bump fc and fix node

[oxarbitrage]

While doing some development i bumped FC to the last version and when doing so I noticed that it will not build with error:

/home/alfredo/CLionProjects/lua/libraries/net/node.cpp: In constructor ‘graphene::net::detail::node_impl::node_impl(const string&)’:
/home/alfredo/CLionProjects/lua/libraries/net/node.cpp:320:11: error: ‘rand_pseudo_bytes’ is not a member of ‘fc’
       fc::rand_pseudo_bytes(&_node_id.data[0], (int)_node_id.size());

This is due to changes introduced here: bitshares/bitshares-fc#84

The same compiler provided a hint to fix it after the error msg:

/home/alfredo/CLionProjects/lua/libraries/net/node.cpp:320:11: note: suggested alternative: ‘rand_bytes’
       fc::rand_pseudo_bytes(&_node_id.data[0], (int)_node_id.size());
           ^~~~~~~~~~~~~~~~~
           rand_bytes

This pull request bumps FC and add the fix changing rand_pseudo_bytes to rand_bytes. I am honestly not totally sure about the implications but just know this way it will build.
Appreciate your comments to confirm if this is a good fix or if need to do it different. Alternative can be to bring back rand_pseudo_bytes or revert the FC pull fully(as changes were done only to fix some warnings)

[jmjatlanta]

Yes, that is correct. The FC PR to fix the warnings also included the removal of deprecated methods. The difference between rand_pseudo_bytes and rand_bytes was small, and rand_pseudo_bytes was not recommended. Hence, the deprecation.

Compiled with Boost 1.67 / OpenSSL 1.10 / Ubuntu 18.04

[cogutvalera]

We can merge it IMHO. Thanks !

[abitmore]

Why docker cloud build is failing?

[cogutvalera]

How to see docker cloud's logs ? After clicking on details link I see 404 error

[oxarbitrage]

I think it timed out or something, in the logs i see "Waiting for logs..." and nothing more. The duration of the failed build was 2 hours, generally successful builds take an hour. I restarted now, lets see how it goes.

[cogutvalera]

sometimes travis & docker cloud must be restarted and it will help. Travis I can restart (have permission), but docker cannot, maybe somebody may help me here to give such access too, Thanks !

[oxarbitrage]

Travis is faster and have less issues in general, need to be restarted on each new pull request to populate the cache @pmconrad added to it, only fails when the pull is new.
Docker takes more time and can have issues when to many builds are done in a short period of time, we use it mainly because we have Dockerfile in the project, if not we could just use travis alone.

@cogutvalera you need to have a dockercloud account, then go to https://cloud.docker.com/u/bitshares to see the builds of bitshares-core.
I am not sure if just an account is enough or we have to ask @xeroc to add you to the organization as he was IIRC the creator.

[cogutvalera]

@oxarbitrage Thanks ! I've docker cloud account, but BitShares builds are forbidden for me.

[pmconrad]

It keeps running into "Build canceled" after about 50%. Perhaps compilation of database.cpp eats too much memory.

[pmconrad]

Perhaps try cmake-option -D GRAPHENE_DISABLE_UNITY_BUILD=ON in Dockerfile.

[oxarbitrage]

Perhaps try cmake-option -D GRAPHENE_DISABLE_UNITY_BUILD=ON in Dockerfile.

tried but failed at the same place.

[oxarbitrage]

This is unrelated to this particular code as cogutvalera:issue_1171 is failing with the same error.

I think we should try to fix it outside the context of this pull request to don't start adding several tests and unrelated commits to it.

[cogutvalera]

#1171 (#1382 PR) must be rebased after merging this PR

[jmjatlanta]

UNITY stuff retracted. Approving again to verify that dockercloud does not error out.