Recover not scanning enough internal addresses?

[ZmnSCPxj]

So I wanted to try out the fidelity bonds thing, and I had a long-running maker and decided to upgrade.

According to the docs, the procedure is to use the wallet-tool.py recover and give the original seed words and instead this time indicate to use a fidelity bond wallet. So I moved my existing wallet.jmdat to wallet_backup.jmdat and did a recover.

When I first did wallet-tool.py wallet.jmdat it told me to trigger a rescan on bitcoin-cli, which I did. While it was rescanning, doing a wallet-tool.py wallet_backup.jmdat showed the expected amount, while doing a wallet-tool.py wallet.jmdat gave only a tiny amount that I had recently moved into the wallet.

Unfortunately for me, once the bitcoind rescan had completed, wallet-tool.py wallet.jmdat did not show all my expected funds. Worse, wallet-tool.py wallet_backup.jmdat had somehow also lost funds. Both have the same seed words but diverge on the amount they report.

I have done wallet-tool.py wallet.jmdat -g 500 but funds are still missing.

By my memory, the internal addresses had reached up to something like m/84'/0'/0'/1/300 or thereabouts. Setting the gap limit to 500 or so seems to work only on the external addresses, not the internal addresses. Is there a way to force a proper rescan recovery with the gap limit applying to the internal addresses? Is there a convenient way for me to get the address derivation for the internal addresses?

[ZmnSCPxj]

Hmm, wallet-tool.py wallet.jmdat displayall -g 500 seems to have gotten everything. Subsequent wallet-tool.py wallet.jmdat seems to have gotten all the funds recovered again.

This could probably be better-documented....

Closing.

[AdamISZ]

@ZmnSCPxj please note this repo is not the active Joinmarket one (hasn't been for several years). I will archive it now so people no longer make the mistake.

Please do feel free to open issues at https://github.com/Joinmarket-Org/joinmarket-clientserver .