test: correct blockencodings_tests

[barrystyle]

This PR corrects the unit tests for blockencodings_tests.cpp

Running this test as it was previously, resulted in an error message that differed per run (!), indicating a bad memory access. Once the hashing function was correctly set; this no longer occurred.

This test can be launched by running (from src/test):

./test_digibyte --run_test=blockencodings_tests

[barrystyle]

This test's outcome does seem a bit far fetched, enclosed is the output (pre) correction:

root@capnmagneto:~/digibyte/src/test# ./test_digibyte --run_test=blockencodings_tests
Running 7 test cases...
test/blockencodings_tests.cpp(214): error: in "blockencodings_tests/NonCoinbasePreforwardRTTest": check partialBlock.FillBlock(block3, {block.vtx[0]}) == READ_STATUS_OK has failed

*** 1 failure is detected in the test module "DigiByte Test Suite"
root@capnmagneto:~/digibyte/src/test# ./test_digibyte --run_test=blockencodings_tests
Running 7 test cases...
test/blockencodings_tests.cpp(109): error: in "blockencodings_tests/SimpleRoundTripTest": check partialBlock.FillBlock(block3, {block.vtx[1]}) == READ_STATUS_OK has failed
test/blockencodings_tests.cpp(214): error: in "blockencodings_tests/NonCoinbasePreforwardRTTest": check partialBlock.FillBlock(block3, {block.vtx[0]}) == READ_STATUS_OK has failed

*** 2 failures are detected in the test module "DigiByte Test Suite"

[gto90]

utACK

[SmartArray]

@barrystyle This PR is a great example of how we might proceed.
I will give an utACK too.

The current state of the source code won't let me compile on Mac OS. I have had experienced this before.

validation.cpp:2493:94: error: use of undeclared identifier '_1'

It seems like Bitcoin-Core fixed this long ago.

So I ask myself, as this looks like a deadlock, how can be proceed now? Fixing the tests is an option, but I will be unable to test the changes.

Should we first proceed to merge current code with bitcoin upstream?
@gto90 What would be the preferred way of doing that? One single PR to merge upstream might be a bit difficult to review? I don't we have another chance, though.

[barrystyle]

@barrystyle This PR is a great example of how we might proceed.
I will give an utACK too.

The current state of the source code won't let me compile on Mac OS. I have had experienced this before.

validation.cpp:2493:94: error: use of undeclared identifier '_1'

It seems like Bitcoin-Core fixed this long ago.

So I ask myself, as this looks like a deadlock, how can be proceed now? Fixing the tests is an option, but I will be unable to test the changes.

Should we first proceed to merge current code with bitcoin upstream?
@gto90 What would be the preferred way of doing that? One single PR to merge upstream might be a bit difficult to review? I don't we have another chance, though.

Please compile again, building and linking against the dependencies generated in the depends folder; the error you're running into is related to boost.

[gto90]

@barrystyle This PR is a great example of how we might proceed.
I will give an utACK too.

The current state of the source code won't let me compile on Mac OS. I have had experienced this before.

validation.cpp:2493:94: error: use of undeclared identifier '_1'

It seems like Bitcoin-Core fixed this long ago.

So I ask myself, as this looks like a deadlock, how can be proceed now? Fixing the tests is an option, but I will be unable to test the changes.

Should we first proceed to merge current code with bitcoin upstream?
@gto90 What would be the preferred way of doing that? One single PR to merge upstream might be a bit difficult to review? I don't we have another chance, though.

@SmartArray ,

As part of our improved peer review process, we would like to make each PR as atomic as is possible in order to ensure a more focused and efficient review process. When the PRs grow beyond the scope of a single topic, it risks reviewers getting overwhelmed and can lead to rubber stamping approvals.

Let's see if we can compile and test using @barrystyle 's suggestion above.

As always, thanks so much for everything you all do!

[SmartArray]

Tests compiled and working now.

@barrystyle Thanks for pointing me into the right direction. I had to add boost header file in order to make it compile. See this commit: 0637d9b

[SmartArray]

ACK.

[digicontributer]

ACK

I was able to build against depends as barry suggested.

Cheers