and the inputs get more confirmations does the probabality of these lower fee transactions confirming become higher ?

[AzzAz]

OK so is mempool growing or what?

[sturle]

I suspect the main "problem" is that people use fixed fees instead of good estimates (like e.g. via core, electrum or "cointap" provides). At least some of the problematic TX come from services that are most likely based on API calls with fixed fees of 10k satoshi per start kbyte.

As these transactions age and the inputs get more confirmations does the probabality of these lower fee transactions confirming become higher ?

Yes, but very slowly if depending on priority alone.  The worst spam storm may be over, btw.  The number of fee paying transactions in my mempool has been slowly decreasing for the last few hours.

[richardsNY]

I suspect the main "problem" is that people use fixed fees instead of good estimates (like e.g. via core, electrum or "cointap" provides). At least some of the problematic TX come from services that are most likely based on API calls with fixed fees of 10k satoshi per start kbyte.

Doesn't it also depends on the pools that mine the blocks? I have noticed that when BTCChina pool mines the next block, my transaction gets confirmed directly. But when an other pool mines the next block, it ignores my transaction. It happened several times. I always include a fee of 10,000 Satoshi.

[bargainbin]

You don't understand the difference between a hard and a soft fork.  A soft fork adds new rules, which can be done without breaking the consensus defined by Bitcoin Core (all versions).

Drop the patronizing tone. I understand the difference between soft & hard forks; it is irrelevant to my point.
The point being: who gets to define consensus. Which, in the linked article, is Core.
P.S. Are you the *humph* author of the linked piece of Core shillery?

Yes, Core defines consensus for Bitcoin, and the core consensus rules were set in stone from the point when Satoshi released version 0.1.  If you don't believe me, read his words for it here.  If you see two posts later down, you will see Gavin defining his future role in Bitcoin.

You are taking the post out of context, if you want to quote satoshi like granny quotes her red letter Bible, here is what satoshi thought of changing the blocksize:

It [higher blocksize limit -ed] can be phased in, like:

if (blocknumber > 115000)
    maxblocksize = largerlimit

It can start being in versions way ahead, so by the time it reaches that block number and goes into effect, the older versions that don't have it are already obsolete.

When we're near the cutoff block number, I can put an alert to old versions to make sure they know they have to upgrade.

Attempts to break the consensus rules will make an altcoin.

Once more: No one is breaking consensus rules -- Core doesn't get to decide what constitutes consensus whenever it feels threatened.

 In principle anyone can do a soft fork by enforcing new rules, stricter than the rules in Core, and this will make a soft fork.  It will not succeed however, unless there is a clear majority supporting it.  Older nodes, all versions, will continue working after a soft fork.

Either you don't understand the difference between a hard and a soft fork, or are having trouble staying focused.
Increasing the blocksize limit is, necessarily, a hard, a hard fork.  

I have been using Bitcoin since 2010, and exchange bitcoins for a living.  The health and development of Bitcoin is vital to my business, and for that reason I follow it closely.  

Don't mean to sound cold now that you're sharin', but I couldn't care less. Irrelevant to the topic at hand.

[sturle]

Yes, Core defines consensus for Bitcoin, and the core consensus rules were set in stone from the point when Satoshi released version 0.1.  If you don't believe me, read his words for it here.  If you see two posts later down, you will see Gavin defining his future role in Bitcoin.

You are taking the post out of context, if you want to quote satoshi like granny quotes her red letter Bible, here is what satoshi thought of changing the blocksize:

It [higher blocksize limit -ed] can be phased in, like:

if (blocknumber > 115000)
    maxblocksize = largerlimit

It can start being in versions way ahead, so by the time it reaches that block number and goes into effect, the older versions that don't have it are already obsolete.

I am not religious.  I quote Satoshi for facts, not beliefs.  Smarter people have taken over development, and have found much more viable ways to scale bitcoin.  You need to understand the difference between how it is and how it can be (or could have been if things were done differently back in 2010).

Attempts to break the consensus rules will make an altcoin.

Once more: No one is breaking consensus rules -- Core doesn't get to decide what constitutes consensus whenever it feels threatened.

Satoshi defined the consensus rules back in 2009, and it is enforced by nodes running the software called Bitcoin Core.  Core cannot change this consensus, and it is not affected by feelings or whatever.  Nobody can.  Not even by trolls on Bitcointalk.  Anyone can make the rules stricter, Core as well, and this will be successful if a majority of hashrate agrees.  Single miners can even enforce stricter rules for only the blocks this miner produce, and it will work.  This won't do any harm.  All nodes validating the blocks will find them to be valid, and all services will continue to work.

There are plenty altcoins adhering to different consensus rules.  Their blocks are invalid to Bitcoin.

In principle anyone can do a soft fork by enforcing new rules, stricter than the rules in Core, and this will make a soft fork.  It will not succeed however, unless there is a clear majority supporting it.  Older nodes, all versions, will continue working after a soft fork.

Either you don't understand the difference between a hard and a soft fork, or are having trouble staying focused.
Increasing the blocksize limit is, necessarily, a hard, a hard fork.

Yes, I do.  A hard fork will essentially make an altcoin.  One valid chain, and one which is only valid for the altcoin.  There are several attempts to make new altcoins with larger blocksizes.  E.g. "BitcoinXT", "Classic",  "Unlimited", "BitPay Core" and whatever.  I have lost count.  Bitcoin has 1 MB blocks.  It is defined by the consensus rules.  Larger blocks will be invalid to Bitcoin nodes.  Those specific altcoins are special in the way they are premined.  You can import a Bitcoin wallet into an altcoin node, split your coins by e.g. mixing with altcoin, and then exchange your altcoins for Bitcoin on Cryptsy or one of the other altcoin exchanges. 

I have been using Bitcoin since 2010, and exchange bitcoins for a living.  The health and development of Bitcoin is vital to my business, and for that reason I follow it closely.  

Don't mean to sound cold now that you're sharin', but I couldn't care less. Irrelevant to the topic at hand.

If you didn't want to know, why did you ask if I was an author?

[toknormal]

I am not religious.  I quote Satoshi for facts, not beliefs.  Smarter people have taken over development, and have found much more viable ways to scale bitcoin.

People also quote Satoshi in the interests of maintaining a coherent vision of a monetary model, not just for 'religious' reasons or for referencing 'facts'.

I've read a lot of your posts and you make a very solid case for supporting the bitcoin "consensus", however you've got to acknowledge that those priorities may come with a potentially damaging price tag of ending up with a very different monetary reality than what was originally envisaged by the "Satoshi" ideal.

For a start, what you call "more viable ways to scale bitcoin" are not really scaling bitcoin at all. They are putting in place scaffolding out in the eco-system so that bitcoin doesn't have to scale. It's basically just 1990's Microsoft Transaction Server funnelling traffic into more digestable chunks. But what are the implications of this for the original vision of a single-tier monetary model ?

Thats the real question that needs investigated IMO. The more bitcoin loses any integral ability to scale or transact instantly, the more it can kiss goodbye any pretence of a role as electronic "cash" in a true monetary sense. The settlement-layer concept is a very different type of animal and is not a financial model that should be stumbled into unconsciously. You can't just say at the start "hey, we're electronic cash" and then later "hey we changed our minds, we're settlement layer for credit transactions cos we had a few technical problems".

The fact is that commercial pressure is remoulding Bitcoin from the original vision of a universally accessible, ubiquitous electronic token, into no more than just another asset class. Although expressed in good faith, IMO your equally religious obsession with definitions of "consensus" is helping to push it down that road.

Nor are concepts like sidechains going to help anything in this regard. All they do is address specialisation at the expense of fungibility (because you're having to recast the currency in a different identity to address a technical deficiency and therefore make one unit distinguishable from another). So all these monetary rules and principles are being broken just in order to address a load of technical deficiencies and preserve the "consensus" fork.

Those priorities seem to me to be the wrong way around. I think that bitcon's technical properties are optimal when they are in harmony with its monetary ones - not in conflict with them. However, the implications of that view are not too palatable to many of the core devs because it means that the "missing space" has to be taken up by altcoins and a true monetary market allowed to develop. They're not too hot on that idea.

That's your problem right there. It isn't technical, it's political.

[stan.distortion]

Interesting discussion, watching.

[hv_]

The mempool was nearly empty at around 1-1.5 two days ago.

No it wasn't-- maybe on that site, but only due to it filtering things.

Because of ongoing low level spam attacks any long running 0.12 node has had a 300MB mempool (which is the limit) for months now.   There are about 800 MBytes of unconfirmed valid transactions that I'm aware of... 

None of this is especially interesting: one should always assume that the number of transactions at a fee of 0 is effectively infinite.

The spammers moved to paying more than 1e-8 BTC per byte; which is still much lower than most ordinary transactions; but high enough to move the needle on tradeblock... this lets people spin a bunch of FUD.  Meanwhile for most people transactions continue to function like normal: a total non-event.

It's NOT a total non-Event: Proof: You posted.

 

[bargainbin]

Yes, Core defines consensus for Bitcoin, and the core consensus rules were set in stone from the point when Satoshi released version 0.1.  If you don't believe me, read his words for it here.  If you see two posts later down, you will see Gavin defining his future role in Bitcoin.

You are taking the post out of context, if you want to quote satoshi like granny quotes her red letter Bible, here is what satoshi thought of changing the blocksize:

It [higher blocksize limit -ed] can be phased in, like:

if (blocknumber > 115000)
    maxblocksize = largerlimit

It can start being in versions way ahead, so by the time it reaches that block number and goes into effect, the older versions that don't have it are already obsolete.

I am not religious.  I quote Satoshi for facts, not beliefs.

That's exactly what Granny tells me about her red letter Bible. And you know what else? When Jesus says something she don't like, like

It [higher blocksize limit -ed] can be phased in, like:

if (blocknumber > 115000)
    maxblocksize = largerlimit

...just like yourself, she points out that her Bible is outdated, and...

Smarter people have taken over development the world, and have found much more viable ways to scale bitcoin make better laws.

> A hard fork will essentially make an altcoin.
In that case, you've been using an altcoin since March 2013. Surprise!

[sturle]

Yes, Core defines consensus for Bitcoin, and the core consensus rules were set in stone from the point when Satoshi released version 0.1.  If you don't believe me, read his words for it here.  If you see two posts later down, you will see Gavin defining his future role in Bitcoin.

You are taking the post out of context, if you want to quote satoshi like granny quotes her red letter Bible, here is what satoshi thought of changing the blocksize:

It [higher blocksize limit -ed] can be phased in, like:

if (blocknumber > 115000)
    maxblocksize = largerlimit

It can start being in versions way ahead, so by the time it reaches that block number and goes into effect, the older versions that don't have it are already obsolete.

I am not religious.  I quote Satoshi for facts, not beliefs.

That's exactly what Granny tells me about her red letter Bible. And you know what else? When Jesus says something she don't like, like

It [higher blocksize limit -ed] can be phased in, like:

if (blocknumber > 115000)
    maxblocksize = largerlimit

...just like yourself, she points out that her Bible is outdated, and...

This was a suggestion.  Satoshi dropped it, because it was a bad idea.  Everyone (almost) can see it is a terrible idea, so it wasn't implemented by anyone else either.  Satoshi had a lot of ideas which he later dropped.  He even asked people to stop GPU mining, because he thought every user should mine with their CPUs.  Go home and tell your granny bitcointalk isn't the Bible.  Or demand a PoW change to make it possible for everyone to mine with their CPUs again, or whatever.  (This is exactly why many of the altcoins exists, btw.)

Satoshi had many conflicting views.  He initially thought Bitcoin would be much more anonymous than it turned out to be.  He also thought the blockchain would reside in a big data center while all users ran SPV wallets, which would completely remove all anonymity.  This is what Coinbase wants, and basically mimics how credit cards work.  Satoshi didn't think of advances like Lightning and sidechains, which would improve on both anonymity and decentralization, but he made the scripting system flexible enough to allow for future advances in the technology.  Why is it so important to you to stop progress?

> A hard fork will essentially make an altcoin.
In that case, you've been using an altcoin since March 2013. Surprise!

Not really.  This was a bugfix.  Entirely uncontroversial.  So far not a single (known to anyone) block has been produced on the other side, so strictly speaking it isn't even a fork.  This did not come as a surprise to me, perhaps to you.

[Laosai]

I'm sorry for the question I'm going to ask but there is something I just can't understand.

Blocks are not at all all full.
So why are the tx in the mempool not filling the blocks which are not full?

[pedrog]

I'm sorry for the question I'm going to ask but there is something I just can't understand.

Blocks are not at all all full.
So why are the tx in the mempool not filling the blocks which are not full?

Some mining pools, AntPool, BW and BTCC, get a lot of empty blocks, miners don't really need to include transactions.

[Laosai]

I'm sorry for the question I'm going to ask but there is something I just can't understand.

Blocks are not at all all full.
So why are the tx in the mempool not filling the blocks which are not full?

Some mining pools, AntPool, BW and BTCC, get a lot of empty blocks, miners don't really need to include transactions.

What's the point of allowing void blocks to me mined? Isn't it a waste of time and energy?

Is there anything to win for the mining pools to not include transactions?

[sturle]

I am not religious.  I quote Satoshi for facts, not beliefs.  Smarter people have taken over development, and have found much more viable ways to scale bitcoin.

People also quote Satoshi in the interests of maintaining a coherent vision of a monetary model, not just for 'religious' reasons or for referencing 'facts'.

I've read a lot of your posts and you make a very solid case for supporting the bitcoin "consensus", however you've got to acknowledge that those priorities may come with a potentially damaging price tag of ending up with a very different monetary reality than what was originally envisaged by the "Satoshi" ideal.

What is Satoshi's ideal?  He wanted to create anonymous cash for the internet.  It turned out to be not anonymous at all.  Should we fix that?

He didn't see any good ways to scale, except for increasing blocksize and centralize the blockchain.  Now that we know better, should we do it?

He didn't want people to mine with their GPUs, becuase he had ment that every user should mine ("vote") with their CPUs.  Then we got FPGA mining, and now we even have large centralized ASIC mining factories.  Should we fix that?

For a start, what you call "more viable ways to scale bitcoin" are not really scaling bitcoin at all. They are putting in place scaffolding out in the eco-system so that bitcoin doesn't have to scale. It's basically just 1990's Microsoft Transaction Server funnelling traffic into more digestable chunks. But what are the implications of this for the original vision of a single-tier monetary model ?

More viable ways scales up the parts of bitcoin which needs to scale up.  We need to make it possible to transact faster, cheaper and more often.  We do not need to increase the amount of resources needed for every node in the system.

Thats the real question that needs investigated IMO. The more bitcoin loses any integral ability to scale or transact instantly, the more it can kiss goodbye any pretence of a role as electronic "cash" in a true monetary sense. The settlement-layer concept is a very different type of animal and is not a financial model that should be stumbled into unconsciously. You can't just say at the start "hey, we're electronic cash" and then later "hey we changed our minds, we're settlement layer for credit transactions cos we had a few technical problems".

The Bitcoin blockchain was designed as a settlement layer.  It has never been possible to transact instantly with Bitcoin.  0 confirmation transactions have the same security as "the check is in the mail".  If I had sent bitcoin to an excange an hour ago now, the transaction would only have four confirmations.  I would need two more confirmations before my coins are credited, and I can exchange them for money.  Lightning can enable instant transactions by building on top of the settlement layer below, where the transaction, or the start and endpoint of a chain of transactions, will be commited later.

The fact is that commercial pressure is remoulding Bitcoin from the original vision of a universally accessible, ubiquitous electronic token, into no more than just another asset class. Although expressed in good faith, IMO your equally religious obsession with definitions of "consensus" is helping to push it down that road.

The commersial pressure from Coinbase and others, which are under strong regulatory pressure to do chain analysis and watch how their users are spending their money, have failed to gain much support.  Instead they fake support by paying for lits of nodes at data centers, like this.  By enabling new technologies, Bitcoin can finally be this ubiquitous electronic token which is usable for everything from large settlements to instant micropayments.

Nor are concepts like sidechains going to help anything in this regard. All they do is address specialisation at the expense of fungibility (because you're having to recast the currency in a different identity to address a technical deficiency and therefore make one unit distinguishable from another). So all these monetary rules and principles are being broken just in order to address a load of technical deficiencies and preserve the "consensus" fork.

I think the opposite is true.  Traditionally Bitcoin has only been usable for relatively fast (~ 1 hour) and cheap settlements worldwide 24/7.  Not really for e.g. paying your bar tab at the end of a long evening.  Not all bartenders will accept "the check is in the mail" as payment when you are about to leave.  Both sidechains, payment channels and Lightning will make Bitcoin more flexible, useful for more purposes, and accessible to more people.

Those priorities seem to me to be the wrong way around. I think that bitcon's technical properties are optimal when they are in harmony with its monetary ones - not in conflict with them. However, the implications of that view are not too palatable to many of the core devs because it means that the "missing space" has to be taken up by altcoins and a true monetary market allowed to develop. They're not too hot on that idea.

I don't think an altcoin for every single purpose is a good idea.  I remember travelling through Europe before the Euro.  I was used to it then, but now I hope we never return to the pre-Euro days.  It was a hassle to keep exchanging my money from one currency to another at every border, and of course the exchangers took a high percentage every time.  Even if I didn't spend anything, I would get home with much less money than I had when I left.  Imagine if each of the states in the US had a different currency.  It would certainly limit free movement of capital inside the country. 

[AliceWonderMiscreations]

I'm sorry for the question I'm going to ask but there is something I just can't understand.

Blocks are not at all all full.
So why are the tx in the mempool not filling the blocks which are not full?

Some mining pools, AntPool, BW and BTCC, get a lot of empty blocks, miners don't really need to include transactions.

What's the point of allowing void blocks to me mined? Isn't it a waste of time and energy?

Is there anything to win for the mining pools to not include transactions?

You can't force miners to include transactions. If you tried, they would just make a bunch of transactions that pay themselves.

[sturle]

What's the point of allowing void blocks to me mined? Isn't it a waste of time and energy?

Is there anything to win for the mining pools to not include transactions?

You can't force miners to include transactions. If you tried, they would just make a bunch of transactions that pay themselves.

You can, via a soft fork (it is among the proposals for the segwit soft fork), force the miners to include proof of validation of the block they build on.  This will get us most of the way.  By dropping the validation of the previous block, or not even downloading it, miners will get a head start on mining the next block.  If they have to validate the previous block, the cost of adding some transactions to the next block is very low compared to the fees they make.

[AliceWonderMiscreations]

What's the point of allowing void blocks to me mined? Isn't it a waste of time and energy?

Is there anything to win for the mining pools to not include transactions?

You can't force miners to include transactions. If you tried, they would just make a bunch of transactions that pay themselves.

You can, via a soft fork (it is among the proposals for the segwit soft fork), force the miners to include proof of validation of the block they build on.  This will get us most of the way.  By dropping the validation of the previous block, or not even downloading it, miners will get a head start on mining the next block.  If they have to validate the previous block, the cost of adding some transactions to the next block is very low compared to the fees they make.

That still doesn't force them to add transactions from the network.

Increased fees make it desirable for them to add transactions from the network, and that only happens if we don't make the blocks bigger.

[sturle]

That still doesn't force them to add transactions from the network.

Correct, it just removes most of the the incentive to produce empty blocks.  Meaning it makes more economic sense to include transactions for extra income.

[AzzAz]

Just received small transaction from SUNDAY. It's moving ...