A filing, statute, agency record, map, or dataset published by the responsible public body.
Methods & corrections
A public position can still use disciplined evidence.
The Observatory is not neutral about Pacifico Cedar Creek at this location. It is rigorous about what the record proves, what remains an assertion, and what evidence could change a conclusion.
Two dimensions
Source type is not claim status
An applicant statement can appear inside an official government filing. The source is official; the underlying assertion is still the applicant’s.
A statement about what an official record contains. This does not independently prove every underlying applicant assertion.
A fact, forecast, description, or claim made by the applicant and represented accurately here without automatic adoption as fact.
A modeled outcome based on assumptions. It is not a guarantee of future jobs, taxes, earnings, or spending.
A calculation or comparison performed by the Observatory from cited source data.
A technically supported conclusion that remains subject to project-specific evidence and can be revised.
A question worth investigating, not a proven condition. It is paired with the evidence needed to test it.
Claim lifecycle
How a conclusion enters the public site
- 01
Acquire the source
Record publisher, date, version, URL, and project linkage.
- 02
Index exact support
Create page-level findings before turning a long filing into a public claim.
- 03
Classify the claim
Separate applicant assertion, record content, cross-check, inference, and hypothesis.
- 04
Name the next evidence
A conclusion is stronger when the public can see what would confirm, narrow, or overturn it.
- 05
Publish material corrections visibly
Errors and materially misleading wording are dated and explained. New records and routine site updates are handled separately.
Published corrections
What we changed—and why
This log is limited to material factual, analytical, or editorial corrections. New records, permit developments, meeting notices, and routine site improvements appear under Updates.
LIV town-hall registration status
We previously said registration for the August 20 LIV Pacifico Town Hall had closed. LIV’s event page did not support that statement. We changed the guidance to direct readers to check current RSVP availability and confirm the walk-in policy.
County Item E.4 guidance
Our August 9 and August 10 updates told residents to support draft Item E.4 while asking the County to preserve existing tools. That guidance was wrong. We changed it to oppose the draft as written and ask the County to use its existing lawful authority, preserve every option, and not return the draft unchanged.
Amazon and EdgeConneX tract extents
Earlier map depictions showed undersized ownership snapshots. We replaced them with county parcel boundaries. Amazon is now shown as an approximately 1,300-acre union of 24 former CTX/Creekside parcels; EdgeConneX mapping separates AUS01, AUS02, Campus 2, and the full Campus 3 host parcel, with the unfiled Campus 3 project footprint clearly qualified.
Pacifico boundary and distance measurements
We replaced earlier project geometry with the eight property IDs listed on J0049 page 7 and changed distance measurements from center-point circles to offsets from the filed project perimeter. The filed tract descriptions total 2,810.0553 acres; Pacifico’s narrative says approximately 2,842 acres.
JETI compelling-factor requirement
We previously suggested dispatchable generation might be exempt from JETI’s compelling-factor requirement. That was wrong. SB 2322 did not become law, and the current statute and Comptroller FAQ retain the requirement.
EDCAUS11 and EDCAUS12 owner of record
We corrected the owner-of-record labels to Burr Computer Environment Inc., as listed in the TDLR filings. Any other corporate or tenant relationship remains unproven.
Sources for the JETI correction
SB 2322 proposed the dispatchable-generation exemption; it did not enact it.
Current §403.609(b)(3) and the Comptroller’s current FAQ still state the compelling-factor requirement.
Privacy
Residences are context; residents are not identifiers
An anonymous public address-point pattern and broad supporter zones can show community proximity and voluntary opposition without publishing a household’s street address, ownership, contact information, or private submission.
Public: official parcels, filed project locations, agency records, public infrastructure, named public facilities, aggregate counts, and residential points stripped to coordinates only.
Protected: resident names, street addresses, ownership, contact details, private submissions, and any tooltip or export that could identify a household. The opposition feature does not ask for, collect or publish supporter names.
Voluntary opposition map: exact addresses are reviewed privately and assigned to broad, fixed areas rather than moved or “jittered” pins. Bubble centers and boundaries are fixed in advance and are never calculated from submitted addresses. After private review, an approved anonymous count may be published immediately in the selected broad-area bubble; there is no minimum batch size or waiting period. The readable address is scrubbed before the count becomes public. Duplicate entries in the launch list were consolidated; new entries use a one-person attestation, private contact-email confirmation, automatic protected location grouping, and organizer review to reduce duplicates. Multiple adults at one location may count separately as supporters but share one represented location. The result is a voluntary supporter count—not a scientific survey of nearby residents.
No names: the submission form and review workflow do not request a supporter name or provide any public-name option. Public opposition data consists only of totals, represented- location counts, fixed broad-area labels, and the count date.
Withdrawals: an organizer can remove a support entry and its anonymous contribution to the public counts immediately. The submitted address itself was already scrubbed at approval and is never displayed publicly.
Retention and abuse controls: unreviewed entries older than 90 days are removed during routine queue cleanup. A readable address is deleted at approval. A protected duplicate-check code, the assigned fixed broad area, and the private contact email remain so the entry can be verified and managed. A keyed, non-readable network code rotates hourly to limit repeated submissions; stale rows are pruned after 48 hours during later submission activity, and the source network address is never stored.
Anonymous site analytics: the Observatory counts page views, browser-level visitors, sessions, referrer hostnames, active-page time, device category, and approximate city, region, and country. It also counts public interaction choices such as links, downloads, successful form submissions, and claim explanations opened or closed. Raw IP addresses, exact coordinates, search queries, form contents, and private submissions are not stored in traffic analytics. Browser Do Not Track and Global Privacy Control signals are honored.
Public-page monitoring signals: the Observatory separately counts requests to the homepage, Updates index, and RSS feed so repeated automated checks can be distinguished from ordinary reader analytics. For 90 days, it retains the requested path, time, response status, broad client category, country, network operator, and a keyed one-way source hash. Raw IP addresses and full user-agent strings are not stored. A repeated pattern or network operator does not establish who directed a request.