Original analysis and research design · The current ZIP matches the measured reference digest. No peer review or independent replication is claimed.
A record that can be corrected.
This register connects each question to a method, sufficient inputs, alternatives, results, and a next check. It covers completed archive observations and synthetic controls, followed by proposed investigations. A proposal is a research design; it is not a participant outcome or an access approval.
Independent developers, students, historians, and preservation volunteers can use the public measurements to challenge these accounts. A correction or a well-documented unsuccessful attempt can be as useful as a replication. There is no claim of institutional partnership, independent replication, peer review, or a reviewed access program.
Download the project register · JSON. The observations refer to the identified reference, not to every copy with the same catalogue label.
COMPLETED WORK
These entries report measurements or documentary comparisons already performed, with the method’s limits retained.
R01 · completed observation
Composition of a dated installation
What does this reference contain beyond its main executable?
- Method
- Enumerate the ZIP directory and aggregate file counts and uncompressed sizes by extension; distinguish directory entries from files.
- Inputs
- ZIP directory metadata; executable size; the whole ZIP is read separately for its container digest.
- Alternatives considered
- Published CSV is enough to inspect the reported totals; a local inventory can reproduce them without transferring protected payloads.
- Finding / current status
- 15,476 files and 113 directory entries; 15,870,202,692 uncompressed bytes. The main executable is 14,858,360 bytes, about 0.094 percent of that total.
- Limits
- Extensions were not validated as formats. Counts and sizes do not establish cultural importance, runtime dependencies, or a complete service.
- Next test
- Compare a separately acquired manifest and identify disagreements before drawing a version-history conclusion.
R02 · completed observation
Literal interface composition
Which explicit include relationships connect five fixed default-interface roots to companion files?
- Method
- Resolve XML Include text and HTML include-body file attributes in two declared directory scopes. Record exact and case-only matches, hashes, parse limits, and graph availability masks.
- Inputs
- Five roots, matched include targets, and a reliable inventory. The tool hashes the ZIP in a separate pass to identify its input.
- Alternatives considered
- The published graph supports review of results. A targeted local subset can repeat these relationships; the entire client is unnecessary for the measured graph.
- Finding / current status
- 180 reached files and 175 include occurrences; 174 exact matches and one case-only match. InventoryWindow.html reaches six files totaling 159,775 bytes.
- Limits
- Literal includes do not establish load order, rendering, complete runtime dependencies, or gameplay behavior. No parse or traversal limit was reached in this run.
- Next test
- Independently repeat the graph with a lawfully held copy and report any disagreement, including case-resolution assumptions.
R03 · completed documentary comparison
Version chronology and provenance boundaries
How does the presence of selected interface filenames compare with dated publisher announcements?
- Method
- Compare case-insensitive selected-basename observations with official September 2025, October 2025, and January 2026 update records; distinguish embedded build strings from publication dates.
- Inputs
- Selected-file inventory and the dated publisher records linked in the paper.
- Alternatives considered
- Readers can evaluate the comparison using the published observations and sources. A publisher-authenticated manifest would provide stronger provenance evidence.
- Finding / current status
- ClaimWindow.html occurs once, at 2,712 bytes. SendMoneyWnd.html and BankWindow.html were not found. The executable contains the string Sep 11 2025.
- Limits
- Chronological consistency does not authenticate the publisher, independently verify release number 630, or prove that all files belong to one release.
- Next test
- Seek an independently obtained comparison manifest and preserve contradictions instead of discarding them.
R04 · completed synthetic controls
Container identity and member identity
Can packaging change a ZIP digest while leaving its named file payloads unchanged?
- Method
- Create original tiny XML fixtures in memory. Compare baseline, repackaged, changed-payload, and missing-member containers against declared predictions.
- Inputs
- The authored fixtures and standard-library Python method; no EverQuest files are used.
- Alternatives considered
- The JSON exposes observed digests and comparisons. The small public script independently reproduces the experiment.
- Finding / current status
- Repackaging changed the ZIP digest while preserving member digests. Separate controls detected a changed payload and an absent member.
- Limits
- These are method controls, not observations about the client's actual packaging history or behavior. Container bytes can vary between compressor implementations.
- Next test
- Repeat with a different ZIP implementation and compare member-level findings even when container bytes differ.
PROPOSED FOLLOW-UP
These designs have no reported outcomes. Each starts with narrower inputs and identifies what would justify extending the inquiry.
P01 · proposed; no independent comparison obtained
Independent whole-installation coherence comparison
Is this reference consistent with an independently and lawfully acquired installation, or does it combine components from different histories?
- Method
- Begin with local file manifests: normalized relative name, size, and content digest. Compare matching, missing, additional, and changed entries; then investigate a bounded set of disagreements.
- Inputs
- Independent reference evidence and consent to publish sanitized aggregate differences. Full file transfer is not a starting requirement.
- Alternatives considered
- Run tools where each copy is held and exchange manifests or aggregate disagreements. Request selected inputs only when a specific unresolved question requires them and a lawful basis exists.
- Finding / current status
- No comparison result is reported. Coherence, modification history, and acquisition provenance remain unresolved.
- Limits
- Two matching installations could share the same prior source. Agreement is not automatically independent publisher authentication.
- Next test
- Define independence criteria before accepting a comparison and record conflicting evidence and the acquisition facts that can be disclosed safely.
P02 · proposed; no runtime study performed
Controlled presentation behavior
Which relationships predicted by static interface definitions are exercised in a suitable isolated execution environment?
- Method
- Specify an authorized environment and one observable presentation task; establish a baseline, change one permitted variable, and log reproducible outcomes without contacting a live game service.
- Inputs
- A suitable lawful installation, documented dependencies, an explicit task, and sufficient resources for that task. Necessary inputs must be justified separately.
- Alternatives considered
- Try static analysis and locally executed instrumentation before sharing files. A component subset may answer some questions; broader inputs may be justified for others.
- Finding / current status
- No rendering, compatibility, accessibility, performance, or dependency outcome is claimed.
- Limits
- A local rendering result would not reconstruct historical server state or establish access to an official service. The archive does not provide that environment.
- Next test
- Write the success and failure criteria in advance, identify legal and safety constraints, and retain negative or inconclusive outcomes.
P03 · proposed; no custodial consortium or recovery exercise claimed
Preservation adequacy and recovery evidence
What evidence would show that a preserved object remains intact and useful for a stated research task?
- Method
- Define significant properties, record a baseline, verify digests on a schedule chosen by actual custodians, and test recovery against a stated task with documented outcomes.
- Inputs
- A documented preservation role, appropriate storage, verification records, and an independent lawful basis for any complete-copy transfers.
- Alternatives considered
- Publish digest verification and recovery reports without distributing protected file contents. Existing authorized custody may be sufficient; additional copies require their own rationale.
- Finding / current status
- No independent custodians, geographically distributed holdings, periodic checks, or recovery success are asserted by this proposal.
- Limits
- A torrent is a transfer mechanism, not evidence that a preservation arrangement exists. Integrity does not establish usability, provenance, or distribution permission.
- Next test
- Specify the loss or failure scenario, the protected properties, and a test that can fail before claiming preservation success.
After the investigation.
Researchers are expected to make a reasonable, good-faith effort to contribute methods, findings, corrections, or negative results after their work, when possible. Access does not require advance contact, an account, an application, or submission of a result.
Prepare a useful contribution → · Reference ZIP and torrent downloads →
CONTINUE READING