BCHD is written in Google Go Language, which is a pretty impressive but not overly verbose.

Getting txdata remains a chore, resembles pulling teeth more than filling a bucket.

Okay - so I finally got TXIDs streaming out of my beautiful BCHD node

And then in deepest despair I begin to realize it religiously abandons almost all the op_returns from Memo

I am not sure yet if BCHD simply rejects them
or if it just never sees them (network?)
but this is tragic
BCHd has everything I wanted in a node: IPFS + GO and more.
but no Memo data.

BCHD --fastsync option makes reloading a borked instance easy.
However, bchd crashes at 2gb limit on IPFS fastsync.
fix = "pin add" the utxo data so it loads internally instead of over the wire.

Update on the IFPS side of things:
after repeated failures of the downloaded UTXO
disabling the IPFS and trying BCHD --fastsync again seems to run smooth.
perhaps I misconfigured the IPFS. yuk

Problems with OP_RETURN are no longer a mystery.
The original BTCD was written in a savagely rabid "THOU SHALT NOT USE OP_RETURN EVER !#!" manner.
Lots of logic flaws exist but these are easy.

I see a new BCHD peer
104.248.209.144 (/bchd:0.13.0(EB32.0)/)
seems to have go online and is rejecting OP_RETURN transactions
I have fixed it in the code I am using and am almost ready to share

I see Chris just released 0.14.0 a few hours ago (Thanks Chris)
so I guess I will be spending time
seeing if it still discards OP_RETURN.
Even if it does
I will simply recode it anyway.

https://github.com/gcash/bchd/releases/tag/v0.14.0

Hopefully helpful hint deartment:
If you try to run both IPFS and BCHD
on an under-powered server ,
start IPFS first
and then start BCHD
so cpu and mem usage stays sane.

Sadly(®) BCHD 0.14.0 still rejects MEMO protocol transactions.
I will revisit the process of recoding the source to avoid this flaw.

How to hack BCHD v0.14.0 to accept OP_RETURN transactions:
[email protected]:~/Progs$ ipfs add BCHD+OP_RETURN.hack.tar.gz
added QmPF3P7NY2Dabkdn9i3BS49eaiYNaKqMc31ajzpdsNRNwW BCHD+OP_RETURN.hack.tar.gz

2019-04-18 19:19:14.127 [INF] SYNC: New valid peer 139.162.198.175:41186 (inbound) (/bcoinsv:1.1.0/)

Got some videos for ya's.

&t=

Also, BCHD 0.14.2 is out! https://github.com/gcash/bchd/releases/tag/v0.14.2

and v0.14.2 does accept OP_RETURN transactions.
This means it is immediately useful for MEMO protocol operations,
Thanks to the BCHD team

Just a note on what spooling up a full v0.14.2 node entails: 2cores 8gb

https://github.com/gcash/bchd/releases/tag/v0.14.3
Upgrade from 0.14.2 (mid-SYNC) was painless.
Now awaiting the 100% SYNCed moment
v0.14.3 has features to investigate once fully SYNCed.

2019-05-06 12:18:04.370 [INF] SYNC: New valid peer 196.249.98.94:56152 (inbound) (neutrino:0.0.4-beta/)

You can test nodes here:
(require.patience): https://cash.coin.dance/nodes#nodeChecker
2019-05-06 12:24:06.346 [INF] SYNC: New valid peer 34.218.174.49:44678 (inbound) (/bitnodes.coin.dance:0.1/)

SIGSEGV crash appears unrecoverable on BCHD,
Lacks any CLI methods to roll back and reSYNC
Restart SYNC taking weeks puts this out of reach on v0.14.3 at this time.
Parked, awaiting next release.

Which blockchain is this ?!

Which blockchain =812484 blocks ?!
2019-05-30 03:50:27.745 [INF] SYNC: Processed 2 blocks in 10.32s (4496 transactions, height 422369/812484 (51.98%), 2016-07-26 17:46:53 +0000 UTC, ~105 MiB cache)