rebased

Adds PersistMempoolSet and StopAfterBlockImport flags to refactor

In D1525#31277, @schancel wrote:

The idea might be to start moving away from the GetArg() stuff all over the codebase, and instead validate and set proper config variables at launch.

Closed in favor of the work started at https://reviews.bitcoinabc.org/D1525

@deadalnix @jasonbcox adding config param to many method's signatures doesn't feel right. If we include gArgs + ChainParams + ChainBaseParams to config it will lose meaning, you will end up not understanding what the method needs.

Move class def to top.

In D1470#30742, @schancel wrote:

We had discussed out of band that you were going to fix the "Readme.md" file to be a proper MD file. I am requesting those changes here.

Update format of Readme.md to match MarkDown

update doc $USER variable

In D1470#29830, @deadalnix wrote:

I don't think it makes a lot of sens to have this under source control.

Add osx builds asserts + sig

Please armor key (send the .asc) file. Thanks for your comment on the doc, I have updated it.

More LIBTOOL_APP_LDFLAGS config to a more appropiate place

rebased

use C++ initializer

comments and formatting nits

updated std::set. Thanks for the review.

Better std:set for rpc help

update comments

refactor CRPCCommand

As background, I was able to reproduce the reported problem with:

class C {
    public:
        int a;
};

revert abc.cpp

LGTM

ACK

Should we set VISIBILITY_INLINES_HIDDENalso?

Nice! ConceptACK. Initial code review looks fine.

Nice refactor with 'buffer_append', much clear now.

well done Jason!

left some minor comments.

fix large test, leave the test failing waiting for the activation code to be completed

ACK!

Left some questions, thanks!