wallet rpc: fix stale argument metadata and help text #36264

MrHodlX commented at 3:15 PM on September 15, 2026: none

Small fixes to stale argument metadata and help text in the wallet RPCs, in three focused commits:

5fcbbf8 — listtransactions: the "Negative from" error references the skip argument by its old name
9aa8977 — listsinceblock: name the failing argument when target_confirmations < 1 instead of a bare "Invalid parameter"
f214b2c — sendmany: the help examples pass values in the ignored minconf position; describe what the example actually sets and pass null for the placeholder

Required test updates ride in each commit. (Two earlier commits touching deprecated arguments were dropped per maflcko's review: either that deprecated argument gets a proper removal cycle later, or it isn't worth touching.)

DrahtBot commented at 3:16 PM on September 15, 2026: contributor

The following sections might be updated with supplementary metadata relevant to reviewers and maintainers.

Code Coverage & Benchmarks

For details see: https://corecheck.dev/bitcoin/bitcoin/pulls/36264.

Reviews

See the guideline and AI policy for information on the review process. A summary of reviews will appear here.

l0rinc commented at 11:26 PM on September 15, 2026: contributor

Could you please split the related fixes into commits, explaining them briefly in the commit message, describing the problem at a high level and explaining the chosen solution. The changes seem unrelated; by grouping them into focused commits, we can selectively review the suggestion instead of accepting or rejecting all of it.

wallet rpc: name the skip argument in the listtransactions error

The listtransactions error for a negative skip said "Negative from",
referencing the argument by its old name. The argument has been called
skip in the help for years; say "Negative skip".")

5fcbbf8352

wallet rpc: name the failing argument in listsinceblock

listsinceblock rejected target_confirmations < 1 with a bare
"Invalid parameter", not saying which argument was wrong. Name it,
matching the neighboring "Invalid start_height" error.")

9aa8977c02

wallet rpc: fix the misleading sendmany example

The sendmany example claimed to set "the confirmation" by passing a
value in the minconf position, but minconf is an ignored dummy, so the
example taught a no-op. Describe what the example actually sets (the
comment) and pass null for the ignored placeholder in both examples.")

f214b2c9fe

MrHodlX force-pushed on Sep 16, 2026

in src/wallet/rpc/coins.cpp:175 in 3a2a982d4e

 171 | @@ -172,7 +172,7 @@ RPCMethod getbalance()
 172 |                      {"dummy", RPCArg::Type::STR, RPCArg::Optional::OMITTED, "Remains for backward compatibility. Must be excluded or set to \"*\".",
 173 |                          RPCArgOptions{.placeholder = true}},
 174 |                      {"minconf", RPCArg::Type::NUM, RPCArg::Default{0}, "Only include transactions confirmed at least this many times."},
 175 | -                    {"include_watchonly", RPCArg::Type::BOOL, RPCArg::Default{false}, "No longer used"},
 176 | +                    {"include_watchonly", RPCArg::Type::BOOL, RPCArg::Default{false}, "(DEPRECATED) No longer used"},

maflcko commented at 5:34 AM on September 16, 2026:

Seems fine, but I wonder if this should be deprecated/removed for real.

Maintaining (and bike shedding) the deprecated stuff seems confusing and not too helpful.

Not sure if this requires an -deprecatedrpc cycle, but I guess it can't hurt.

Also, would probably require updating the type to OBJ_NAMED_PARAMS for safety.

bitcoin deleted a comment on Sep 16, 2026

MrHodlX commented at 12:10 PM on September 16, 2026: none

Fair point — this commit only makes the marker consistent while the argument exists. A real deprecation/removal cycle (with the OBJ_NAMED_PARAMS type update) can be a follow-up PR rather than expanding the scope here.

maflcko commented at 1:40 PM on September 16, 2026: member

idk, you are adding two commits to modify ~dead code. Either it matters and it should be removed/deprecated properly, or it doesn't matter and the commit can just be dropped.

MrHodlX force-pushed on Sep 16, 2026

DrahtBot added the label CI failed on Sep 18, 2026

DrahtBot removed the label CI failed on Sep 18, 2026