Skip to content

release: SPEC/homepage sync + Wikipedia backfill + 362 verified promotions (2026-09-24) - #243

Merged
Seungpyo1007 merged 75 commits into
mainfrom
develop
Sep 24, 2026
Merged

Seungpyo1007 merged 75 commits into
mainfrom
develop

Conversation

@Seungpyo1007

@Seungpyo1007 Seungpyo1007 commented Sep 24, 2026 •

Copy link
Copy Markdown
Member

Summary

Releases everything accumulated on develop since the last release (#195):

  • Wikipedia URL backfill: 39+ confirmed smartphone/tablet citations across several batches
  • soc_not_after_device false-signal fix effect: 362 records promoted to verified (20.41% → 20.75% record-level)
  • Full dump refresh (191,210 records across all 12 collections, catching up a backlog that had missed CPU/GPU/SoC/laptop for weeks)
  • Homepage meta description + README accuracy fixes (post-game-split wording)
  • Routine verify-status aggregate refreshes and submodule bumps

Test plan

  • python -m app.validate passed on develop tip

Closes #1

TechEngineBot and others added 30 commits September 18, 2026 13:03
Lets a preview build serve develop's UI while reading main's live
public dump instead of regenerating its own — no code needed beyond
setting the env var on the preview deployment.
Move 2025 M4 Air and M5 Pro records out of the 2024 folder, set
availability dates from Apple Newsroom, and normalize chip names
against Support tech specs.
Collapse scrape artifacts such as "Snapdragon Snapdragon X Elite" and
"Intel 210H 210H" to vendor-canonical names, citing Qualcomm/Intel/OEM
specs in source_urls.
Auto-promotions from the verification layer (green+live-source or crossref-confirm).
Each flip is verified:false->true only; see data/_verify/ledger.jsonl. Refs #1
Append an en.wikipedia.org link to source_urls on 107 kaggle-only GPU records. The match required the same desktop/mobile marker, a compatible bus, and at least two agreeing specs. The other rows in the 300-card sample were left unchanged.
Append an en.wikipedia.org link to source_urls on 159 kaggle-only GPU records. The match required the same desktop/mobile marker, a compatible bus, and at least two agreeing specs. The other rows in this 400-card sample were left unchanged.
Append an en.wikipedia.org link to source_urls on 110 kaggle-only GPU records. The match required the same desktop/mobile marker, a compatible bus, and at least two agreeing specs. The other rows in this 400-card sample were left unchanged.
Append an en.wikipedia.org link to source_urls on 25 kaggle-only GPU records. The match required the same desktop/mobile marker, a compatible bus, and at least two agreeing specs. The other rows in this 119-card sample were left unchanged.
Add Wikidata entity URLs for 54 Brand records whose English name or alias, founded year, and official website or country agree. The other 23 records in this last batch stay unchanged.
@github-actions github-actions Bot added data Dataset changes documentation Improvements or additions to documentation enhancement New feature or request labels Sep 24, 2026
@Seungpyo1007
Seungpyo1007 merged commit b3f89a4 into main Sep 24, 2026
@TechEngineBot

Copy link
Copy Markdown
Member

🔎 Data verification — Tier 0 (offline existence/trust)

Scored by app.verify; posted by TechEngineBot. Informational only — the structural gate (app.validate) is separate and authoritative for merge.

Changed records in this PR

No records scored.

Full-dataset baseline

105800 record(s) scored.

%%{init: {"theme":"base","themeVariables":{"pie1":"#3fb950","pie2":"#d29922","pie3":"#f85149","pieStrokeWidth":"0px","pieOpacity":"1"}}}%%
pie showData
    title Verification bands — all records
    "Green" : 23401
    "Yellow" : 81278
    "Red" : 1121
Loading
Category Distribution Total 🟢 🟡 🔴 🟢 %
brand 🟩🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨 207 24 183 0 11.6%
soc 🟩🟨🟨🟨🟨🟨🟥🟥🟥🟥🟥🟥 2104 231 825 1048 11.0%
smartphone 🟩🟩🟨🟨🟨🟨🟨🟨🟨🟨🟨🟥 93396 17900 75429 67 19.2%
tablet 🟩🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨🟥 3455 252 3202 1 7.3%
watch 🟩🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨🟥 433 39 390 4 9.0%
pda 🟩🟩🟩🟩🟩🟩🟩🟨🟨🟨🟨🟨 140 83 57 0 59.3%
gpu 🟩🟩🟩🟩🟩🟩🟩🟨🟨🟨🟨🟨 2030 1211 819 0 59.7%
cpu 🟩🟩🟩🟩🟩🟩🟩🟩🟩🟩🟨🟥 4035 3661 373 1 90.7%
All 🟩🟩🟩🟨🟨🟨🟨🟨🟨🟨🟨🟥 105800 23401 81278 1121 22.1%

Hard violations (forced red):

Count Check
1 !boost_ge_base

green = authoritative source + complete + consistent · yellow = plausible, needs confirmation · red = sparse/weak source or a hard contradiction. Promotion to verified runs in the scheduled verify-network workflow.

@TechEngineBot

Copy link
Copy Markdown
Member

TechEngine change review: PASS

Check Result
python -m app.validate PASS
New hard integrity anomalies vs PR base PASS

Changed data

Category Added Modified Deleted Added verified Added unverified Added Kaggle-sourced
brand 0 0 0 0 0 0
soc 0 0 0 0 0 0
smartphone 0 0 0 0 0 0
tablet 0 0 0 0 0 0
watch 0 0 0 0 0 0
pda 0 0 0 0 0 0
gpu 0 0 0 0 0 0
cpu 0 0 0 0 0 0

Changed record examples

  • No data file changes detected.

Heuristic review

  • Heuristic warnings: none found.

@TechEngineBot

Copy link
Copy Markdown
Member

TechEngine validation stats: PASS

Data summary

Category Total Verified Unverified Missing verified Tracked Verified % of tracked
brand 207 24 183 0 207 11.6%
soc 2104 184 1920 0 2104 8.7%
smartphone 93396 17495 75901 0 93396 18.7%
tablet 3455 248 3207 0 3455 7.2%
watch 433 39 394 0 433 9.0%
pda 140 74 66 0 140 52.9%
gpu 2030 1021 1009 0 2030 50.3%
cpu 4035 2868 1167 0 4035 71.1%
all 105800 21953 83847 0 105800 20.7%

Warning

Tracked verified coverage is below 50% for tablet 7.2% (248/3455), soc 8.7% (184/2104), watch 9.0% (39/433), brand 11.6% (24/207), smartphone 18.7% (17495/93396), all 20.7% (21953/105800).
Tracked coverage excludes records missing the verified field; see the Missing verified column for those records.
This does not fail validation. Keep imported records verified: false until manual audit, but treat this as follow-up verification work before relying on the affected categories as curated data.

Validation notes

  • The integrity gate blocks only hard anomalies introduced by this PR relative to its merge base.
  • Full advisory outlier listings are suppressed on successful runs because they are dataset-wide and mostly stable between PRs.
  • Failure runs still include a detailed log excerpt for debugging.

Key output:

## app.validate
## integrity_check.py (PR head compared with PR base)
loaded CPU=4035 GPU=2030
Integrity section Flagged lines
structural 18
CPU name/tier consistency (desktop mainstream only) 0
CPU single>multi (cinebench/geekbench — should be multi>=single) 0
CPU era-vs-score outliers 8
CPU cross-source ratio outliers (possible wrong-variant) 155
GPU cross-source ratio outliers + sanity 18

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

data Dataset changes documentation Improvements or additions to documentation enhancement New feature or request

Projects

Status: Done

Development

Successfully merging this pull request may close these issues.

2 participants