FOLLOW US

Largest alleged file-size reductions — May 13–14 revision

FBI 62-HQ-83894 Section 6371 MB → 61 MB (−310 MB)

DoW Box 7 Incident Summaries 1–100247 MB → 33 MB (−214 MB)

DoW Box 7 Incident Summaries 101–172243 MB → 29 MB (−214 MB)

All FBI 62-HQ-83894 sectionsMultiple reductions

The second tracker claimed to have preserved both the pre- and post-change versions of all 70 files using content-addressed storage — meaning the original bytes, as distributed by the government on May 8, were allegedly still accessible in the researcher's archive even after war.gov's URLs began serving the smaller replacements.

The researcher was careful with language. They described the phenomenon as a potential "cover-up-vs-cleanup signal" and explicitly noted that size reduction alone could have benign explanations — re-compression, removal of duplicate pages, conversion to lower-resolution scans. But, they argued, the combination of silence and scale warranted scrutiny.

"File-list integrity is the easy half of the problem," the researcher wrote in the public thread. "Byte-level integrity is where the cover-up-vs-cleanup signal actually lives."

A Dispute Over Methodology — And What It Reveals

The two researchers — whose projects were entirely independent and had not coordinated prior to the public thread — quickly engaged each other on a technical point that illuminates the complexity of tracking government document changes: what, exactly, counts as a "removal"?

The first tracker reported five genuine file removals. The second tracker, using a different reconciliation algorithm, classified most of those same files not as removals but as renames — arguing that the underlying numeric IDs matched entries that reappeared under new labels. Under that reading, the geographic changes (e.g., "Japan, 2023" becoming "INDOPACOM 2023") were bureaucratic relabeling rather than deletion.

The second tracker offered this reconstruction in the public forum:

Disputed classification — rename vs. removal

Djibouti 2025 (PR43)→ Africa 2025 (PR043)

Southern US 2020 (PR40)→ Middle East 2020 (PR040)

East China Sea 2024 (PR46)→ INDOPACOM 2024 (PR046)

Japan 2023 (PR47)→ INDOPACOM 2023 (PR047)

Indo-PACOM 2024 (PR48)→ INDOPACOM 2024 (PR048)

Under this reading, only one file — a duplicate entry for FBI 62-HQ-83894 Section 6 with no rename candidate — represented a genuine deletion. The two researchers agreed to compare methodologies further, with both acknowledging that the absence of an official changelog made ground truth impossible to establish with certainty.

This methodological dispute is, in a sense, the story within the story: that independent researchers cannot agree on what happened because the government has provided no authoritative record of what it changed, why, or when.