FAQ: There's a constant average rate of new Bitcoins created

[Stephen Gornick]

Is that still true?

In the past 24 hours, there were 263 blocks created.  That's an annual rate of 95,995 -- which is more than all the blocks that exist today (92643 blocks, per http://www.bitcoinwatch.com )

From the FAQ, http://www.bitcoin.org/faq
> As of October 30th 2009, there are about 26,000 blocks in the block chain,

This means that since October 30th, 2009, there have been an average of about 174 blocks created each day (calculated as (92,643 - 26,000) / 383 days). Today's completion of 263 blocks is about 50% higher than that average daily number then.  So I'm confused.

And also, ... just so I understand this:
So today's miners of the 263 blocks got 50 BTC per block, thus the day's inflation of the Bitcoin supply was 13,150 BTC, (or roughly about US $3,200.)

Is that right?

[caveden]

There's another topic talking about this too.

The difficulty factor is updated only at each 2016 blocks if I'm not mistaken (why such a high number, anybody knows?). When it's updated, the new difficulty factor is calculated based on the average value that would be necessary to keep the 6 blocks per hour rate.
But if the computing power in the network keeps raising, even after the update the new difficulty will be to small to keep that rate, since it was calculated with the average of the last 2016 blocks.

And the computing power has been raising a lot with all these GPU miners.

[MoonShadow]

There's another topic talking about this too.

The difficulty factor is updated only at each 2016 blocks if I'm not mistaken (why such a high number, anybody knows?).

(6 blocks per hour target) * (24 hours) * (14 days) = 2016

Two weeks is the target interval between difficulty adjustments, but it is never exactly two weeks because difficulty adjustments occur at a block, not at a time.

[MoonShadow]

Is that still true?

It's a goal of the protocol, not an absolute, but generally speaking it is true.

In the past 24 hours, there were 263 blocks created.  That's an annual rate of 95,995 -- which is more than all the blocks that exist today (92643 blocks, per http://www.bitcoinwatch.com )

24 hours is not a very reliable data set, but as new users join the network and new clients become more efficient at mining, the total-proof-of-work of the network as a whole increases, but difficulty adjusts only every two weeks and is limited to a change of 4 times.  So over the past several months, the hashing ability of the network has grown faster than the protocol's ability to adapt, resulting in more blocks generated over that same time period than was predicted.

[Stephen Gornick]

[...] but difficulty adjusts only every two weeks and is limited to a change of 4 times.

What does that "4 times" mean.

Found this, ...
http://nullvoid.org/bitcoin/difficultiez.php

[FreeMoney]

[...] but difficulty adjusts only every two weeks and is limited to a change of 4 times.

What does that "4 times" mean.

Found this, ...
http://nullvoid.org/bitcoin/difficultiez.php

Do you see where it went from 45 to 181? It would have increased by more, but it is limited to a 4x increase or decrease. This is to keep a temporary surge of power on the network from making blocks take too long for a very long time.

[MoonShadow]

[...] but difficulty adjusts only every two weeks and is limited to a change of 4 times.

What does that "4 times" mean.

Changes in the difficulty are limited to a factor of four, if the difficulty is 4 it maxs out at 16 or drops to 1, or somewhere in the middle.

[caveden]

(6 blocks per hour target) * (24 hours) * (14 days) = 2016

Two weeks is the target interval between difficulty adjustments, but it is never exactly two weeks because difficulty adjustments occur at a block, not at a time.

Yes, I got it, I just don't see why such a large time interval. Why not daily, for example? It would be much more precise...

[MoonShadow]

(6 blocks per hour target) * (24 hours) * (14 days) = 2016

Two weeks is the target interval between difficulty adjustments, but it is never exactly two weeks because difficulty adjustments occur at a block, not at a time.

Yes, I got it, I just don't see why such a large time interval. Why not daily, for example? It would be much more precise...

Well, part of it is simply an arbitrary design decision.  But much of it is based upon assumptions about how the network would act in a future with a huge user base.  Much like the 10 minute target block interval is much longer than is currently neccessary, but in the future the number of nodes and number of transactions in transit would burden a network that tried to move at a much shorter interval.  In the same way, the target adjustment was chosen, in part, to prevent the network from getting out of wack due to an overly-precise time definition across the network.  If the interval is too short, there is the possibility that parts of the network could end up with a slightly different calculation of the target; resulting in parts of the network that can't produce a block acceptable to the rest of the network for some period of time.  This wouldn't be catastrophic, and might actually happen to some degree anyway, but shortening the target to, say, once every 24 hours would magnify this effect.  Also, shortening the interval would also negate the max. difficulty change rule; which limits changes in difficulty to a factor of four.  This rule exists to prevent a network attack intended to drive up the difficulty level to great heights before removing the excessive generation power and forcing the network into an unacceptablely long block interval for an extended period of time.  It limits extreme changes in difficulty, which could otherwise be manipulated into forcing normal users into paying transaction fees higher than would be reasonable, or simply annoy the network users in an effort to undermine faith in the currency system.

I don't really know as much about programming or encryption as most of these guys, but I do know the Economics of a monetary system, and this one is very well thought out.