Skip to content
Merged
Show file tree
Hide file tree
Changes from 37 commits
Commits
Show all changes
44 commits
Select commit Hold shift + click to select a range
efee42d
feat(model): add the legacy identity recovery plan and additive merge
lklimek Jul 28, 2026
3bab9a9
feat(database): read one legacy identity row through the shared row d…
lklimek Jul 28, 2026
b677776
feat(identity): restore keys stranded in the previous version's saved…
lklimek Jul 28, 2026
c2e859f
feat(masternodes): offer the previous version's stranded keys on the …
lklimek Jul 28, 2026
c274558
feat(identities): offer stranded legacy keys on the Key Info screen
lklimek Jul 28, 2026
e484589
docs(identity): record the legacy-key recovery flow and close the mig…
lklimek Jul 28, 2026
63991b7
fix(model): only offer legacy keys that still correspond to the identity
lklimek Jul 28, 2026
f405098
fix(identity): stop holding the storage guard across the password prompt
lklimek Jul 28, 2026
7caf529
fix(ui): keep the restore offer and the record it wrote in step
lklimek Jul 28, 2026
fc125f4
docs(identity): record what the recovery flow actually shipped
lklimek Jul 28, 2026
24d3081
fix(model): stop the legacy blob from vouching for its own voter key
lklimek Jul 29, 2026
4c42e55
fix(identity): gate the recovery offer on legacy rows, not on data.db…
lklimek Jul 29, 2026
cdde7f1
fix(ui): re-read the key screen on arrival and attribute a finished r…
lklimek Jul 29, 2026
be5299b
test(identity): pin the mid-flight protection guard against regression
lklimek Jul 29, 2026
ad22a2c
fix(model): stop the legacy blob from vouching for its own operator key
lklimek Jul 29, 2026
bff08f7
fix(masternodes): drop the missing-voter message that can never be shown
lklimek Jul 29, 2026
171c512
fix(ui): let the recovery section read as a whole when it can restore…
lklimek Jul 29, 2026
1612fc7
docs(identity): stop promising a voting-key restore that cannot happen
lklimek Jul 29, 2026
2ade1b8
fix(identity): give the identity keys list a way into each key
lklimek Jul 29, 2026
8db5341
test(identity): pin the restore outcome and the key naming on the key…
lklimek Jul 29, 2026
b1e2f28
test(identity): report both restore outcomes on the keys list in the …
lklimek Jul 29, 2026
ef665b1
fix(identity): make the keys list and Key Info agree on where a key i…
lklimek Jul 30, 2026
a237497
fix(identity): return from a key to a current list, in the identity's…
lklimek Jul 30, 2026
cb5b0de
fix(identity): one attribution rule for both hosts of the recovery offer
lklimek Jul 30, 2026
f36b4ca
fix(identity): identify a key by its public half, and name it the sam…
lklimek Jul 30, 2026
aa1234f
fix(identity): find a key's private half where it is filed, not where…
lklimek Jul 30, 2026
2dbc996
fix(identity): Key Info names a user's keys the way the keys list does
lklimek Jul 30, 2026
d39457b
fix(identity): a disabled key the device holds is still reported as held
lklimek Jul 30, 2026
3228b2a
fix(identity): resolve a key's store on demand instead of threading i…
lklimek Jul 30, 2026
91fbc73
fix(masternodes): a key opened from a node's page keeps the name the …
lklimek Jul 30, 2026
66a80a8
refactor(identity): make the key map private so the resolver cannot b…
lklimek Jul 30, 2026
5f82b74
docs(identity): record the key-placement resolution design and the fix
lklimek Jul 30, 2026
4042c01
refactor(identity): one held-key resolution rule for every Manage key…
lklimek Jul 30, 2026
a028d7e
refactor(identity): make an upstream key field a build failure, not a…
lklimek Jul 30, 2026
c95f067
Merge fix/889-key-info-navigation into the key-placement resolver
lklimek Jul 30, 2026
c898a70
test(identity): pin that removing one key spares a different key shar…
lklimek Jul 30, 2026
001b3d5
Merge fix/889-key-info-navigation: port the key-field exhaustiveness …
lklimek Jul 30, 2026
0e41d5d
fix(masternodes): resolve the Add-protection CTA's held key like ever…
lklimek Jul 30, 2026
1a944eb
Merge remote-tracking branch 'origin/v1.0-dev' into fix/889-key-info-…
lklimek Jul 30, 2026
f334875
fix(identity): reach Key Info's re-read, share the restore offer, nam…
lklimek Jul 30, 2026
2c264bb
docs(user-stories): point IDN-008 at the key list that shipped
lklimek Jul 30, 2026
17feeea
Merge remote-tracking branch 'origin/fix/889-key-info-navigation' int…
lklimek Jul 30, 2026
22ee879
fix(identity): route legacy-recovery test fixtures through insert_at/has
lklimek Jul 30, 2026
9c99fe6
Merge remote-tracking branch 'origin/v1.0-dev' into fix/889-key-targe…
lklimek Jul 30, 2026
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
69 changes: 69 additions & 0 deletions CHANGELOG.md

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🟠 MEDIUM · blocking — the changelog now argues with itself, inside a single section.

Lines 87-98 add "A saved voting key can now actually sign" to ## [Unreleased] → ### Fixed. Six lines further down, still in the same ### Fixed section, the pre-existing entry at 125-133 continues to advise:

One known limitation, for a voting key stored on the identity itself rather than on a separate voting identity: … saving or removing one by hand can affect a voting key of the same number on a linked voting identity, and removing it may leave the original in place. So until then, after saving or removing a voting key on an identity like that, open the keys list and check that each key still reads as you expect, and re-enter any key that should be saved but no longer is. This will be closed by the in-progress key-placement resolution fix.

This PR is the in-progress key-placement resolution fix. remove_held_private_key now removes every placement candidates selects, and removing_one_key_leaves_a_different_key_sharing_its_id_alone pins it. get_commit on 5f82b74 confirms CHANGELOG.md gained 13 lines and removed exactly zero — the new entry was appended, the obsolete one was never reconciled.

So the release notes will simultaneously tell a masternode operator the bug is fixed and instruct them to hand-audit every key after every save and remove, and promise a future fix that shipped in the same release. Three mutually incompatible statements, one section. The PR checklist ticks "CHANGELOG.md updated", which is true in the narrowest possible sense.

Fix: delete the "One known limitation …" paragraph (125-133), or reduce it to the residual that genuinely survives. Per your own design.md §8 that residual is the orphaned vault secret on delete — not the wrong-key delete, which this PR actually closed. Whatever you keep, the sentence "This will be closed by the in-progress key-placement resolution fix." must not survive in a section that announces that fix.

Copy link
Copy Markdown
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Confirmed still present at HEAD 001b3d5c. CHANGELOG.md's ### Fixed section still carries both entries: the new "A saved voting key can now actually sign" bullet, and — a few lines later, inside the "An identity's keys are reachable again" bullet — the stale "known limitation ... This will be closed by the in-progress key-placement resolution fix" paragraph, which now self-referentially describes this same PR as still pending. Not reconciled.

Original file line number Diff line number Diff line change
Expand Up @@ -25,6 +25,27 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).

### Added

- **Restore keys an upgrade left behind**: an identity that was already in the
app before the update — a masternode loaded from its ProTxHash, or one that
held only some of its keys — kept its remaining keys in the previous
version's data with no way to reach them. Its page now offers to bring them
across: the node detail page and the Key Info screen list what can be
restored, named by role, and restore only what you press Restore on. Nothing
happens at launch or during the update, keys already saved are never replaced
or removed, and on a password-protected identity the identity password is
asked for first — cancelling or mistyping it leaves everything as it was. A
saved key that no longer matches one this identity uses — a key rotated or
retired since the previous version saved it — is listed with an explanation
instead of being restored, so a key that could not sign can never make the app
report a role as held. The same applies to a key held on a separate voting or
operator identity that this identity does not currently link to: nothing
outside the old data says that key is still in use, so it is listed with its
explanation and entering it by hand stays the way to bring it back. A
masternode loaded from its ProTxHash alone is exactly that case — its owner
and payout keys come back, its voting key is listed as one that cannot be, and
checking that key against the chain instead is tracked as issue #942. The
previous version's data is only ever read, so this is safe to repeat.

- **Automatic Platform node refresh during upgrades**: migrating a pre-1.0
installation now triggers a best-effort Mainnet or Testnet node refresh.
Failed attempts retry on later launches until fresh addresses are saved and
Expand Down Expand Up @@ -63,6 +84,54 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).

### Fixed

- **A saved voting key can now actually sign**: a voting key held on an
identity's own record — rather than on a separate voting identity — was saved
and shown as being on this device, but nothing could use it. Signing looked for
it in the wrong place, so voting with it failed and the key's page reported it
missing, on the screen whose job is to answer that question. Dash Evo Tool now
finds a key by matching it against the key itself, wherever it is filed, so it
is found whichever version of the app saved it and no key material has to be
moved to fix this. This also means a key is no longer confused with a different
key that happens to share its number, which a masternode has whenever its
voting identity numbers a key the same way as its main identity: removing one
key could remove the other's private half, and a key could be reported as
saved on the strength of an unrelated key being present.

- **An identity's keys are reachable again**: the keys list under an identity's
Settings → Advanced now opens each key's own page, so keys can be inspected
and restored — and, once a key is on this device, signed with or
password-protected — without changing the interface mode and without starting
a payment. Previously that list was a read-only table with no way onward, and
every route to a key's page ran through an action screen — sending,
withdrawing, a token operation — each of which offers it only when the
identity already holds a key of the kind that action needs. So an identity
missing its keys, the one case where this matters most, could not get to them
at all. The offer to restore keys left behind by an earlier version now also
appears on the keys list itself, above the keys, rather than only inside a
key's page. Each key is named by its role and states whether it is saved on
this device. Keys are named for the identity they belong to: a user identity's
keys are described in plain language rather than in masternode registration
terms, which previously appeared on every identity. Leaving a key returns to
the list with both its keys and the restore offer brought up to date, so a
restore made from a key's page is reflected immediately instead of being
offered again.

A key opened from a masternode's page keeps its name too. A voting key is the
node's voting key however it is recorded, and its own page now says so instead
of describing it as another kind of key, which also means the page no longer
reports such a key as missing while the list it was opened from shows it as
saved on this device.

One known limitation, for a voting key stored on the identity itself rather
than on a separate voting identity: the keys list and the key's page now agree
on whether such a key is saved here, but saving or removing one by hand can
affect a voting key of the same number on a linked voting identity, and
removing it may leave the original in place. So until then, after saving or
removing a voting key on an identity like that, open the keys list and check
that each key still reads as you expect, and re-enter any key that should be
saved but no longer is. This will be closed by the in-progress key-placement
resolution fix.

- **Wallet rename consistency**: renaming a wallet no longer overwrites other
saved wallet details when metadata cannot be read. Overlapping renames and
wallet removals also keep displayed aliases and deleted-wallet metadata
Expand Down
18 changes: 14 additions & 4 deletions docs/ai-design/2026-07-13-legacy-identity-migration/design.md
Original file line number Diff line number Diff line change
Expand Up @@ -284,10 +284,20 @@ trip the vault-first downgrade guard. Rather than a heuristic that risks
resurrecting removed keys or failing the whole pass, the safe behaviour is to
skip and defer recovery to a dedicated, provenance-aware flow.

The proper recovery flow — an interactive, opt-in re-import that reads the
preserved legacy blob and merges only genuinely-missing key material under the
identity password — is tracked as a follow-up in
[issue #889](https://github.com/dashpay/dash-evo-tool/issues/889).
**Resolved.** The recovery flow this limitation defers to has since shipped:
`docs/ai-design/2026-07-28-legacy-identity-recovery/design.md`
([issue #889](https://github.com/dashpay/dash-evo-tool/issues/889)). It is an
interactive, opt-in, per-identity action reached from the node detail page and
the Key Info screen. It answers the provenance question by asking the only
party who holds it: detection lists what the legacy blob has and the modern
record does not, and the user's approved list is what the merge writes. The
merge is additive from the fresh modern record, so nothing present is replaced
or removed; a protected identity verifies its password first and every restored
key is sealed before the record is written, so the downgrade guard cannot trip.

The importer itself is unchanged — skip-if-present still stands, `data.db` is
still opened read-only and never modified, and nothing about the recovery flow
runs at migration or launch time.

---

Expand Down
Loading
Loading