This record makes public claims about real businesses. Some of them will be wrong. What follows is what actually happens when one is — described as the mechanics it is, not as a promise.
Owners, staff and readers can contest any record with the form below. It goes to the same inbox as reader tips, marked as a dispute, with the record it is about. You can also email press@seabiz.rip.
The most useful report says what is wrong (open when we say closed, the wrong date, the wrong location, or confused with a similarly named business elsewhere) and links to something showing the correct position. A link helps most, for the same reason we require one to publish.
On timing, honestly: this is a small, independently run project. There is no newsroom behind it and no guaranteed response time. Reports are read by hand, as time allows. Claiming a turnaround we cannot keep would be the kind of promise this page exists to avoid. If you leave an email address, that is where any reply will go.
New closures go through the tip form on the home page. How a record gets published in the first place is on Methodology.
When a closure is withdrawn, its page stops asserting the closure and returns a real HTTP 410 Gone with an explanation of what happened. That status matters: it tells search engines and archives the page was deliberately removed, rather than leaving a silent redirect or a page that pretends it never existed.
The record itself is not deleted. It is marked no longer current, and the prior state and the reason are written to an audit table first. Any withdrawn record can be restored exactly as it was.
When a published closure changes, it is marked no longer current and its prior state and reason are written down first — it is not removed. Every corrective operation on the published record writes its own reversal trail before it changes anything:
This is the same discipline as the citation rule, applied backwards: we do not publish a claim we cannot show you the source for, and we do not remove one without keeping the evidence of what changed.
The honest limit: this covers the published record. Behind it, maintenance work has occasionally merged duplicate rows outright — two database entries that were always the same restaurant. Those merges are recorded, but a merged-away row is not recoverable field-for-field the way a withdrawn closure is.
Some limits are structural rather than accidental, and are published rather than hidden:
The fuller version of what this record does and does not claim is on About, and the methodology is on Methodology.
Each entry says what was wrong and what was done, no more. The date on each entry is the one recorded in the audit document it is drawn from.