US 7791704

I'll search for litigation and docket information on this patent, then compile the summary.

Let me check specifically for litigation involving this patent and any USPTO proceedings.

Let me run a final check for any PTAB or district-court activity specific to this patent.

US Patent 7,791,704 — Summary

⚠️ Scope note on your two requested checks

• USPTO database / CAFC 2026 dockets for this specific number: no results found. I ran multiple targeted searches for the literal strings "7791704" / "7,791,704" paired with litigation, PTAB/IPR, Federal Circuit, and assignee terms. Nothing authoritative surfaced. Every apparent "hit" was a false positive on a different identifier, and per your instruction I did not auto-correct them:
• A Brazilian notary register (jucisrs.rs.gov.br, 2021) listing a personal ID number 7791704.
• A 2025 Santa Catarina municipal gazette publication number 7791704.
• Ormco Corp. v. Lancer Orthodontics (C.D. Cal. 8:14-cv-00545) asserting U.S. 8,469,704 — a different patent.
• Fellowes litigation referencing U.S. 9,724,704 — a different patent.
• I could not directly query USPTO PatentCenter / PTAB API. My bibliographic data below is drawn from the Google Patents record for US7791704B2 (which mirrors USPTO data), fetched 2026-09-27, plus the assignment history it reports. Treat litigation/PTAB status as unverified rather than confirmed-negative, though I have no indication of any proceeding.

Bibliographic data

Cited prior art (2 references, both cited by examiner):

• JPH08129166A — Sanyo Electric Co Ltd, "Liquid crystal display device" (1994-10-31 → 1996-05-21)
• JPH08171095A — Sanyo Electric Co Ltd, "Liquid crystal display device and production of liquid crystal display device" (1994-12-16 → 1996-07-02) — referred to in the specification as "patent document 1"

The specification expressly distinguishes JPH08171095 on the basis that its sealing material is a plain rectangular frame that does not follow the recessed shape of the upper substrate.

Cited by (4): US 2013/0135830 A1 and US 8,982,575 B2 (Au Optronics), US 9,791,753 B2 (E Ink Holdings), and US 12,525,523 B2 (Wuhan China Star Optoelectronics).

Assignee / chain-of-title history (as reported)

1. 2008-05-27 — Assignment from inventor Abe to Hitachi Displays, Ltd. (effective 2008-05-15).
2. 2011-10-14 (effective 2010-06-30 / 2010-10-01) — Company split transferring 50% share from Hitachi Displays to IPS Alpha Support Co., Ltd., then merger of IPS Alpha Support into Panasonic Liquid Crystal Display Co., Ltd.
3. 2023-11-17 records — Hitachi Displays → Japan Display East, Inc. (name change, eff. 2012-04-01) → Japan Display, Inc. (name change, eff. 2013-04-01); and Panasonic Liquid Crystal Display → Panasonic Intellectual Property Corporation of America (nunc pro tunc assignment, eff. 2023-08-28).
4. 2025-07-10 (effective 2025-06-25) — Assignment from Japan Display Inc. to Magnolia Purple Corporation.

Context from press coverage: Magnolia Purple is one of three IP-holding subsidiaries (Magnolia Purple/White/Blue) created by Japan Display (JDI) on 2025-05-19 in Minato-ku, Tokyo, into which JDI contributed LCD and OLED patent assets; JDI sold all shares in these subsidiaries to its majority shareholder Ichigo Trust, completing 2025-07-30, while retaining a royalty-free license. This is consistent with the recorded 2025 assignment — i.e., the patent now sits in a monetization vehicle.

Abstract

A liquid crystal display panel is provided which exhibits excellent durability by protecting portions of a substrate most liable to be cracked or chipped. In a panel with a sealing material between a TFT substrate and a CF substrate and a liquid crystal layer held therebetween, the CF substrate has one-side surface formed in a recessed shape as viewed from above; a flexible printed circuit board (FPC) is connected to a region of the TFT substrate where the two substrates do not overlap; and the sealing material is formed along the shape of the CF substrate. Regions where the substrates overlap on both sides of the FPC can be filled with sealing material with no gap.

Plain-language overview of the claims

Independent Claim 1 — A liquid crystal display panel comprising a sealing material between a lower substrate and an upper substrate, with a liquid crystal layer held between them, wherein:

• the upper substrate has one side formed in a recessed shape as viewed from above (i.e., a cut-out/notch profile rather than a full rectangle);
• a flexible printed circuit board is connected to a region of the lower substrate where the upper and lower substrates do not overlap (the exposed ledge);
• the sealing material is formed along the shape of the upper substrate, including the recessed shape (so it follows the notch, rather than being a plain rectangular frame); and
• the sealing material is wider at the portion that follows the recessed shape than at a portion that does not follow the recessed shape.

The final "wider at the recess" limitation is notable: it is not reflected in the pre-grant abstract/description language ("a sealing material is formed along a shape of the CF substrate"), which suggests it was added during prosecution to secure allowance over the Sanyo references. I flag this as an inference from comparing the abstract to the granted claim text, not a verified prosecution-history finding.

Dependent claims:

• Claim 2 — the upper substrate is a plastic or glass substrate.
• Claim 3 — at the regions where the substrates overlap on both sides of the FPC, the sealing material fills those regions with no gap (expressly defined in the spec as not intentionally leaving a space).
• Claim 4 — a driver chip for controlling panel driving is placed on the lower substrate in the non-overlapping region.
• Claim 5 — the FPC and the driver chip are both arranged parallel to a side surface of the lower substrate.
• Claim 6 — the FPC and the driver chip are arranged parallel to each other AND parallel to a side of the lower substrate (the narrower arrangement, whose stated benefit is shorter interconnect lines between chip and FPC and thus lower wiring resistance).

Note that claims 5 and 6 are worded differently ("a side surface" vs. "a side of the lower substrate"); I have quoted them literally rather than harmonizing them.

Stated purpose of the invention (from the spec): conventional panels have a one-sheet-substrate ledge (where only the TFT substrate exists) that is mechanically weak; external stress can warp it, peel the driver chip/FPC, crack or chip the glass, and disconnection or debris causes display defects. Overlapping the CF substrate at the four corners of the TFT substrate and running the sealant along the CF substrate's shape protects the vulnerable edge regions.

Uncertainty / caveats

• No CAFC 2026 docket or USPTO PTAB proceeding could be located for US 7,791,704. I cannot affirmatively certify that none exists — only that my searches found none, and that all numeric near-matches were unrelated patents (8,469,704; 9,724,704) or non-patent documents.
• Bibliographic data is from the Google Patents mirror of the USPTO record, not a direct PatentCenter retrieval. Item-level details (e.g., the exact scope of the 2011 certificate of correction, and the precise pre-grant claim text) would need confirmation against the official file wrapper.
• The claim-1 "wider sealing material" limitation appears to be a prosecution amendment; I did not verify this against a USPTO transaction-history record.

I'll research litigation involving US patent 7791704. Let me run several targeted searches.

Initial results don't show litigation. Let me check specialized patent litigation databases directly.

Litigation Search Report — U.S. Patent No. 7,791,704

Bottom line

No litigation involving U.S. Patent No. 7,791,704 was found. Based on the searches run, I found no district court complaints, no PTAB proceedings, no Federal Circuit appeals, and no ITC Section 337 investigations asserting or challenging this patent.

What was searched

Important disambiguation notes

Several false positives appeared repeatedly and should be excluded:

• Case Nos. 17-1704 (CAFC) — Coho Licensing LLC v. Glam Media/Ning/AOL/Rovi. This is a docket number (17‑1704), not Patent No. 7,791,704. Do not conflate them.
• U.S. Pat. No. 5,776,170 ('170 patent) and U.S. Pat. No. 7,204,808 — appeared in unrelated invalidity-contentions and Mass Engineered Design dockets; not the patent at issue.
• The "Cited By" entries on the Google Patents record — US20130135830A1 and US9791753B2 (Au Optronics Corp.; E Ink Holdings), and US12525523B2 (Wuhan China Star Optoelectronics, 2026) — are later patent publications that cite 7,791,704 as prior art. These are citations, not litigation involving the '704 patent.

Case-by-case table

Context relevant to the negative finding

The record is consistent with a never-asserted patent:

• Ownership chain never entered monetization hands until recently. Original assignees Hitachi Displays, Ltd. and IPS Alpha Technology, Ltd. (2008); then Panasonic Liquid Crystal Display Co., Ltd. (2011); Panasonic Intellectual Property Corporation of America (2023); Japan Display, Inc.; and finally Magnolia Purple Corporation on 2025‑06‑25 (reel/frame 071890/0202). If the patent were being actively asserted, the usual pattern would show Magnolia Purple (a monetization entity) as plaintiff — but no such case surfaced.
• No post-grant challenges. The entire PTAB/invalidation footprint reported for this patent is empty; the only citations are the two Japanese Sanyo references (JPH08129166A, JPH08171095A) cited during prosecution.
• Maintenance fees paid through the 12th year (2022‑02‑23), and the Google Patents entry lists status Active with an adjusted expiration of 2028‑08‑28 — so the patent is enforceable and could still be asserted; absence of litigation is simply an absence of record, not an abandonment.

Caveats and limitations

I could not directly query the Unified Patents case list, RPX Insight, PACER, or Docket Navigator with authenticated/full access in this session; the searches above relied on indexed web results for those portals. It is possible that:

• a very recent (2026) complaint exists that is not yet indexed by the search engines;
• litigation is pending under a sealed docket or where the patent is asserted only via a related family member (e.g., the CN or JP counterparts, CN101266367B / JP2008225414A) rather than the U.S. patent itself.

If you need certainty, the authoritative checks would be: (1) PACER Docket Navigator "Patents" search on 7,791,704; (2) Unified Patents' litigation database by patent number; (3) USPTO PTAB API for IPR/PGR/CBM on the patent. I am flagging this explicitly rather than representing that the negative result is exhaustive.

I'll verify the structured ODP result against public sources before writing this up.

Confirmed the first pass shows only false positives. Let me run a few more targeted checks specific to the patent's owners and monetization vehicle.

Let me do a final pair of checks for any ex parte reexamination (non-AIA but adjacent) and any NPE/assertion campaign activity.

PTAB proceedings on file — US 7,791,704

Bottom line up front: there are no AIA trial proceedings on this patent. The structured ODP block is unambiguous — zero IPRs, zero PGRs, zero CBM reviews, no institution denials, no settlements. My independent web searches did not surface any proceeding the ODP list might have missed. This section is therefore a verified-negative, and the strategic value is in what that absence means, not in a claim-by-claim kill sheet.

Proceedings overview

Total AIA trials: 0 (0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denials) — meaning there is no claim-level disposition of US 7,791,704 to rely on: all six claims stand exactly as granted on 2010-09-07, as amended only by the Certificate of Correction recorded 2011-09-13, and a defendant facing assertion today must build its invalidity case from scratch rather than riding an existing FWD.

⚠️ Verification notes and a name-collision warning

Because a "no results" finding is only as good as the search behind it, here is exactly what I checked and what I could not check:

Name-collision warning — do not conflate. Searching the current owner surfaced Magnolia Medical Technologies, Inc. v. Kurin, Inc., IPR2026-00097, Paper 17 (Director May 14, 2026) — a precedential Director decision on discretionary denial. That case involves a medical-device patent (U.S. 12,138,052) and a blood-collection company, not Magnolia Purple Corporation, the Japan-based LCD/OLED patent-holding vehicle that acquired this patent on 2025-06-25. The two "Magnolias" are unrelated. This matters practically: if your opponent's demand letter or a vendor's docket alert shows "Magnolia," confirm the entity before treating that precedent as related to your dispute. (The Magnolia/Kurin decision itself is relevant to your IPR calculus — see Strategic summary, below.)

I could not directly query PTAB E2E (ptacts.uspto.gov) or the ODP API from this session. A human confirmation against PTAB E2E and CourtListener is warranted before you treat the zero-count as final for litigation purposes.

Full-text search caveat — where my "no PTAB activity" claim lives

The absence of proceedings should not be read as a guarantee. Two gaps worth stating plainly:

1. AIA-trial lists do not include ex parte reexaminations. Those are prosecuted ex parte and live in a separate USPTO record, not the PTAB proceedings list. I found no indication of a reexam on the '704, but the structured block in this prompt would not capture one either way.
2. Terminal disclaimers, certificates of correction, and reissue are similarly outside the AIA-trial list. The 2011-09-13 Certificate of Correction is confirmed in the bibliographic record, but its precise scope is not resolved here — I could not retrieve the item-level detail. If claim scope is dispositive to you, pull the correction from the official file wrapper rather than assuming it is a typographical fix.

Strategic summary

Claim status: everything is UNTESTED. No claim has been canceled, narrowed by AIA trial, or held patentable by the Board. Specifically:

• Claim 1 (independent) — alive and unadjudicated. Includes the recessed-upper-substrate + sealant-along-the-recess + "width of the sealing material at a first portion that is formed along the recessed shape is wider than a width of a second portion" limitation. Per the earlier-generated analysis, that final "wider" limitation is not reflected in the pre-grant abstract language, suggesting it was a prosecution amendment — an inference from comparing abstract to granted claim text, not a file-wrapper-verified prosecution finding.
• Claims 2–6 (dependent: plastic/glass substrate; gap-free sealant flanking the FPC; driver chip on the non-overlapping ledge; and the two differently-worded parallel-arrangement claims) — all alive and unadjudicated.

Estoppel landscape: nothing is estopped, but nothing is pre-decided either. Because no FWD has ever issued, § 315(e)(2) estoppel has never attached to any party. A defendant today is therefore free to raise any § 102/§ 103 ground in district court, and equally free to file an IPR. But note the corollary: you gain no benefit from prior petitioners' work, because there has been no prior petitioner. There is no art that has already been "spent" and no claim construction the Board has blessed.

Pattern signals: the notable signal is silence from a patent that has been in force for 16 years and just changed hands into a monetization vehicle. Three things stand out:

1. No defensive aggregator. No Unified Patents, RPX, or similar challenger appears anywhere in the chain. The "cited by" set (Au Optronics, E Ink, Wuhan China Star) reflects later patent citations, not challenges.
2. No serial petitioner, no appeal history. Nothing to indicate this patent has been litigated to the Board or appealed to the Federal Circuit.
3. The ownership change is the live variable. Magnolia Purple Corporation acquired the patent from Japan Display Inc. effective 2025-06-25 (recorded 2025-07-10). Magnolia Purple is one of three JDI-created IP-holding subsidiaries (Magnolia Purple/White/Blue, formed 2025-05-19) whose shares were sold to Ichigo Trust — a monetization posture. A patent moving into a dedicated assertion vehicle after 16 years of quiet is a forward-looking risk indicator, and it is precisely the scenario where a first-ever IPR tends to appear.

The clock is the dominant factor. Adjusted expiration is 2028-08-28 — roughly 23 months from today (2026-09-27). That compresses the IPR value proposition sharply:

• An IPR petition filed now would target a FWD in roughly late 2027 (12 months from institution), leaving a thin tail of term to protect.
• Under Director Squires' discretionary-institution regime, the age of the patent cuts against your petition, not for it — the Office has held that "the length of time [a patent] has been in force weighs against institution of an IPR," and that AIA review should not be a "second bite at the apple." The Magnolia Medical v. Kurin precedential decision (IPR2026-00097, Paper 17, Director May 14, 2026) reinforces that a petitioner who already had a full and fair shot at validity in court may be denied institution outright. The one counterweight consistently credited by the Director is a substantial showing of USPTO error — the Padagis v. Neurelis line. If a § 102/§ 103 theory against claim 1 exists, it will have to be framed as examiner error, not merely as art the examiner never saw.
• A further proposed-rules overlay: the October 17, 2025 rulemaking package would deny institution if a district-court trial is "more likely than not" to precede the FWD, and would require a petitioner to stipulate to forgo § 102/§ 103 invalidity defenses in other proceedings as the price of institution — a broader preclusive effect than congressionally mandated § 315(e) estoppel. As of my searches these rules had not been finalized, but they are the direction of travel.

Recommended next steps

If you are a defendant today, and no AIA trial exists:

1. Do not expect an FWD to save you. There is no prior PTAB outcome to lean on. Quote-ready: the Board has issued zero decisions on this patent; claims 1–6 are all in force and unadjudicated. Treat the patent as fully viable for validity purposes and litigate it on the merits.
2. The § 315(b) clock runs from your complaint. If you have already been served, you have one year to petition. If this patent is now in a monetization vehicle, assume a demand letter or complaint is the trigger and calendar the deadline immediately — a 2028-08-28 expiry means a late petition is a wasted petition.
3. Two-track your validity work. Because no estoppel exists yet, you can hedge: develop the full § 102/§ 103 case for district court (where any art type and system prior art is available) while preserving an IPR as an option. Do not assume IPR is the cheaper path here — the patent's age is working against institution.
4. Attack the claim-1 "wider sealing material" limitation directly. It is the most prosecution-fragile element on the face of the claim (absent from the pre-grant abstract, per the earlier analysis). Get the file wrapper and confirm whether it was added by amendment and for what stated reason. If it was the allowance hook, it is also the natural § 112 written-description / indefiniteness target — and that theory is not available in an IPR (IPR is limited to § 102/§ 103 on patents and printed publications), which is a concrete reason to preserve invalidity in district court rather than betting on the Board.
5. Confirm the negative independently. Before committing strategy, run PTAB E2E and CourtListener for the patent number and for the full assignee chain (Hitachi Displays → Japan Display East → Japan Display → Magnolia Purple), and pull any ex parte reexamination record. I found none, and the ODP block lists none, but the structured list would not capture a reexam, and I could not reach the E2E interface myself.
6. Watch the timing, not just the absence. If the patent has been quietly held for 16 years and just landed in an Ichigo Trust monetization subsidiary, the absence of PTAB activity is a pre-assertion baseline — not evidence the patent is weak. The most likely future development is a first IPR filed by a first-sued defendant, and it would be filed against a patent with less than two years of term left. That is a strong candidate for discretionary denial unless the petition demonstrates material examiner error.

One explicit non-finding, stated so it isn't mistaken for a finding: I did not locate any Federal Circuit appeal, any settlement, any institution decision, or any claim-level disposition for this patent — because there is nothing to locate. Any proceeding number, panel, or FWD disposition you see asserted for US 7,791,704 should be treated as unverified until pulled from PTAB E2E.

I'll research the ownership chain, correspondent data, and any assertion activity before reconstructing the record.

Let me probe for assertion activity and correspondent/record data on the specific patent and the Magnolia entities.

Let me make a final attempt at correspondent data and any post-2025 assertion activity.

Ownership reconstruction — US 7,791,704 (Liquid crystal display panel)

⚠️ Method / access note (read first)

I could not query the USPTO Assignment Center (assignmentcenter.uspto.gov) or the legacy assignment.uspto.gov interface directly from this environment, and no web search surfaced the correspondent-of-record fields (recording attorney / firm) for this patent's reel/frame entries. I made several targeted attempts (searching the literal reel/frame strings 021037/0692, 027063/0019, 065615/0327, 071890/0202, and the assignee names) and got no authoritative correspondent data. Every reel/frame number below is reproduced from the Google Patents "Legal Events" mirror of the USPTO assignment record; the correspondent column is therefore blank and is flagged as not verified. This is a genuine gap, not a finding of "no correspondent."

The good news: the reel/frame, conveyance type, assignor, assignee, and execution/recording dates are all present in the mirror, so the chain of title is reconstructible even though the correspondent tell is not.

Inventors

• The sole inventor assigned his rights to Hitachi Displays, Ltd. by an instrument executed 2008-05-15 and recorded 2008-05-27 (Reel 021037/0692), i.e., roughly two months after the 2007-03-16 Japanese priority filing and about two months after the US filing date context. A single-inventor corporate assignment executed near filing is unremarkable.
• No "inventor-departure" pattern is observable. With one inventor and no additional personnel data in the record, there is nothing to support the "all inventors departed within 12 months" fire-sale precursor. I found no evidence either way about Abe's tenure.
• ⚠️ Internal tension to flag: Google Patents lists two original assignees — Hitachi Displays, Ltd. and IPS Alpha Technology, Ltd. — but the only recorded inventor→company assignment (Reel 021037/0692) names Hitachi Displays alone. The later "company split" entry (Reel 027063/0019) transferring a 50% share out of Hitachi Displays suggests Hitachi Displays held the full share as of 2010. The origin of any IPS Alpha Technology interest is not explained by the recorded assignments. I have not harmonized these; treat the co-ownership question as unresolved.

Original assignee

• Hitachi Displays, Ltd. (Japan), with IPS Alpha Technology, Ltd. listed as an additional original assignee on the face of the Google Patents record.
• Primary line of business: TFT-LCD panel design and manufacture (Hitachi's flat-panel display arm). IPS Alpha Technology, Ltd. was a January 2005 joint venture among Hitachi Displays, Toshiba, and Matsushita Electric Industrial, formed to manufacture and sell TFT-LCD panels for televisions (source: consolidated LCD antitrust complaint, N.D. Cal. MDL 3:07-md-1827, ¶60).
• Did they ship a product embodying the claims? Yes — the claims cover an LCD panel/module substrate-and-sealant geometry; Hitachi Displays / IPS Alpha were volume LCD panel makers. This is a genuine product-company origin, not a paper filing.
• Current status:
• Hitachi Displays, Ltd. → renamed Japan Display East, Inc. (eff. 2012-04-01) → Japan Display, Inc. / JDI (eff. 2013-04-01). JDI is a listed but financially distressed operating company (via disclosure: 650 billion yen owed to controlling shareholder Ichigo Trust; net assets declining; Mobara fab production ending by March 2026).
• IPS Alpha Technology → IPS Alpha Support Co., Ltd. → merged into Panasonic Liquid Crystal Display Co., Ltd. (eff. 2010-10-01). Panasonic's LCD business later assigned its side of the patent to Panasonic Intellectual Property Corporation of America (2023).
• Net: the original operating owners, and the patent's practice, have both been wound down/restructured. The patent now sits in a monetization vehicle (below).

Assignment timeline

Chronological, using the Google Patents legal-event record. Correspondent is not verifiable in this environment for any entry.

1. 2008-05-15 (executed) / recorded 2008-05-27 — Reel 021037/0692

   • Conveyance: Assignment
   • Assignor: Abe, Hideaki
   • Assignee: Hitachi Displays, Ltd. (Japan)
   • Correspondent: not retrievable — see access note.
   • Context: ordinary inventor-to-employer assignment of rights (initial title transfer).

2. 2010-06-30 (executed) / recorded 2011-10-14 — Reel 027063/0019

   • Conveyance: Assignment (recorded as "COMPANY SPLIT PLAN TRANSFERRING FIFTY (50) PERCENT SHARE OF PATENTS")
   • Assignor: Hitachi Displays, Ltd.
   • Assignee: IPS Alpha Support Co., Ltd. (Japan)
   • Correspondent: not retrievable.
   • Context: internal corporate reorganization (company split) moving a 50% undivided interest; recorded ~16 months after execution.

3. 2010-10-01 (executed) / recorded 2011-10-14 — Reel 027063/0139

   • Conveyance: Merger
   • Assignor: IPS Alpha Support Co., Ltd.
   • Assignee: Panasonic Liquid Crystal Display Co., Ltd. (Japan)
   • Correspondent: not retrievable. (Reels 027063/0019 and 027063/0139 are consecutive frames from the same recording batch — same correspondent highly likely, but unverified.)
   • Context: internal reorganization — merger of the IPS Alpha entity into Panasonic's LCD company.

4. 2012-04-01 (effective) / recorded 2023-11-17 — Reel 065614/0223

   • Conveyance: Change of Name
   • Assignor: Hitachi Displays, Ltd. → Assignee: Japan Display East, Inc.
   • Correspondent: not retrievable.
   • Context: name change only — no change in beneficial ownership. Recorded ~11 years late (catch-up filing cluster).

5. 2013-04-01 (effective) / recorded 2023-11-17 — Reel 065614/0644

   • Conveyance: Change of Name
   • Assignor: Japan Display East, Inc. → Assignee: Japan Display, Inc.
   • Correspondent: not retrievable.
   • Context: name change only. Recorded ~10.5 years late.

6. 2013-04-17 (effective) / recorded 2023-11-17 — Reel 065654/0250

   • Conveyance: Change of Address
   • Assignor/Assignee: Japan Display, Inc. (same entity)
   • Correspondent: not retrievable.
   • Context: administrative only — corporate address update, no transfer of interest.

7. 2023-08-28 (effective) / recorded 2023-11-17 — Reel 065615/0327

   • Conveyance: Assignment — Nunc Pro Tunc
   • Assignor: Panasonic Liquid Crystal Display Co., Ltd. → Assignee: Panasonic Intellectual Property Corporation of America (California)
   • Correspondent: not retrievable.
   • Context: internal transfer of the Panasonic-side interest into its US IP-holding company; recorded together with entries 4–6 in a single 2023-11-17 hygiene batch.

8. 2025-06-25 (executed) / recorded 2025-07-10 — Reel 071890/0202

   • Conveyance: Assignment ("ASSIGNMENT OF ASSIGNOR'S INTEREST")
   • Assignor: Japan Display Inc. → Assignee: Magnolia Purple Corporation (Japan)
   • Correspondent: not retrievable. This is the entry whose correspondent matters most — if a single attorney/firm appears here, it is the likely candidate for a repeat-player role across the Magnolia Purple/White/Blue family. I could not obtain it.
   • Context: transfer to an IP-monetization vehicle. Per JDI's own 2025-06-26 disclosure, Magnolia Purple Inc. was incorporated 2025-05-19 in Minato-ku, Tokyo; capital JPY 10,000; business description "acquisition, management, operation, procurement, and licensing of intellectual property rights"; representative Koichi Irie; formed to receive LCD/OLED patent assets by contribution-in-kind, then have 100% of its shares sold to Ichigo Trust (JDI's controlling shareholder) with a royalty-free license back to JDI. Share sale executed 2025-07-30.

Also on the record but NOT an assignment: a Certificate of Correction recorded 2011-09-13 (no reel/frame surfaced). It corrects the printed patent and is not a title document.

Timeline diagram

timeline
    title Ownership of US 7791704
    2007 : Japanese priority filed
    2008 : US application filed
         : Inventor Abe assigns to Hitachi Displays
    2010 : Patent granted
         : 50 percent share split to IPS Alpha Support
         : IPS Alpha Support merges into Panasonic LCD
    2012 : Hitachi Displays renamed Japan Display East
    2013 : Renamed Japan Display Inc
    2023 : Panasonic LCD interest to Panasonic IP Corp America
         : Late name change recordings filed in one batch
    2025 : Japan Display sells patent to Magnolia Purple
         : Magnolia Purple stock sold to Ichigo Trust

NPE / troll-pattern signals

1. Shell-entity transfer — PRESENT. The patent moved from an operating display maker (Japan Display Inc.) to Magnolia Purple Corporation, an entity created days before the transfer whose sole stated business is "acquisition, management, operation, procurement, and licensing of intellectual property rights," capitalized at JPY 10,000, with one representative director and no manufacturing. Evidence: Reel 071890/0202 (exec. 2025-06-25 / rec. 2025-07-10) plus JDI's 2025-06-26 disclosure. Caveat: this is a Japanese IP-holding subsidiary with a license-back to JDI, not a classic anonymous Delaware/Texas single-member LLC at a registered-agent address — so the "shell" characterization is well-supported on the no-products/licensing-only dimension but differs in form from the textbook US NPE shell.

2. Known asserter in the chain — NOT PRESENT. None of the assignees (Hitachi Displays, IPS Alpha Support, Panasonic LCD, Panasonic IP Corp of America, Japan Display, Magnolia Purple) match Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, or Spangenberg entities, and Magnolia Purple did not surface on any RPX/Unified asserter list I could reach. (False-positive warning: search hits for "Magnolia Medical Technologies" and "Purple LLC / Purple Innovation" are unrelated blood-collection and mattress companies, respectively — not this chain.)

3. Repeat correspondent across the chain — UNCLEAR. Cannot be assessed: the Assignment Center correspondent/recording-attorney fields were not obtainable for reels 021037/0692, 027063/0019, 027063/0139, 065614/0223, 065614/0644, 065654/0250, 065615/0327, or 071890/0202. Note that reels 027063/0019 and 027063/0139 are adjacent frames from one batch, and the four 2023-11-17 entries (065614/0223, 065614/0644, 065654/0250, 065615/0327) share a recording date — consistent with one filer handling each batch — but I have no name to cite, so this is not a finding.

4. Cascading transfers — NOT PRESENT (as defined). There is no chain of consecutive assignments through chained LLCs within 24 months. What exists is a 2023-11-17 recording cluster of four long-delayed instruments (name changes effective 2012 and 2013, a change of address, and a 2023 nunc pro tunc assignment) — i.e., record-hygiene clean-up ahead of the later sale, not an LLC cascade. Then a single 2025 transfer. Flag the cluster as interestingly timed relative to the 2025 divestiture, but it does not meet the "multiple LLCs in <24 months" test.

5. Pre-litigation transfer — UNCLEAR / NOT PRESENT for this patent. The transfer to Magnolia Purple (2025-06-25) is recent (~15 months before today), which is the kind of timing that can precede assertion. However, no infringement suit, ITC action, or IPR naming US 7,791,704 was found, so the transfer cannot be tied to a first suit. Marked unclear rather than present.

6. Bankruptcy fire-sale — NOT PRESENT (as defined). No Chapter 7/11 filer appears in the chain. Context worth noting: JDI is severely distressed (650 billion yen owed to Ichigo Trust; declining net assets) and divested these patents to repay debt to its controlling shareholder — economically a distressed-asset sale, but executed as a private share sale, not a bankruptcy estate sale.

7. Privateering — UNCLEAR. The structure (operating company contributes patents to new IP subsidiaries, sells 100% of those subsidiaries' shares to its controlling financial shareholder, takes a royalty-free license back) is the anatomy of patent privateering/monetization-by-proxy. What is missing is the defining act: no evidence that Magnolia Purple (or Ichigo Trust) has asserted these patents against JDI's competitors. On the current record this is a monetization vehicle with latent assertion potential, not a demonstrated privateering campaign. This is the single most important open question — a first Magnolia Purple complaint would convert this signal to PRESENT and would likely move the verdict to high confidence.

8. Defensive aggregator — NOT PRESENT. The chain terminates at Magnolia Purple Corporation, not at RPX, AST, LOT Network, Unified Patents, or OIN. The patent has not been neutralized.

Verdict

NPE — moderate confidence.

Two things are established by record evidence rather than naming inference. First, the current owner of record is Magnolia Purple Corporation per Reel 071890/0202 (executed 2025-06-25, recorded 2025-07-10), an entity whose own corporate disclosure describes a licensing-only IP business with JPY 10,000 capital and no products — the "shell-entity transfer" signal is PRESENT. Second, the immediately preceding owner, Japan Display Inc., is an operating company that divested the patent under financial duress to a vehicle whose equity it sold to its controlling shareholder, Ichigo Trust, retaining only a license-back. That is 1 strong signal plus meaningful monetization context, which maps to moderate rather than high confidence. No litigation naming this patent, no known-asserter match, and no defensive-aggregator endpoint were found, and the correspondent-of-record field that would let me test the "repeat lawyer" signal could not be retrieved — so I stop short of high confidence.

Verify at: https://assignmentcenter.uspto.gov/ (search patent 7,791,704; also indexed at https://assignment.uspto.gov/patent/index.html). Retrieving the correspondent name/firm on reel 071890/0202, and checking whether that same correspondent appears across the Magnolia Purple/White/Blue filings, is the highest-value next step and would resolve signal 3.

Key caveats: (a) all reel/frame numbers derive from the Google Patents legal-events mirror, not a direct PatentCenter read; (b) the correspondent field is unavailable and is not a finding that no correspondent exists; (c) the possible joint Hitachi Displays / IPS Alpha Technology original ownership is unexplained by the recorded assignments; (d) the 2023-11-17 recording cluster's legal significance (pure housekeeping vs. sale preparation) is an inference, not a verified fact.

I'll search for the specific patent and its cited references.

Let me get details on the two cited Japanese references.

US 7,791,704 — Prior-Art Identification and § 102 Mapping

Cross-reference note: This builds on the earlier bibliographic/history section. Nothing in this task contradicts that section; the two examiner-cited references already listed there are the entirety of the "Patent Citations" of record, which I confirm again below. I have not repeated the bibliographic table, claim summary, or chain-of-title.

1. Identification of the document (literally "7791704")

I did not auto-correct the number. As noted in the earlier section, numeric near-matches in the search corpus (e.g., US 8,469,704 in Ormco v. Lancer Orthodontics; US 9,724,704; the Brazilian notary document listing a personal ID 7791704) are different identifiers and were not merged into this analysis.

Verification caveat: the "Patent Citations" list below is taken from the Google Patents record that mirrors the USPTO file ("Patent Citations (2)," https://patents.google.com/patent/US7791704/en). I could not directly retrieve the USPTO PatentCenter/"References Cited" page or the file-wrapper transaction history in this session. The U.S. pre-grant publication US 2008/0225221 A1 is in the same family and is not, of course, prior art against US 7,791,704.

2. Citation 1 — JPH08129166 A

Full citation: JP H08-129166 A (Japanese Patent Laid-Open Hei 8-129166), "Liquid crystal display device" (液晶表示装置), Sanyo Electric Co., Ltd. — priority/filing 1994-10-31, published 1996-05-21. Cited by the examiner; not discussed by name in the US 7,791,704 specification.

Brief description (from the fetched reference text, https://patents.google.com/patent/JPH08129166A/en): A COG-type liquid crystal display device in which a first insulating substrate and a second insulating substrate of different size are bonded via a seal, with liquid crystal injected. The non-superposed region carries the bare chip (chip 13, face-down or wire-bonded) and the flexible cable. Its express point of novelty is the substrate profile: "The lateral side L of the second insulatable substrate 11 is formed to a nonlinear shape. Then, if, for example, the device is provided with the two chips 13, recessed parts 15 corresponding to the number are formed and both sides of these recessed parts 15 are effectively utilized as the display region." Its background section also criticizes the prior art practice of forming flexible substrates "in a rectangular shape by cutting straight lines."

§ 102 analysis: This reference is materially closer to the invention's substrate-profile idea than its brief "cited by examiner" treatment suggests, because it discloses (i) two substrates of unequal size, (ii) a liquid crystal layer sealed therebetween, (iii) a non-linear (recessed) lateral profile of the upper substrate as viewed from above, and (iv) a chip and flexible cable occupying the non-superposed region. Those elements map onto the preamble plus limitations (c) and (d) of claim 1, and onto the subject matter of dependent claims 4–6 (driver chip and flexible printed circuit board on the non-overlapping region of the lower substrate).

However, on the text retrieved, it does not disclose the two limitations that the granted claim 1 makes mandatory: the sealing material formed along the shape of the upper substrate including the recess, and the sealing material being wider at the recess-following portion than at a non-recess portion. The seal (seal 3) is described only as bonding the substrates and enclosing the inject port. Consequently I do not assess JPH08129166A as anticipating any of claims 1–6 under 35 U.S.C. § 102 — every claim depends from claim 1, which requires both limitations. Its proper role is as § 103 art against the substrate-shape and COG/FPC-placement aspects.

3. Citation 2 — JPH08171095 A (the specification's "patent document 1")

Full citation: JP H08-171095 A (Japanese Patent Laid-Open Hei 8-171095), "Liquid crystal display device and production of liquid crystal display device" (液晶表示装置およびその製造方法), Sanyo Electric Co., Ltd. — priority/filing 1994-12-16, published 1996-07-02. This is the reference the US 7,791,704 specification names as "patent document 1."

Brief description: A liquid crystal display device (and its manufacturing method) in which the second (upper) substrate is arranged on the first (lower) substrate and one side of the second substrate is formed in a recessed shape as viewed from above. According to the US 7,791,704 specification's own characterization of this document: "With respect to the constitution of the liquid crystal display module described in the previously mentioned patent document 1, the second substrate is arranged on the first substrate, and one-side surface of the second substrate is formed in a recessed shape. However, the sealing material arranged between the first substrate and the second substrate is formed in a rectangular frame shape as shown in FIG. 11. That is, the sealing material described in patent document 1 is not formed along a shape of the upper substrate different from the present invention."

§ 102 analysis: By the patentee's own admission, JPH08171095A discloses limitations (a)–(d) of claim 1 substantially in full — the recessed upper substrate, the liquid crystal layer between the two substrates, and the FPC/driver-chip region on the exposed lower-substrate ledge — but discloses the sealing material as a plain rectangular frame, i.e., not formed along the upper substrate's shape and not wider at a recess. It therefore does not anticipate claim 1 or any of dependent claims 2–6 under § 102. It is the single most relevant prior-art reference for this patent: it is the reference the claimed invention is expressly framed against, and limitation (e)/(f) of granted claim 1 appears to exist precisely to carve out the difference over it. Claims 2–6 (plastic/glass substrate; seal filling the overlap regions flanking the FPC with no gap; driver chip on the lower substrate; parallel FPC/chip arrangements) each add further features not shown in this reference's description as characterized.

Honest limitation of this analysis: I was unable to retrieve the full text/abstract of JPH08171095A directly in this session — the searches returned the US 7,791,704 record's citation entry for it rather than the reference document itself. The description above for this reference therefore rests on (i) the bibliographic citation entry and (ii) the US 7,791,704 specification's own discussion of "patent document 1," which is the authoritative text supplied. A complete § 102 read would require the actual JP H08-171095 A disclosure (including its drawings), which I could not verify here.

4. Additional references considered in the family (not U.S. citations of record)

The US record separately lists three "Family Cites Families" documents — i.e., art cited in the CN 101266367 / JP family members, not confirmed as having been cited against the U.S. claims:

• JP 2003-280028 A — Seiko Epson Corp., "Electro-optical device and electronic equipment," priority 2002-03-25, published 2003-10-02.
• JP 2004-272224 A — Sony Corp., "Liquid crystal display device and manufacturing method thereof," priority 2003-02-18, published 2004-09-30.
• JP 4552637 B2 — Epson Imaging Devices Corp., "Liquid crystal display," priority 2004-12-09, granted 2010-09-29.

I retrieved only bibliographic data for these three; I could not obtain their technical content in this session, so I make no § 102 or § 103 assessment of them. They are nonetheless the candidates most likely to be relevant to the "substrate cut-out / sealant routing" subject matter, and a full prior-art study should read them.

The four documents in the "Cited By" table — US 2013/0135830 A1 and US 8,982,575 B2 (Au Optronics), US 9,791,753 B2 (E Ink Holdings), and US 12,525,523 B2 (Wuhan China Star Optoelectronics) — all postdate the 2007-03-16 priority date and are therefore not available as prior art against US 7,791,704. They are forward citations only.

5. Statutory basis and conclusion

• Effective pre-AIA regime. US 12/073,943 was filed 2008-03-12, before the 2013-03-16 AIA first-to-file cutoff, so pre-AIA § 102 governs. Both JP publications issued in 1996 — more than one year before the earliest relevant U.S. date (2007-03-12, i.e., one year before the U.S. filing date) — so both are § 102(b) printed publications, and a fortiori qualify under § 102(a)/(e) (both are by another — Sanyo Electric — and predate the 2007-03-16 JP priority date). Neither is applicant's own work.
• Most relevant prior art: JPH08171095 A (the specification's "patent document 1"), because it is the reference from which the granted independent claim is expressly distinguished, followed by JPH08129166 A for the recessed/non-linear substrate profile and COG-region chip + flexible-cable placement.
• Anticipation outcome: On the material available to me, neither cited reference anticipates any of claims 1–6 under § 102, because claim 1 requires (i) the sealing material formed along the upper substrate's shape including the recess, and (ii) the seal being wider at the recess-following portion than elsewhere — and neither reference appears to disclose either limitation. The two references are best characterized as § 103 combination/elementary-art references directed at the recessed upper substrate and at placing the driver chip and FPC on the non-overlapping ledge.

6. Items I could not verify (stated explicitly rather than assumed)

1. The official USPTO "References Cited" / file-wrapper contents — my citation list derives from the Google Patents mirror of the US 7,791,704 record, which lists exactly two cited patents, both flagged "cited by examiner."
2. The full text, abstract, and drawings of JPH08171095 A were not retrieved; its description here rests on the US patent's own characterization plus its citation entry.
3. The technical content of the three family-cited JP documents (2003-280028, 2004-272224, 4552637) was not retrieved.
4. Whether limitation (e)/(f) of claim 1 was added by amendment during prosecution — the earlier section flagged this as an inference from comparing the abstract to the granted text; I did not obtain the transaction history to confirm it, so I do not treat it as established here.

I'll research the prior art references on this patent's record to ground the § 103 analysis.

Let me retrieve the full text of the two exam-cited Sanyo references.

Let me pull the full text of JPH08171095A and the three "family cites" references.

Obviousness Analysis — U.S. Patent No. 7,791,704 B2

"Liquid crystal display panel" — 35 U.S.C. § 103

⚠️ Scope and sourcing note. Per instruction, this analysis is built from the Prior Art section of the US 7,791,704 record — i.e., the two references cited on the face of the patent (JPH08129166A and JPH08171095A, both Sanyo) and the three references appearing in the "Family Cites Families" list (JP2003280028A, JP2004272224A, JP4552637B2). I was able to retrieve the full English text of JPH08129166A, and the full Japanese text of several Sanyo seal-related publications, but I was unable to retrieve the full text of JPH08171095A, JP2003280028A, JP2004272224A, or JP4552637B2 in this session. Where I rely on those, I say so and rely only on their titles and on the '704 specification's own characterization — I do not fabricate disclosure content. Consistent with the earlier sections, I also flag that I could not verify the prosecution file wrapper (transaction history), so the reasons the "wider sealing material" limitation was added remain an inference.

1. Governing law and date framework

The application was filed 2008-03-12 with a JP priority of 2007-03-16, so pre-AIA § 103(a) applies, as construed in KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), and Graham v. John Deere. Key consequences:

• No rigid TSM requirement. A reason to combine may come from the prior art, the problem to be solved, or "common sense"; predictable variations of known elements are obvious.
• "Obvious to try" is available where there is a finite number of identified, predictable solutions.
• Both Sanyo references published in 1996 and therefore qualify as § 102(b) prior art (more than one year before the 2007-03-16 priority date) — no priority/date contest. Because they are Japanese publications, an English translation would be required as an exhibit (37 C.F.R. § 1.98), but the substantive art date is not in doubt.
• § 103(c) common-ownership safe harbor is unavailable — the Sanyo references are not commonly owned with Hitachi Displays/IPS Alpha.

Flag for correctness: the family-cited references appear in the family prosecution record, not in the "Patent Citations (2)" list on the U.S. face of the patent. They are therefore secondary authority here; the two Sanyo references are the references the U.S. examiner actually relied on.

2. Level of ordinary skill in the art

A bachelor's degree in electrical engineering, mechanical engineering, materials science, or applied physics, or equivalent, plus approximately 2–3 years of experience in LCD panel/module design (substrate layout, sealant dispensing, COG/COF mounting), or a master's degree with less experience. This is a mature, incremental mechanical/assembly art by 2007 — a fact that weighs in favor of obviousness under KSR ("a court must ask whether the improvement is more than the predictable use of prior art elements according to their established functions").

3. Construction of the three limitations that carry the case

(a) "the upper substrate has one-side surface thereof formed in a recessed shape as viewed from above."
Read against the '704 spec and FIG. 5A/5B, the "recessed shape" is a plan-view notch in the outline of the upper (CF) substrate — the CF substrate overlies the rectangular TFT substrate but with a notch on the FPC side, so that the CF substrate still overlaps the four corners of the TFT substrate (spec: "the upper substrate is arranged to overlap with four corners of the lower substrate"). It is not a depression in a major face. ⚠️ The literal phrase "one-side surface … recessed" is grammatically imprecise and is a claim-construction risk; I flag it rather than harmonize it.

(b) "the sealing material is formed along a shape of the upper substrate that includes the recessed shape."
The seal path must deviate inboard to follow the notch, i.e., it is not a plain rectangular frame.

(c) "a width of the sealing material at a first portion that is formed along the recessed shape is wider than a width of a second portion of the sealing material that is not formed along the recessed shape."
A relative-width (quantitative-by-comparison) limitation, with no numerical bound.

4. Element-by-element mapping — JPH08129166A (Sanyo)

Retrieved text: https://patents.google.com/patent/JPH08129166A/en

Conclusion: JPH08129166A discloses every element of claim 1 except the relative seal-width limitation. On its face, that is a prima facie case of obviousness (and arguably of § 102 anticipation as to the disclosure, absent the width limitation).

Notably, JPH08129166A's background section also recites the same deficiency the '704 patent attacks — rectangular, straight-cut substrates and the wasted/structurally exposed non-overlap region — and its own solution is to make at least one side non-linear and to run the seal non-linearly. That is precisely the '704 concept.

5. Grounds of rejection

Ground 1 — Claim 1 obvious over JPH08129166A alone

• Difference: the relative seal-width limitation.
• Rationale (KSR — "predictable variation / result-effective variable"): where the seal must negotiate a step or turn to follow a non-linear substrate edge, widening the bead at that location is the inherent and predictable result of dispenser-applied sealant (flow rate is constant while traverse speed drops at turns, and the step region must be bridged without a void). Widening the seal there is a design choice within the ordinary skill level, producing no unexpected result.
• Reinforcement from the reference itself: JPH08129166A claim 2 and claim 6 teach surround/backfill concepts — "a flow stopper made of the same material as the seal is provided so as to surround the semiconductor element" (claim 2) and "[t]he arrangement area is surrounded by the sealing material" (claim 6) — i.e., the art already contemplates locally augmenting seal material in the non-overlap/notch region. This supplies an express reason to make the seal material there larger, not smaller.

Ground 2 — Claim 1 obvious over JPH08129166A in view of JPH08171095A

• JPH08171095A is the spec's own "patent document 1": per the '704 specification, it discloses "the second substrate is arranged on the first substrate, and one-side surface of the second substrate is formed in a recessed shape," but with the "sealing material … formed in a rectangular frame shape."
• Motivation: both references are Sanyo LCD-device references in the identical field, addressing the same substrate-size-asymmetry/seal-placement problem; a POSITA would consult one with the other. JPH08171095A supplies the recessed upper substrate concept (and its recognized benefit of exposing the mounting ledge); JPH08129166A supplies the conformal, non-linear seal and the FPC/driver placement in the non-overlap region. Combining them yields claim 1 as a matter of predictable use of known elements according to their established functions.
• ⚠️ Caveat: I could not retrieve JPH08171095A's text, so I cannot confirm whether it independently discloses the seal-width differential. I would not assert that without the document.

Ground 3 — Dependent claims obvious over JPH08129166A (with or without JPH08171095A)

6. Motivation to combine — articulated reasons

1. Same field, same problem, same solution family. Both Sanyo references and the '704 patent address the same structural problem: an asymmetric-substrate LCD panel whose single-substrate ledge is exposed and weak. JPH08129166A's background states the identical premise (rectangular, straight-cut substrates; non-overlap region "could not be effectively utilized").
2. Express suggestion in the primary reference. JPH08129166A does not merely suggest — it claims the non-linear upper-substrate side (claim 1) and the non-linear seal side (claim 4). A POSITA seeking to protect the notch/ledge would naturally carry the seal along that non-linear edge.
3. Predictable result. Following a part's outline with a seal is the ordinary function of a conformal seal; the whole point of a seal is to bond the substrates wherever they overlap, which mechanically requires the seal to follow the notched outline rather than cut across it.
4. Known benefit of widening/backfilling at the notch. JPH08129166A's flow-stopper/surround teachings (claims 2, 6) make it natural, not inventive, to locally increase seal material where the driver/FPC sits.
5. Ordinary engineering optimization. The remaining layout choices (chip + FPC parallel along one edge) are the routine, well-known COG/COF arrangement, with results (narrower frame, shorter interconnect) that are foreseeable and expressly acknowledged as such in the specification.

7. The weak link — the "sealing material is wider at the recess" limitation

This is where the obviousness case is narrowest, and I will not overstate it:

• It is not expressly disclosed in JPH08129166A or (as far as I can verify) elsewhere on the record. It also is not reflected in the '704 abstract or description, which describe only that "a sealing material is formed along a shape of the CF substrate" — consistent with (though not proof of) a prosecution amendment added to secure allowance. Because I could not obtain the file wrapper, this remains an inference; if verified, the amendment and any applicant arguments would shape both validity and any later estoppel.
• Best available challenge routes:
  (a) Inherency / result-effective variable — seal width is a function of dispense rate and traverse speed; turns in a non-linear seal path inherently produce a wider bead. This should be supported with evidence (e.g., sealant-dispensing literature or an expert declaration) rather than asserted.
  (b) Design choice — the seal must bridge the step where the upper substrate steps inboard to form the notch; widening it there to avoid a void is a predictable mechanical necessity, which also dovetails with claim 3.
  (c) Non-record art, flagged explicitly. My searches surfaced JP2002268072A (Sanyo) — a recess (凹部 27) formed at the seal-application start position so the accumulated seal bead (シール玉 26a) is accommodated and the seal width does not balloon. It confirms that varying/controlling seal width at a local substrate feature was known in this art, but it is NOT on the '704 record, and it points the opposite direction (suppressing width growth). It is a double-edged reference: useful to show seal-width-at-a-locale was a known, controllable design variable; unhelpful if used to show a desired widening.
• Practical takeaway: absent the file wrapper, I would expect the "wider at the recess" clause to be the limitation the patentee leans on to survive a § 103 challenge. The rejection is strong on all other elements and on claims 2–6, but on claim 1 it turns on whether the relative-width clause is entitled to weight beyond an obvious design choice.

8. Secondary (Graham) considerations

No evidence of secondary indicia appears on the record: no unexpected results, no teaching-away, no long-felt-but-unsolved need, no failure of others, no copying, no commercial-success nexus. JPH08129166A is not taught away from the '704 approach — if anything it teaches toward it. The 2013–2022 forward citations (Au Optronics, E Ink Holdings, Wuhan China Star) show the area remained active but do not establish unexpected results.

9. Bottom line

Confidence: high on the claim-to-reference mapping for JPH08129166A (full text retrieved and quoted); moderate-to-low on JPH08171095A, JP2003280028A, JP2004272224A, and JP4552637B2 (titles and '704's characterization only — full texts not retrieved). The claim-1 width limitation is the decisive open issue; resolving it would require the USPTO file wrapper (transaction history / amendment and remarks) and, ideally, an English translation plus full text of JPH08171095A. I flag that I could not verify the file wrapper, and I have not auto-corrected any reference number: JPH08129166A and JPH08171095A are distinct Sanyo publications, neither to be confused with the similarly numbered U.S. patents (e.g., 8,469,704; 9,724,704) noted in the earlier sections.