fix(blockchain): clarify DeclarationInfo::withdraw_at and its behavior - #424
Open
youngjoon-lee wants to merge 1 commit into
Open
fix(blockchain): clarify DeclarationInfo::withdraw_at and its behavior#424youngjoon-lee wants to merge 1 commit into
DeclarationInfo::withdraw_at and its behavior#424youngjoon-lee wants to merge 1 commit into
Conversation
madxor
added a commit
that referenced
this pull request
Aug 27, 2026
…rved epoch Two problems on the withdrawal side of the same seam, both about which epoch a stored field names. The first is the name. withdraw_at reads as "the epoch at which the withdrawal is executed", which is not what it holds: 58f636f deliberately made it record the epoch the withdrawal was requested in, and dropped the guard that the value be higher than the current epoch. The name has been pulling readers back toward the old meaning ever since, most recently in #424, which proposes setting it to the current epoch plus two. Rename it to withdrawn, in the past tense, which cannot be read as a future execution moment. That also lines the three lifecycle fields up with the three event types the Indexing section already defines: created, active, withdrawn. created and withdrawn record when a message landed, active records what a message claimed. Storing the requested epoch rather than the removal epoch is the right side of that choice for the same reason active stores the attested epoch: it is the only value observable when the field is written, and it does not bake the finalization delay into ledger state, where a later change to that delay would falsify every stored value. The second is the service window, which #424 is right about. A withdrawal included in epoch e reaches services only through the snapshot for epoch e+2, so the declaration is still listed for e+1 and the node serves it. The previous text claimed the node "provides the service through the withdrawal epoch e and stops afterwards", which the snapshot rule contradicts. Correcting that exposes an epoch the node serves but can never be paid for. The report attesting to e+1 is submitted during e+2, and the declaration was removed at the start of e+2. So the node sat in the active set for an epoch it had no reason to serve, which for the Blend Network means peers connecting to it and counting it toward the minimum network size. Move removal to e+3: the e+1 report lands during e+2, its reward is paid in the first block of e+3, and removal follows in that same block, exactly as the existing rule already promises when it says a declaration is never removed before its final reward is paid. The stake stays locked one epoch longer, across every epoch the node was in service. Mantle carries the removal predicate and the write, so both move with it: withdrawn <= current_epoch - 3, and declare_info.withdrawn = current_epoch. #424 changed the field's meaning in the SDP document alone, which would have left the Mantle predicate firing at e+4 while the prose still said e+2. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
13 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The definition of
DeclarationInfo::withdraw_atis written correctly:But, the description in the
Withdrawsection does not match with the definition. So, I polished some sentences in the section.In short,
withdraw_atshould be set withe + 2, whereeis the epoch of block containingWithdrawMessage.I remember that we renamed
withdrawn_atto ‘withdraw_at' for clarity.