Time to roll out bigger blocks
------------------------------
I'm going to submit a pull request to the 0.11 release of Bitcoin Core that
will allow miners to create blocks bigger than one megabyte, starting a little
less than a year from now.
I will be writing a series of blog posts, each addressing one argument against
raising the maximum block size, or against scheduling a raise right now. These
are the objections I plan on writing about; please send me an email
(gavinandresen@gmail.com) if I am missing any arguments for why one megabyte is
the best size for Bitcoin blocks over the next few years.
* "There is no need to raise the maximum block size right now, the average
block size is only about 400,000 bytes."
* "There is no need to raise the maximum block size because the Lightning
;I%6b
>P6|
Network / Sidechains / ChainDB / Treechains / Factom will solve the scaling
problem."
* "The network will be more secure with one megabyte blocks, because there will
be more competition among transactions, and, therefore, higher transaction
fees for miners."
* "More transactions means more bandwidth and CPU and storage cost, and more
cost means increased centralization because fewer people will be able to
afford that cost."
* "More transactions makes it more difficult to keep Bitcoin activity private
from oppressive governments."
* "Larger-than-one-megabyte blocks have had insufficient testing and/or
insufficient research into economic implications and/or insufficient security
review of the risks versus benefits."
* "Any change that requires a hard fork will open up Pandora's Box and destroy
*jJN
!<VU.
the confidence people have in the stability of Bitcoin; if the one megabyte
maximum block size limit can be changed, why not the total number of bitcoins
issued?"
If I am being unfair in how I am stating any of the above objections, please
let me know via email.
-http://gavinandresen.ninja/time-to-roll-out-bigger-blocks
Why increasing the max block size is urgent
-------------------------------------------
Perhaps the most common objection I hear to raising the maximum block size from
one megabyte is that "blocks aren't full yet, so we don't have to do anything
(yet)."
It is true that we're not yet at the hard-coded one megabyte block size limit;
on average, blocks are 30-40% full today.
PLX9
gWas
$oY0D
There is a very good blog post by David Hudson at hashingit.com analyzing what
will happen on the network as we approach 100% full blocks. Please visit that
link for full details, but basically he points out there is a mismatch between
when transactions are created and when blocks are found - and that mismatch
means very bad things start to happen on the network as the one megabyte limit
is reached.
Transactions are created steadily over time, as people spend their bitcoin.
There are daily and weekly cycles in transaction volume, but over any
ten-minute period the number of transactions will be roughly equal to the
number of transactions in the previous or next ten minute period.
Blocks, however, are created via a random Poisson process. Sometimes a lot of
blocks are found in an hour, sometimes all the miners will be unlucky and very
few (or none!) will be found in a hour.
U!cW
The mismatch between the steady submission of transactions to the network and
the random Poisson distribution of found blocks means we will never have blocks
that are 100% full all of the time. Sometimes miners will find a lot of blocks
in a row, clearing out the queue of waiting transactions.
Conversely, very bad things can happen when miners happen to be unlucky. The
queue of transactions waiting to be confirmed will grow, using more and more
memory inside every full node. Full nodes could (and probably will in a future
release of Bitcoin Core) start to drop transactions from the queue, which will
make transaction confirmation less reliable.
If the wallet re-broadcasts transactions if they are not confirmed after a few
blocks (the Bitcoin Core wallet does), then bandwidth usage spikes as every
wallet on the network rebroadcasts its unconfirmed transactions.
If the number of transactions waiting gets large enough, the end result will be
V"y=
aQb6
~ZR5
{k?,
!nb&"
an over-saturated network, busy doing nothing productive. I don't think that is
likely - it is more likely people just stop using Bitcoin because transaction
confirmation becomes increasingly unreliable.
-http://gavinandresen.ninja/why-increasing-the-max-block-size-is-urgent
B$]|