Reviews Inquire

Confirm your archive with reliable checksums

An archived file can open cleanly and still differ from the version that was originally published. A checksum provides a compact fingerprint of the file’s bytes, allowing you to detect whether a copy has been altered, truncated or damaged in transit. The familiar SHA-256 method produces a 64-character hexadecimal digest. Compare that newly calculated value with a trusted reference: an exact match confirms that both files share the same content, while a mismatch gives you a reason to investigate.

Checksums are especially useful for historical web material, where a saved page may sit beside a database export, an image set and several revised drafts. They provide a repeatable answer to a simple question: has this particular artefact changed? Their value lies in the evidence they preserve. A dated hash can later demonstrate that the file examined during a research session matched the copy recorded by a library, publisher or volunteer steward.

Verification does not automatically establish who created a file, whether its claims are true or whether it is safe to open. It confirms byte-for-byte consistency with a supplied reference value, and that reference must come from a reliable source. The hash also describes one file sequence, not its filename, location or surrounding archive. A page saved with browser-added scripts can therefore look identical while producing a different digest from the original capture.

Australian research often crosses several preservation systems. Records surfaced through Trove, the catalogue operated by the National Library of Australia, may point towards material held by the Australian Web Archive or a local library collection. Each hop can involve a different file package and a different record-keeping practice. A checksum workflow helps you keep those layers visible, so a preserved title remains traceable even when its bits, formats or storage locations change over time.

Understand what a checksum proves

A SHA-256 value is calculated from the file’s exact sequence of bytes. Moving a document to another folder or changing its name normally leaves the value unchanged. Adding metadata, rewriting line endings or saving the same document through another application can change it. The checksum identifies a specific digital artefact rather than its general subject or visual appearance.

The test only means something when the expected value is anchored in trusted material. A checksum copied from an unattributed forum post may be outdated or attached to the wrong file. Independent records, repository manifests or confirmation from the organisation that produced the file provide a stronger basis for comparison.

Before treating a download as part of the collection, use the domain resource hub to understand the site’s background and locate its preserved sections. The hub can supply context; the authoritative checksum should come from a signed manifest, a trusted depositor or another independent record. Keeping those sources separate makes the final verification easier to explain to another researcher.

Prepare a reproducible verification workspace

Start with a working copy rather than the only surviving original. Preserve the directory structure and filenames used by the archive, because a manifest may identify objects by relative path. Record where the reference checksum came from, its publication date and the algorithm used. Check that the download finished completely before calculating anything; a valid-looking filename offers no protection against a short transfer.

For archived HTML, compare the artefact actually captured by the archive with the file supplied for checking. Browser “Save page” tools may inject scripts or rewrite markup, changing the bytes even when the text looks the same. For compound packages, identify whether the checksum applies to an individual object or to a ZIP container. These small decisions prevent a correct hash from being attached to the wrong material.

Audio and moving-image material deserves the same care. The National Film and Sound Archive in Canberra provides a useful Australian reference point for preservation workflows that handle substantial media files. Confirm the checksum of a preservation master before transcoding, editing metadata or moving it to a new collection-management system. A digest measured after those changes answers a different question from the one an archivist originally intended.

Calculate and compare hashes across systems

Use the same algorithm on every system involved. On most Linux systems, run this in the directory containing the file:

sha256sum document.pdf

On macOS, use:

shasum -a 256 document.pdf

In Windows PowerShell, use:

Get-FileHash -Algorithm SHA256 .\document.pdf

Read the 64 hexadecimal characters printed by the command and compare them with the trusted reference. Uppercase and lowercase letters represent the same digest, so a case difference is harmless. Spaces in a copied text file can still cause a false mismatch. Hash the exact file in the archive rather than a shortcut that points to another location, and record the command, algorithm, time and result together.

Community readers arriving through archived reporting material can apply the same procedure to any files linked from those pages. Checksums are not encryption, and publishing a digest does not make a suspicious file safe. Avoid uploading confidential material to an unknown online hash checker simply because it is convenient. For this domain’s historical material, a local calculation keeps the file under your control while producing a value that another person can reproduce.

For collections with many objects, store each result in a plain-text manifest instead of comparing files by eye. Automating the command across a directory creates a consistent baseline for later checks. It also leaves a human-readable trail that can be reviewed when the archive is migrated or someone asks how the integrity of a download was established.

Diagnose mismatches and record the evidence

A different hash is a signal, not a diagnosis. Common causes include an incomplete download, the wrong release, a transcoded derivative, text-mode line-ending conversion, or a checksum calculated for another object. Renaming a file should not affect it, and moving it should not either. Check those two facts first, then compare file sizes and the capture metadata.

A trustworthy comparison record should contain:

If the values differ, re-download from the original source where possible and run the calculation again on a clean copy. A mismatch report should record:

Do not overwrite a failed result with a later pass. Keep both entries and explain what changed between them. If a new download matches, the archived original may still need attention; if it does not, seek a second copy or a refreshed manifest from the custodian. A signed statement can help establish provenance, but the digest remains the direct evidence of byte-for-byte agreement.

Preserve manifests and local provenance

A manifest turns individual hash checks into a durable record. Use a simple format such as filename, SHA-256 value, file size in bytes and capture timestamp, then keep it with the associated archive. Record timestamps in UTC, with the Australian local offset added if a human-readable time is needed. Melbourne, Sydney and Hobart move between AEST and AEDT during daylight-saving changes, whereas Brisbane remains on AEST, so a local timestamp can otherwise be ambiguous.

Keep an additional copy outside the main working folder. The Australian Cyber Security Centre’s guidance on resilient backups reinforces this separation, particularly for collections held on shared drives. A manifest copied onto the same disk as the files cannot provide independent evidence if that disk fails. Version control can help track revisions to the manifest, while periodic restoration tests confirm that both the files and their reference values remain usable.

Domain stewardship adds another Australian layer. The .com.au and .au namespaces operate under auDA policies, and eligibility or presence requirements can affect renewals and transfers. Anyone assessing a related domain name should check the current rules instead of assuming that an expired listing can be claimed. This matters for an archival project because continuity of the name can help visitors find the records and contact the people responsible for them.

Dated entries in the preserved news pages can help connect individual artefacts with earlier announcements and community submissions. Add a source column to the manifest so that each digest points back to the record that supplied it. Where possible, link to a stable capture rather than a page that may move again.

Keep verification as an ongoing practice

Treat checksums as part of routine stewardship rather than a one-off rescue. Recheck a sample after archive migrations, storage updates and changes to capture software, then run a broader comparison on a schedule that suits the collection. In an Australian archive team, a quick “hash check” is often the fastest way to settle whether two copies are identical when visual comparison is inconclusive.

Make the reference value available to the people who will maintain the material after you. Publish a manifest, explain the algorithm in plain English and keep a signed copy in more than one location. If new files arrive, calculate their digests before joining them to the existing set; mixing unverified objects into a verified collection weakens the meaning of the whole manifest.

Choose one archived file today, calculate its SHA-256 digest and save a dated verification record. Keep a second copy of that manifest, and contact the domain maintainers through the resource hub when the original checksum or file history cannot be located. One careful check creates a small, solid piece of evidence that future researchers can trust and reproduce.