Your register is a fiction

The comment register in your project folder is a historical document. It records what was true when someone last updated it, which is not the same as what is true now. Ask the project manager and the lead engineer, separately, whether it is current. Both will pause. Both will say something like "probably" and then qualify it. That qualification is the real answer. Most spreadsheet-based registers exist primarily because auditors expect them and clients find them reassuring. Their function is to demonstrate that the office has a process. It does not function to tell you where a comment actually stands.

The mechanism of drift is not complicated. A class surveyor at ClassNK or Bureau Veritas issues a comment on a structural drawing. The comment is logged. A response is drafted, a revision is issued, and a disposition is entered in the register. Then the surveyor comes back: the response is partially accepted, a further revision is requested, the scope of the original comment narrows. This exchange happens by email. The register is not updated, because updating the register is a separate task from answering the surveyor, and the person answering the surveyor is not thinking about the register. By the third exchange the register entry is a relic. The live status of the comment is in an email thread that has acquired a subject line no one chose and which contains the names of three people who are no longer on the project.

This is not a discipline failure. The register is updated at intervals because updating it continuously is not how work gets done. Comment exchanges with DNV or Lloyd's Register happen at the pace of the review cycle. They involve drawings, phone calls, sometimes sketches attached as jpegs. The resolution of a comment is a negotiation, not an event. It does not have a clean timestamp. The register demands a clean entry — status, disposition, revision reference — and so it receives clean entries describing an earlier and simpler version of what actually happened. The gap between the formal record and the operational reality is not sloppiness. It is structural. The architecture of the register assumes that comments resolve discretely. They do not.

Email is the actual record. This is worth saying plainly because most offices treat it as an embarrassing workaround rather than a fact about how approval correspondence works. The thread is where dispositions are actually recorded, where the surveyor's exact language appears, where the date of a verbal agreement is approximated in a follow-up message that says "as discussed." When a correspondence dispute arises — over whether a point was closed, over which revision governs, over what the agreed scope of a re-submission is — the parties go to email. Not to the register. The register is consulted for the project file, for handover documents, for the client's monthly report. Email is consulted when someone needs to know what was actually said.

Revision drift is the same problem in a different column. Every project with more than six months of approval correspondence has drawings at different revisions in different inboxes. The yard has Rev D. The owner's technical team has Rev B, issued before the last two comment cycles. The register references Rev C, current when the relevant comment was entered. These are not filing errors. They are the result of issuing revised drawings to multiple recipients at different points in a review cycle, where not every party receives every revision. Nobody updated the register entry. Nobody told the owner's team that Rev B was superseded. The drawings exist in three states and the register does not resolve which governs. It was not designed to. It was designed to list comments.

The correspondence dispute that follows — the claim that a comment was closed on Rev C, the counterclaim that the surveyor's acceptance was conditional on a change in Rev D, the question of whether the owner formally accepted the arrangement — cannot be resolved from the register. It can be resolved, if at all, from the email thread, provided the right people are reachable and the right emails have not been deleted. That is a fragile basis for dispute resolution on a newbuild. The fragility is not incidental. It follows from the formal record and the working record being different documents, maintained in different places, by different people, for different purposes. The register is a fiction the office agrees to maintain. The correspondence is the truth nobody has filed.