Data Recovery Case File · Trust, Practice & Honest Limits · Inherited and Irreplaceable

Excavation Is Destructive, So the Record Is the Evidence

This enquiry concerns data nobody in the building created. A home-assembled computer "transferred to us when an archaeological unit ceased trading. We believe it contains some important records of some excavations. We are looking for the data to be recovered and made available to us, as the unit appears to be damaged or corrupted." Unknown hardware, unknown configuration, and contents nobody can describe precisely. And a kind of data that is irreplaceable in a stricter sense than almost anything else in this archive — because the sites those records describe no longer exist in the state they recorded.

The decode: why this is two problems, and the one people forget

Why this data is irreplaceable in an unusual way: excavation destroys what it examines. A site is dug once, and the act of digging removes the stratigraphy that was the evidence. The record made at the time — context sheets, plans, levels, finds registers, photographs — is the site, permanently and for everybody afterwards. There is no re-examining the original. Most irreplaceable data in this archive is personal; this is irreplaceable in the sense that no one can produce it again by any means.

Why the machine's history is the first practical problem: a self-assembled computer of unknown age may hold drives of any vintage on any interface, may have been stored badly, and has been powered by people trying to see what is on it. The first step is not to keep switching it on but to remove the storage, identify what it is, and capture it — because an old drive that still works is a drive with a limited number of start-ups left, and nobody knows how many have been used.

Now the part that gets forgotten, and it is the reason a recovery here can succeed and still fail. Getting the bytes off is half the job. A recovered file in an application format nobody can open is not yet a recovered record. Units of that era used a mixture of general software and specialist packages — site databases, planning and survey applications, CAD formats — some of which no longer exist, some of which changed format across versions, and some of which wrote proprietary files readable by nothing else.

So the deliverable has to include an inventory: what formats are present, from roughly what era, and what reads them now. That turns a folder of unopenable files into a known problem with known routes — a modern package that imports the old format, a conversion tool, or a specialist who still runs the original software. Asking for formats to be identified and reported alongside the data costs nothing at recovery time and is very expensive to work out later.

What is worth doing beyond the recovery: material of this kind usually has a proper home — a regional historic environment record, a national archaeological archive, or the relevant local authority. Placing it there is what actually preserves it, and archives have deposit standards worth knowing before the files are reorganised.

The obligation worth checking: records from a dissolved unit may carry conditions from whoever commissioned the work, and establishing who holds rights is worth doing early.

On the bench

The machine was not powered repeatedly, an unknown-age drive having a limited number of start-ups remaining and nobody knowing how many had been used. Storage was removed and identified before any host attempt, then imaged write-blocked under per-sector timeouts. Recovered material was inventoried by file format and era, and formats identified and reported alongside the data — since a recovered file in an application format nobody can open is not yet a recovered record, and working that out later is far more expensive than at recovery time.

The outcome

The storage captured before further powering, and the recovered material delivered with a format inventory alongside it. Free assessment, one fixed written figure including VAT; where a drive has to be opened, 50% of parts and labour is payable upfront with the balance only on success — otherwise no recovery, no fee. The decode, for anyone recovering inherited institutional records: getting the bytes off is half the job. Specialist software of that era wrote proprietary formats, some of which no longer exist — so ask for the formats to be identified and reported with the data. And excavation records are irreplaceable in the strictest sense, because the site was consumed in the making of them.

Our case files are drawn from genuine enquiries received by our laboratory over the past ten years, anonymised to protect client confidentiality. Each one describes the diagnostic and recovery procedure our engineers apply to that fault, using the equipment listed.