25 mining proof of work #66

david-julien · 2017-06-10T22:57:08Z

Initial mining proof of work code, not all tests written yet

…ange nonce to uint64

jordanschalm

Really good! I think there are some more simplifications we can add

jordanschalm · 2017-06-10T23:07:36Z

blockchain/hash.go

@@ -1,9 +1,38 @@
 package blockchain


Move all the miner stuff into its own package

jordanschalm · 2017-06-10T23:07:57Z

blockchain/hash.go

+	// HashLen is the length in bytes of a hash.
+	HashLen = 32
+	// MaxDifficultyHex is the maximum difficulty value represented as a hex string
+	MaxDifficultyHex = "000000FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF"


Can this be replace by using math.big?

jordanschalm · 2017-06-10T23:09:45Z

blockchain/hash.go

@@ -26,3 +55,43 @@ type Marshaller interface {
 func HashSum(m Marshaller) Hash {
 	return sha256.Sum256(m.Marshal())
 }
+
+// CompareTo compares two hashes, it returns true if the operation of the first hash on the second hash specified by the comparator is true, and false otherwise
+func CompareTo(h1 Hash, h2 Hash, comparator int) bool {


We will only ever be checking for hashes being less than right? Can we simplify this accordingly and maybe remove those comparator constants unless we need arbitrary comparisons?

yes we need all of them for miningheader verification purposes before mining, we'll see if i can clean this up when we start using math.big

math.big uses the exact same arbitrary constants for comparison, so if we wanted to compare two hashes we would have to convert to math.big and then use the same constants.

https://golang.org/src/math/big/float.go?s=43528:43561#L1637

jordanschalm · 2017-06-10T23:10:26Z

blockchain/hash.go

+}
+
+// HexStringToHash converts a big endian hex string to a hash
+func HexStringToHash(s string) Hash {


Can this be replaced by math.big?

jordanschalm · 2017-06-10T23:13:03Z

blockchain/miner.go

+
+// Mine continuously increases the nonce and tries to verify the proof of work until the puzzle is solved
+func (mh *MiningHeader) Mine() bool {
+	if mh.VerifyMiningHeader() == false {


use ! rather than comparing to false

jordanschalm · 2017-06-10T23:13:23Z

blockchain/miner.go

+		return false
+	}
+
+	for mh.VerifyProofOfWork() == false {


jordanschalm · 2017-06-10T23:14:27Z

blockchain/miner.go

+}
+
+// AppendUint32ToSlice converts a uint32 to a byte slice and appends it to a byte slice
+func AppendUint32ToSlice(s *[]byte, num uint32) {


Do we need this function? (and below) seems like they're only called once

jordanschalm · 2017-06-10T23:15:41Z

blockchain/miner.go

+}
+
+// SetMiningHeader sets the mining header
+func (mh *MiningHeader) SetMiningHeader(lastBlock Hash, rootHash Hash, target Hash) {


Rather than a setter can we just use composite instantiation? What does the setter buy us?

ie. mh := MiningHeader{lastBlock, rootHash, target}

it sets the time to the time variable to the current time in the miningheader struct, not quite sure how to set default values in golang structs

The default value is the zero value for the type. You can do something like:

mh := MiningHeader{ LastBlock: lastBlock, ... timeField: time.Now(), }

bfbachmann

Great job 🥇

coveralls · 2017-06-13T21:47:05Z

Coverage decreased (-7.1%) to 79.248% when pulling 0666a58 on 25-mining-proof-of-work into b120ac0 on dev.

coveralls · 2017-06-13T21:49:27Z

Coverage decreased (-7.1%) to 79.248% when pulling 0666a58 on 25-mining-proof-of-work into b120ac0 on dev.

…rs and testing

…o 25-mining-proof-of-work

coveralls · 2017-06-14T18:42:18Z

Coverage decreased (-0.06%) to 86.316% when pulling 9401181 on 25-mining-proof-of-work into b120ac0 on dev.

chadlagore · 2017-06-17T19:31:12Z

blockchain/hash.go

+// HexString returns the hexadecimal representation of the hash
+func (h Hash) HexString() string {
+	var hashBigEndian Hash
+	for i := 0; i < HashLen; i++ {


I notice this pattern of inverting the hex string a couple of times, perhaps a helper function?

jordanschalm

Awesome, if we can remove dependence on literal hex strings and use math.big instead that would be awesome, otherwise ready to merge with some more test coverage.

jordanschalm · 2017-06-17T19:59:40Z

blockchain/hash_test.go

+		t.Fail()
+	}
+
+	if h.ParseString("ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff") != nil {


@david-julien Is it possible to replace these huge hex strings with math.big. I don't like putting big string literal hexes all over the code.

Rather than doing hash.ParseString() to get a hash from a big number, can we use the math.big exponentiation function like z.Exp(2, 50, 0) to get 2^50.

If we need the ParseString function, do func ParseString(s string) Hash rather than having the hash as the receiver.

Was trying to articulate this very thing.

jordanschalm · 2017-06-17T20:00:37Z

blockchain/hash.go

+}
+
+// ParseString sets the hash to the the value represented by a hexadecimal string, and returns the hash
+func (h *Hash) ParseString(s string) *Hash {


I'd like to remove the need for hex strings if possible. See comment below

chadlagore

This is great 🥇 Nothing blocking but I think ParseString could be broken out into helper functions, and the API around CompareTo could be a little more clear. Up to you for changes, otherwise 👍.

chadlagore · 2017-06-17T19:33:08Z

blockchain/hash_test.go

+	"testing"
+)
+
+func TestParseString(t *testing.T) {


👏 Great coverage. But I wonder if ParseString couldn't be broken out into a set of helper functions, with a test for each so that when something breaks we have a clearer idea of what part broke?

chadlagore · 2017-06-17T19:56:00Z

miner/miner.go

+
+// VerifyProofOfWork computes the hash of the MiningHeader and returns true if the result is less than the target
+func (mh *MiningHeader) VerifyProofOfWork() bool {
+	if mh.DoubleHashSum().CompareTo(mh.Target, bc.LessThan) {


Non-blocking, but in terms of convention we should consider the use of bc as a namespace vs a concrete instance of a Blockchain. I've been using it as the later, but I'm game for whatever as long as we're consistent.

Also here the namespacing does not really tell me what LessThan is. I wonder if

hash.GreaterThan(otherHash)

hash.LessThan(otherHash)

hash.EqualTo(otherHash)

is a more clear API?

You could make this a one liner by hitting a little:

return mh.DoubleHashSum().CompareTo(mh.Target, bc.LessThan)

chadlagore · 2017-06-17T21:12:14Z

miner/miner.go

@@ -0,0 +1,94 @@
+package miner


Can we introduce an interface for MiningHeader or maybe Miner? I can imagine a scenario where we want to do an integration test where we mine a block (without waiting).

bfbachmann

🥇

bfbachmann · 2017-06-17T21:10:58Z

miner/miner.go

+
+// VerifyProofOfWork computes the hash of the MiningHeader and returns true if the result is less than the target
+func (mh *MiningHeader) VerifyProofOfWork() bool {
+	if mh.DoubleHashSum().CompareTo(mh.Target, bc.LessThan) {


You could make this a one liner by hitting a little:

return mh.DoubleHashSum().CompareTo(mh.Target, bc.LessThan)

…hpad/cumulus into 25-mining-proof-of-work

coveralls · 2017-06-20T23:10:42Z

Coverage decreased (-3.08%) to 84.516% when pulling 8446f1b on 25-mining-proof-of-work into 17eeffd on transaction-pooling.

…ing-proof-of-work

coveralls · 2017-06-20T23:38:13Z

Coverage decreased (-0.2%) to 84.516% when pulling ef80df5 on 25-mining-proof-of-work into 4148a01 on transaction-pooling.

coveralls · 2017-06-21T01:09:31Z

Coverage increased (+1.2%) to 70.697% when pulling f110208 on 25-mining-proof-of-work into e6bf31e on dev.

David Julien and others added 11 commits June 7, 2017 15:08

Hash helpers and converters

ba7b9f5

minining functions

df7c3e9

Clean up in hash.go

900ca56

Clean up in miner.go

5d395c9

move BaseConverter to test_utils

c7e58ee

Start writing tests for hash.go

2e9a038

start writing tests for miner.go

7cb8898

Change target from type uint32 (compact) to Hash

6f0dd8b

remove time and version from mining header

940659e

remove unused functions from mining implmementation

93a58f2

add time to mining header, add errors in mining header validation, ch…

9401181

…ange nonce to uint64

david-julien requested a review from jordanschalm June 10, 2017 22:57

jordanschalm approved these changes Jun 10, 2017

View reviewed changes

bfbachmann reviewed Jun 11, 2017

View reviewed changes

David Julien and others added 3 commits June 12, 2017 18:49

move miner to its own package

3ff4c9c

HashToHexString

9673713

Merge branch 'dev' into 25-mining-proof-of-work

0666a58

David Julien added 3 commits June 13, 2017 17:11

change hash string representation to 0x hexadecimal form and add erro…

82896d5

…rs and testing

Merge remote-tracking branch 'remotes/origin/transaction-pooling' int…

6397882

…o 25-mining-proof-of-work

change pointers in hash functions to non-pointers, and add mining tests

79614ae

chadlagore reviewed Jun 17, 2017

View reviewed changes

jordanschalm changed the base branch from dev to transaction-pooling June 17, 2017 19:46

jordanschalm approved these changes Jun 17, 2017

View reviewed changes

chadlagore approved these changes Jun 17, 2017

View reviewed changes

chadlagore reviewed Jun 17, 2017

View reviewed changes

bfbachmann approved these changes Jun 17, 2017

View reviewed changes

Merge branch '25-mining-proof-of-work' of https://github.com/ubclaunc…

cf17cfc

…hpad/cumulus into 25-mining-proof-of-work

Cleaned up code according to new review

8446f1b

Merge remote-tracking branch 'origin/transaction-pooling' into 25-min…

ef80df5

…ing-proof-of-work

jordanschalm changed the base branch from transaction-pooling to dev June 21, 2017 01:06

Merge branch 'dev' into 25-mining-proof-of-work

f110208

jordanschalm merged commit a988383 into dev Jun 21, 2017

jordanschalm deleted the 25-mining-proof-of-work branch June 21, 2017 03:11

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

25 mining proof of work #66

25 mining proof of work #66

david-julien commented Jun 10, 2017

jordanschalm left a comment

jordanschalm Jun 10, 2017

jordanschalm Jun 10, 2017

jordanschalm Jun 10, 2017

david-julien Jun 12, 2017 •

edited

Loading

david-julien Jun 13, 2017 •

edited

Loading

jordanschalm Jun 10, 2017

jordanschalm Jun 10, 2017

jordanschalm Jun 10, 2017

jordanschalm Jun 10, 2017

jordanschalm Jun 10, 2017

david-julien Jun 12, 2017

jordanschalm Jun 13, 2017

bfbachmann left a comment

coveralls commented Jun 13, 2017

coveralls commented Jun 13, 2017

coveralls commented Jun 14, 2017

chadlagore Jun 17, 2017

jordanschalm left a comment

jordanschalm Jun 17, 2017

chadlagore Jun 17, 2017

jordanschalm Jun 17, 2017

chadlagore left a comment

chadlagore Jun 17, 2017

chadlagore Jun 17, 2017

chadlagore Jun 17, 2017

bfbachmann Jun 17, 2017

chadlagore Jun 17, 2017 •

edited

Loading

bfbachmann left a comment

bfbachmann Jun 17, 2017

coveralls commented Jun 20, 2017

coveralls commented Jun 20, 2017

coveralls commented Jun 21, 2017

25 mining proof of work #66

25 mining proof of work #66

Conversation

david-julien commented Jun 10, 2017

jordanschalm left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

david-julien Jun 12, 2017 • edited Loading

Choose a reason for hiding this comment

david-julien Jun 13, 2017 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

bfbachmann left a comment

Choose a reason for hiding this comment

coveralls commented Jun 13, 2017

coveralls commented Jun 13, 2017

coveralls commented Jun 14, 2017

Choose a reason for hiding this comment

jordanschalm left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

chadlagore left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

chadlagore Jun 17, 2017 • edited Loading

Choose a reason for hiding this comment

bfbachmann left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

coveralls commented Jun 20, 2017

coveralls commented Jun 20, 2017

coveralls commented Jun 21, 2017

david-julien Jun 12, 2017 •

edited

Loading

david-julien Jun 13, 2017 •

edited

Loading

chadlagore Jun 17, 2017 •

edited

Loading