However, this is just my opinion. It seems Amaury shared it. I hope we don't prevent ourself from making needed organizational changes indefinitely because of backports.

In D3013#71320, @schancel wrote:

I think more and more stuff should be better organized. Core doesn't seem to care about separating stuff out much.

However, this is just my opinion. It seems Amaury shared it. I hope we don't prevent ourself from making needed organizational changes indefinitely because of backports.

Ah I see. Well the original organization was definitely crap, as both Core and Amaury recognized. I guess my mental picture here is that if the backport of PR10095 had already been done, then the tojson.h file would never have been made in the first place and a divergence would not have occurred. I also personally think the blockchain.h is the right place for these functions (being defined in blockchain.cpp).

Okay, I see. That's weird -- still not sure *toJSON functions should go there, but okay.

schancel accepted this revision.May 13 2019, 18:45

This revision is now accepted and ready to land.May 13 2019, 18:45

In D3013#71477, @schancel wrote:

Okay, I see. That's weird -- still not sure *toJSON functions should go there, but okay.

Yeah, as I was typing that I was thinking "why are they in blockchain.cpp anyway?".. Well, once these functions are caught up we can look at messing around with them.

Closed by commit rABC4967c68c3891: [rpc] Move tojson.h into blockchain.h (authored by Mark Lundeberg <36528214+markblundeberg@users.noreply.github.com>). · Explain WhyMay 13 2019, 19:41

This revision was automatically updated to reflect the committed changes.

nakihito mentioned this in D3293: Merge #13163: Make it clear which functions that are intended to be translation unit local.Jun 12 2019, 23:30

Mark Lundeberg <36528214+markblundeberg@users.noreply.github.com> added a commit: rSTAGING4967c68c3891: [rpc] Move tojson.h into blockchain.h.Sep 18 2019, 23:30