Before this change, if you made a typo such as `-parellel-8` when trying to run
tests, you'd get a backtrace that's difficult to interpret. With this change,
you'll get a better error message and a non-zero exit code.
f0f50cf9c introduced a compile error on qt4 because qHash only takes
an int argument from qt5+. This fixes that by converting the seed
into a string and both prepending and appending the seed to the
datadir whenever compiling on qt4.
Adds checkpoint and assumevalid for mainnet block 5050000:
- hash: e7d4577405223918491477db725a393bcfc349d8ee63b0a4fde23cbfbfd81dea
- work: 000000000000000000000000000000000000000000000e993d2aa86cf246a49b
- txs: 226128837
- timestamp: 1705383360 (2024-01-16)
Compares tx delta with block from 2023-01-16, 4556625:
- hash: 269ad8789a2792c9e6db17a3c98314d3c3931407689fa730075c9909511100e0
- txs: 92752025
(226128837 - 92752025) / 31536000 = 4.23 tx/sec
---
Adds checkpoint and assumevalid for testnet block 5900000:
- hash: 199bea6a442310589cbb50a193a30b097c228bd5a0f21af21e4e53dd57c382d3
- work: 00000000000000000000000000000000000000000000000000540f2f62ff9a26
- txs: 7391928
- timestamp: 1703511130
Because testnet has seen increased mining for 9 months or so but this
has since ceased, the tx delta was calculated over the whole period since
last checkpoint:
(7391928 - 5353803) / (1703511130 - 1657647467) = 0.04 tx/s
- Removes zipping of XCode instruction
- Removes the boost source compile, no longer needed since 41406bf
- Removes the bdb source compile, no longer needed since e23f14c
- Adds instructions for using protobuf@21
- Adds libevent as a dependency
- Use berkeley-db@5
- Formats to be closer to the other build guides
- Remove $-signs in front of commands, to make it easier to c&p
- Updates the brew installation script
- Replaces the old build-osx guide because that is unmaintained
and this one is better
- Now tested on Sonoma and Ventura, x86_64 and arm64
Remove unmaintained translations mentioned in 2a0c36b6, except for
bitcoin_fr_CA.ts, as there has been activity on this, indicating
that this is important to support.
This commit adds a configuration option, `enable-bip70`, to enable this
feature. Only enable this if you know what you're doing.
Please note that BIP-21 payment links continue to work.
Co-authored-by: Old Dip Tracker <patricklodder@users.noreply.github.com>
rebranded strings from "satoshi" to "koinu"
Revert "rebranded strings from "satoshi" to "koinu""
This reverts commit f03c63a4d3489ed4945eebbb63eef496438e6a4c.
Revert "rebrand strings containing "satoshi(s)""
This reverts commit ae0c5ebfff83297165a65dc94be1ad13fbb2b99c.
rebrand strings (koinu)
rebrand strings (koinu)
changed dates back
Changes the static parameters used in CBlockPolicyEstimator to
support the Dogecoin network, inspired by changes made in Bitcoin
Core 0.15, and field-tested to survive and recover from peak
inflow of transactions.
In addition to changing the parameters:
- documented each changed parameter and the rationale in-line
- changed the fee estimator tests to take into account the new
configuration. Note that only 8 buckets can be tested on a set
of 10, when confidence is set to 80%
1. Changes the decay rate of statistical history from a half-life
of 346 blocks to a half-life of 60 blocks. (DEFAULT_DECAY)
This means that the weight of previous block feerate stats on the
current fee estimates becomes smaller, and more recent block
results weigh heavier, so that the estimates react faster to
temporary peaks.
Weight:
age | previous | now
1h | 0.887 | 0.498
2h | 0.786 | 0.248
3h | 0.697 | 0.124
4h | 0.618 | 0.062
5h | 0.548 | 0.031
7h | 0.431 | 0.008
10h | 0.301 | 0.001
24h | 0.056 | 0.001
2. Changes the historical confidence level (the percentage of
transactions included at a certain level) from 95% to 80%.
(MIN_SUCCESS_PCT)
This lowers the successful inclusion rate because miners in the
Dogecoin ecosystem tune mempool policy for orphan prevention
more than transaction inclusion (because subsidy > fees). This
negatively impacts the precision of the estimate, but with 10%
of empty blocks and at least 2 orders of magnitude difference
between miner enforced minimum fee rates, higher precision will
cause extremely high feerate estimate, yet still have low
effectivity rates. The value was picked by taking Bitcoin Core's
half-confidence of 60% used for 24-block half-time, and standard
confidence of 85% and picking a workable value between those.
3. Decreases the number of transactions per bucket required to
consider a feerate valid from 1 to 0.1 (SUFFICIENT_FEETXS)
This makes more buckets relevant while filtering for fees to
consider and offsets the more rapid half-life from point 1, as
we still want to consider a feerate included with a single
transaction 10 blocks ago, yet because of decay would not be
sufficient for inclusion (because effective relevance would
be 0.89 with the new parametrization). Value taken from Bitcoin
Core because it's been the default there for v0.15 until v25.0
4. Set the minimum and maximum feerates to be considered in
line with default mempool policies: min 0.001 DOGE/kB and
max 10 DOGE/kB (MIN_FEERATE, MAX_FEERATE)
This is copied 1-on-1 from an earlier attempt by Ross Nicoll that
needed further investigation.
5. Lowers the spacing inside feerate buckets from exponent
1.1 to 1.05 (FEE_SPACING)
This increases the granularity of the buckets and is copied
directly from Bitcoin Core, where this has been the parameter
from v0.15 until v25.0
