US 8965932

I'll research US Patent 8965932 across USPTO and CAFC sources.

Let me search for the CAFC 2026 docket and the actual claim text.

Let me retrieve the actual claim text and confirm the CAFC docket status.

US Patent 8,965,932 — Research Summary

Method note: I ran web searches across Google Patents, Justia, PTAB/Unified Patents portals, and docket aggregators. I did not have direct authenticated access to USPTO PatentCenter/PAIR or the CAFC docket system, so docket-status items below are drawn from third-party aggregators and should be verified against the official records. I treat the patent-number string "8965932" literally and did not substitute any similar number.

1. Bibliographic data

2. Abstract (verbatim)

"A method of sequencing and scheduling web resources, via a software application that collects URLs and feeds them to a Web browser, so that the amount of clicks and decisions are reduced when browsing the Internet. The method begins after a computer user with Net-access encounters a URL that leads to a collection of other URLs that the software application has arranged to be presented in a continuous, show manner. This collection of resources then plays automatically, going from one resource to the next without the user being required to make a decision or click. The decision-making process is not inhibited, it is made optional. A method of inserting full-screen advertising, public service announcements, news, etc. has been developed. The server software application module described herein allows a publisher to stream URLs to a user who does not have the client-side software. The publishing module allows a user to format resources to a particularly desirable size and characteristic that promotes readability during a show viewing session. This same module can publish collections of nodes for others to view as a whole sequence, accessible via e-mail, screen saver or as a web resource."

3. Plain-language overview of the claimed subject matter

Independent claim 1 (as reproduced in the pre-grant publication US 2006/0218479 A1, which corresponds to this application):

1. A method for displaying on a user's computer, content derived from a plurality of resources in an organized arrangement comprising the steps of:
   • creating a show structure of nodes, each node identifying a resource from a plurality of accessible resources;
   • accessing each of said resources identified by each of said nodes;
   • retrieving a content corresponding to each of said accessed resources; and
   • displaying each of said retrieved contents automatically in accordance with said show structure.

In ordinary language, the independent claim covers building a linked "show structure" (a set of nodes, each pointing at a resource such as a web page, executable file, or data file), pulling the content at each node, and then auto-playing that content in the order defined by the structure — so the user is not required to click from page to page. The specification describes this suite of programs under the name RoboCast (RoboSurf client plug-in, RoboSites server module, RoboGuide personalized guide, RoboPublisher authoring tool), including timed display durations per node, timer-driven advancement, "surf" suspend/resume, node arrays with sequentially and/or concurrently traversed paths, caching of the next resource, personalized filtering via user profiles (Firefly/Automated Collaborative Filtering), and interspersed advertising/PSAs.

Uncertainty flag: The full granted claim set of the '932 patent is longer than the single claim shown in the pre-grant publication I retrieved, and some of it was not reproduced in the sources returned. I therefore cannot state with authority how many independent claims the grant contains, their exact language, or the precise dependent-claim scope. Someone needing claim-level fidelity for claim charts should pull the issued patent's claims directly from PatentCenter or the patent PDF.

4. Litigation and PTAB activity involving 8965932

• IPR2022-01125 — Unified Patents, LLC v. Robocast, Inc., filed June 30, 2022, challenging U.S. Patent 8,965,932 (Respondent Application No. 11,316,209; Tech Center 2100). Institution denied (decision Jan. 20, 2023).
• IPR2023-00592 — Google LLC et al. v. Robocast, Inc., filed Feb. 21, 2023. Institution denied (decision Sept. 18, 2023).
• IPR2023-00593 — filed Feb. 21, 2023, listed as not instituted (merits).
• Delaware District Court cases 1:22-cv-00304 and 1:22-cv-00305 (Robocast enforcement actions, per Unified Patents litigation data).
• CAFC: Google Patents lists a Federal Circuit case, 25-1651, associated with this family. Docket number format suggests a 2025 filing.

On the specific "CAFC 2026 dockets" request: I found no 2026 Federal Circuit decision, opinion, or merits ruling for 8,965,932 in the sources searched. I could only confirm the existence of the Federal Circuit case identifier 25-1651 (per Google Patents/Unified Patents litigation data) and no proof that it involves a 2026 outcome for this patent. I do not have authoritative information that a 2026 CAFC docket entry for 8965932 exists — if one is needed, it should be checked directly in the CAFC docket (PACER/CM-ECF, appeal no. 2025-1651 or the plaintiff's counsel's filings), because my search tools did not surface any 2026 appellate docket for this patent.

Earlier prosecution context worth noting: the same family was asserted in Robocast, Inc. v. [Microsoft Corp.](/litigations/by-plaintiff/Microsoft%20Corp.) and Robocast, Inc. v. Apple Inc., where a claim-construction order dated June 28, 2013 construed terms of the '451 parent patent (cited as Ex. 1011 in the Unified Patents IPR petition).

5. Bottom line

US 8,965,932 is a Robocast, Inc. patent (sole inventor Damon Torres) claiming a 1996-priority way to auto-sequence and display web/network resources as a continuous "show," issued Feb. 24, 2015 on an application filed Dec. 21, 2005, now expired (fee related) with an adjusted expiration of Dec. 10, 2020. It survived three PTAB challenges (two IPRs against it denied institution in Jan. 2023 and Sept. 2023). The most recent signal is a Federal Circuit case number 25-1651, but I found no confirmed 2026 CAFC action — treat any 2026 appellate activity as unverified until checked against the official docket.

I'll research litigation involving US patent 8965932 using multiple sources.

Let me get details on the outcomes of these cases and the Federal Circuit appeal.

Litigation Involving U.S. Patent No. 8,965,932

Important scoping note: U.S. 8,965,932 ("Automated content scheduler and displayer") issued February 24, 2015 — well after Robocast's earlier suits against Microsoft (D. Del. 1:10‑cv‑01055) and Apple (D. Del. 1:11‑cv‑00235) were filed. Those earlier campaigns asserted the parent '451 patent, not the '932 patent. I found no litigation asserting the '932 patent prior to 2022. The '932 patent was asserted only in the two 2022 Delaware actions and the resulting Federal Circuit appeal, all of which also involved the sibling patents '451 and '819 (all three share a common priority claim to Sept. 3, 1996).

1. Robocast, Inc. v. Netflix, Inc.

Additional procedural history: Netflix's partial motion to dismiss was granted-in-part on Nov. 14, 2022 (dismissing indirect/willful and vicarious-liability claims, leaving direct infringement only). Netflix also alleged the asserted patents were unenforceable due to inequitable conduct (a false Torres declaration in the parent '063 application). Netflix styled its counterclaims in part as declaratory judgments of non-infringement, invalidity, and unenforceability of the '932 patent.

2. Robocast, Inc. v. YouTube, LLC and Google LLC

3. Robocast, Inc. v. Netflix, Inc. (Federal Circuit)

Related PTAB (IPR) Proceedings on the '932 Patent

These are administrative validity challenges, not district-court litigation, but they are directly tied to the '932 patent's enforcement:

(Netflix separately filed its own IPR petitions directed at the '451 and '819 patents, IPR2023‑00081 and IPR2023‑00182, not the '932 patent.)

Summary

• Two district-court cases asserted the '932 patent: Robocast v. Netflix (D. Del. 1:22‑cv‑00305, filed 3/7/2022) and Robocast v. YouTube/Google (D. Del. 1:22‑cv‑00304, filed 3/7/2022).
• Netflix won summary judgment of § 101 invalidity on all asserted claims (Feb. 21, 2025); the YouTube/Google case is stayed.
• The Netflix judgment is on appeal at the Federal Circuit, No. 25‑1651 (docketed April 17, 2025); final appellate outcome is not confirmed in my retrieved sources.
• Three IPR petitions (Unified Patents IPR2022‑01125; Google/YouTube IPR2023‑00592 and IPR2023‑00593) were denied institution.
• No other litigation asserting U.S. 8,965,932 was found.

Caveat on completeness: My searches relied on Unified Patents' portal, Docket Alarm, CourtListener, Justia dockets, and Law360 secondary reporting. I could not directly query PACER for a complete post‑April 2025 docket history, so any very recent developments (e.g., a Federal Circuit decision, settlement, or dismissal in 25‑1651) are outside what I can confirm.

I'll research the PTAB proceedings referenced in the patent record and verify their outcomes.

Let me get the details on the other two proceedings and any appeals.

Let me confirm the -00592 institution date and the related Delaware/CAFC posture.

PTAB Proceedings on US 8,965,932 ("Automated content scheduler and displayer")

Record note / discrepancy flag. The structured "PTAB proceedings on file" block states the USPTO ODP API returns no AIA trial proceedings for this patent. That is inconsistent with the patent's own public record. The Google Patents page for US8965932 carries three PTAB litigation entries (IPR2023‑00592, IPR2023‑00593, IPR2022‑01125, each labeled "Not Instituted – Merits"), and each is separately corroborated by PTAB docket data and party filings. I therefore report three proceedings and flag the ODP ingest as incomplete rather than treating the absence as meaningful. Proceeding numbers below are taken verbatim from those records — none are invented.

Proceedings overview

There are three AIA trial proceedings on US 8,965,932 — IPR2022‑01125, IPR2023‑00592, and IPR2023‑00593 — and all three were denied institution, before any trial, any Final Written Decision, and any cancellation of claims; there are zero claims invalidated, zero claims sustained on the merits, zero settlements, and zero institutions (3 × institution denied).

Bottom line for a defendant: this is not a hardened patent and not a dead patent — it is an untested one. No PTAB panel has ever reached the merits of any of the '932 patent's 67 claims, so no claim has been canceled by the Board and no statutory estoppel has attached to any challenger. The PTAB shield is unavailable to you. The only merits adjudication of validity on this patent family comes from district court, where the Delaware court held the asserted claims ineligible under § 101 in a 2025 order now on appeal at the Federal Circuit (25‑1651). Any invalidity defense you build should be built in district court on § 101 and § 112 / prior-art theories, not on a PTAB record.

IPR2022‑01125 — Unified Patents, LLC v. Robocast, Inc.

• Type: Inter Partes Review
• Filed: 2022‑06‑30
• Status: Institution Denied (verbatim from source data). Plain English: the Board declined to institute; the proceeding terminated 2023‑01‑20 without a trial.
• Judge panel: Patrick M. Boucher, Nabeel U. Khan, Michael T. Cygan (Boucher writing).
• Claims challenged: 1–9, 13–15, 17, 18, 20, 22–27, and 33–43 (32 of the patent's 67 claims).
• Petition grounds: four grounds under 35 U.S.C. § 103. I could not retrieve the specific reference combinations from the sources available to me; I am not going to guess at them.
• Institution decision: 2023‑01‑20, denied. The Board's language, as quoted by Patent Owner in its public statement, is: "we conclude that the information presented in the Petition does not establish a reasonable likelihood that Petitioner would prevail in showing that at least one of claims 1–9, 13–15, 17, 18, 20, 22–27, and 33–43 of the '932 patent is unpatentable. Accordingly, we deny the Petition and decline to institute an inter partes review." This is a merits-based denial under § 314(a), not a discretionary Fintiv denial. Patent Owner had also argued that (i) the Board lacks jurisdiction because the '932 patent expired on 2020‑12‑10, and (ii) Fintiv discretionary denial — the Board did not need to reach either.
• Final Written Decision: none issued — no trial occurred.
• Settlement / termination: no settlement; the case terminated on the denial. Petitioner's request for refund of post-institution fees was granted 2023‑01‑31.
• Appeal: none. Institution denials are non-appealable (35 U.S.C. § 314(d); Thryv).
• Defensive value: useful as a signal, not a shield. The country's most active defensive aggregator attacked 32 claims with four grounds and lost at the institution stage — that tells you the '932 claims are not low-hanging fruit for a quick § 103 kill at the Board. But it also means nothing in this record estops you or anyone else, and nothing here was adjudicated.

Links: Unified's case portal — https://portal.unifiedpatents.com/ptab/case/IPR2022-01125 · Docket data — https://ipverse.greyb.com/ptab-web/cases/case-details/IPR2022-01125

IPR2023‑00592 — Google LLC and YouTube LLC v. Robocast, Inc.

• Type: Inter Partes Review
• Filed: 2023‑02‑21
• Status: Institution Denied. Terminated 2023‑09‑18 without trial.
• Judge panel: David C. McKone, Patrick M. Boucher, Michael T. Cygan (Boucher writing).
• Claims challenged: claims 1–21 and 63–67 (26 claims; independent claims 1 and 63 are representative).
• Petition grounds (all § 103):
  • Ground 1 — claims 1–11, 14, 16–21 over Shimizu (US 5,634,062), Benedict (Using Harvard Graphics Version 2 for Windows), Fox (HTML Web Publisher's Construction Kit), Wolfe (US 5,715,445)
  • Ground 2 — claims 63–67 over Shimizu, Benedict, Fox, Wolfe, and Sagman (Microsoft PowerPoint 4 for Windows)
  • Ground 3 — claims 11–16 over the Ground 1 art plus Glaser (US 5,793,980)
  • Ground 4 — claim 65 over the Ground 2 art plus Robertson (US 6,486,895 B1)
  • Supporting declaration of Christopher M. Schmandt.
• Institution decision: 2023‑09‑18, denied. The decision states: "Applying the standard set forth in 35 U.S.C. § 314(a), which requires demonstration of a reasonable likelihood that Petitioner would prevail with respect to at least one challenged claim, we deny the Petition and do not institute an inter partes review." Again a merits denial, not Fintiv. Patent Owner had pressed two substantive merits failures: that Shimizu's hypertext nodes cannot be the claimed "nodes" of claims 1 and 63, and that Benedict cannot supply the "interactively variable duration information" limitation; it also argued the Petition was an improper second petition (see ‑00593).
• Final Written Decision: none — no trial.
• Settlement / termination: none.
• Appeal: none (institution denial is non-appealable).
• Defensive value: the Google/YouTube art set (Shimizu + Benedict/Fox/Wolfe) is now public and reusable — you are free to run it in district court, and no § 315(e)(2) estoppel blocks it, because there was no final written decision. But you should assume it is not a winner at the Board: a merits panel already found no reasonable likelihood it prevails.

Links: PTAB portal — https://portal.unifiedpatents.com/ptab/case/IPR2023-00592 · Institution decision (Paper 16, 2023‑09‑18) — https://www.docketalarm.com/cases/PTAB/IPR2023-00592/Google_LLC_v._Robocast_Inc/docs/09-18-2023-Board/Institution_Decision__Deny-16-Denying_Institution_of_Inter_Partes_Review_35_USC_%C2%A7_314a.pdf

IPR2023‑00593 — Google LLC (and YouTube LLC) v. Robocast, Inc.

• Type: Inter Partes Review
• Filed: 2023‑02‑21 (filed concurrently with ‑00592)
• Status: Institution Denied. Terminated 2023‑09‑18.
• Judge panel: David McKone, Michael Cygan, Patrick Boucher.
• Claims challenged: claims 22–62 (41 claims; independent claims 22 and 46, drawn to search-result-based show structures). Deliberately non-overlapping with ‑00592: together the two Google petitions reached all 67 claims of the '932 patent.
• Petition grounds: § 103, using different art from ‑00592 (the petitions explain that the claims 22–62 features around "search results" required a separate reference set, including Guinan). I could not retrieve the complete ground-by-ground table for this petition and will not reconstruct it.
• Institution decision: 2023‑09‑18, denied. I did not retrieve the full text of this denial, so I am not quoting its reasoning. Patent Owner separately filed an Objection to Petitioner's Filing of Multiple Petitions, arguing the Board should deny at least one of the two Google petitions for improper multiple-petition filing — a live issue the panel resolved by denying both on the merits.
• Final Written Decision: none — no trial.
• Settlement / termination: none.
• Appeal: none.
• Defensive value: closes out the "all claims at the Board" strategy. Google spent two petitions to cover 67 claims and got zero institutions. Any defendant contemplating a fresh IPR on this patent should assume the Board will view the art landscape as already thoroughly explored (§ 325(d) / General Plastic risk) and should instead look for genuinely new art or attack under § 101/§ 112 in court.

Links: PTAB portal — https://portal.unifiedpatents.com/ptab/case/IPR2023-00593 · Docket data — https://gaeflexstaging-dot-docketupdate.appspot.com/cases/PTAB/IPR2023-00593/Google_LLC_v._Robocast_Inc/

Related proceedings (not on the '932 patent — do not conflate)

• IPR2023‑00081 and IPR2023‑00182 — filed by Netflix against U.S. Patent 7,155,451, the '932 patent's parent. These are on the '451 patent, not the '932 patent, and are referenced throughout the '932 record. (Note: some filings cite ‑00081, others ‑00182, for the Netflix '451 petition — I could not reconcile which number attaches to which of Netflix's two '451-related filings from the available sources.)
• IPR2023‑00590, ‑00591, ‑00594 — Google/YouTube petitions filed the same day as ‑00592/‑00593 against the other two asserted Robocast patents. Only ‑00592 and ‑00593 concern the '932 patent.
• District court: Robocast, Inc. v. YouTube, LLC et al., No. 1:22‑cv‑00304 (D. Del.); Robocast, Inc. v. Netflix, Inc., No. 1:22‑cv‑00305 (D. Del.) — both filed 2022‑03‑07. Earlier family litigation: Robocast v. Microsoft (1:10‑cv‑01055) and Robocast v. Apple (1:11‑cv‑00235).
• Federal Circuit: Robocast, Inc. v. Netflix, Inc., No. 25‑1651 (Fed. Cir.), docketed 2025‑04‑17 (appeal received 2025‑04‑14); Robocast is appellant, appellant's brief due 2025‑06‑16. This appears to be Robocast's appeal from the Delaware court's § 101 patent-ineligibility ruling in the 1:22‑cv‑00305 action (a February 2025 order holding the asserted claims abstract and lacking an inventive concept, citing Bluebonnet v. Pandora, Chewy v. IBM, and Broadband iTV v. Amazon). I could not confirm the Federal Circuit's disposition from the sources retrieved at my cutoff — verify this docket before relying on it. Docket: https://dockets.justia.com/docket/circuit-courts/cafc/25-1651 · District court § 101 order excerpt: https://www.courtlistener.com/opinion/[10337880](/patent/10337880)/robocast-inc-v-netflix-inc/

Strategic summary

Claim status: everything is UNTESTED at the PTAB; nothing is canceled and nothing is sustained. The '932 patent issued with 67 claims (independent claims include 1, 22, 46, and 63). Across Unified's petition (claims 1–9, 13–15, 17, 18, 20, 22–27, 33–43) and Google/YouTube's two petitions (claims 1–21 and 63–67, plus 22–62), all 67 claims were placed in front of the Board at least once — and the Board instituted on exactly none of them. There is therefore no FWD, no claim-cancellation, no "surviving claim" list, and no useful PTAB claim-construction or validity finding to import into your case. The single most important structural fact for a defendant is this: the PTAB has never reached the merits of this patent.

Estoppel landscape: there is none. Statutory estoppel under 35 U.S.C. § 315(e)(2) (and § 315(e)(1) before the Office) attaches only after a final written decision. Because all three petitions were denied institution, Unified Patents, Google, YouTube, Netflix, and their privies are not estopped by anything in this record, and neither are you. The Google/YouTube § 103 art (Shimizu + Benedict/Fox/Wolfe, plus Glaser, Sagman, Robertson) and Unified's grounds remain fully available as district-court invalidity theories — subject only to the ordinary rules of disclosure and to General Plastic / § 325(d) exposure if you try to refile the same art at the Board. Conversely, a defendant today is not protected by anyone else's failed petition either: a denial is not a validity holding, and the patent retains its presumption of validity.

Pattern signals. (1) A defensive aggregator is in the chain — Unified Patents filed IPR2022‑01125 as a third-party challenger in 2022, before either district-court defendant petitioned, and lost on the merits. (2) The same two practice groups have been fighting this family since 2010 — McKool Smith and Cantor Colburn for Robocast; Robocast has consistently litigated rather than settled, and its public statements after the Unified denial were aggressive ("Big Win for Robocast," emphasizing "the high quality of our patent assets"). (3) The Google/YouTube cluster (−00590 through −00594) is the largest single coordinated attack on the family; every one of the '932-targeted petitions was denied. (4) The patent expired on 2020‑12‑10, which caps damages exposure to pre-expiration conduct and removes injunctive relief from the table — the practical leverage here is retrospective royalties only.

Recommended next steps

• Do not build your defense on the PTAB. There is no FWD to cite and no canceled claim to point at. If a demand letter or complaint asserts claims 1, 22, 46, or 63 of the '932 patent, the Board record gives you nothing to quote.
• Anchor on the district court § 101 ruling and its appeal. The Delaware court held the asserted claims ineligible under Alice step one/two in a February 2025 order, reasoning that the claims recite the abstract idea of "collecting, organizing, and automatically displaying content (e.g., a playlist of Internet content)" without any specificity as to how, and rejecting Robocast's "technical solution to a technical problem" framing. That ruling is now before the Federal Circuit in No. 25‑1651. Check CourtListener and the CAFC docket for the current status and any opinion before you file anything — if the ineligibility judgment is affirmed, the '932 patent is effectively dead against accused playlist/tile functionality, and a demand letter citing claims 1 or 63 is worth pushing back on hard.
• Reuse the public IPR art, but re-verify it. The Shimizu/Benedict/Fox/Wolfe combination and Unified's grounds are public and un-estopped. Before you adopt them, pull the institution decisions and POPRs and confirm what the panels rejected, so you can fix the gaps rather than repeat them.
• Calendar the appeal, not the trial. With all three IPRs terminated and no trial before the Board, there are no PTAB milestones to track. The live dates are the Federal Circuit briefing and argument in 25‑1651 and the district-court scheduling in 1:22‑cv‑00304 / 1:22‑cv‑00305.
• If you are considering your own IPR petition: the combination of (a) three prior denials, (b) a 2020‑12‑10 expiration, and (c) General Plastic / § 325(d) exposure on the same or similar art makes a new petition a low-probability play. If you file anyway, lead with genuinely new art and a § 325(d) Advanced Bionics showing, and expect the panel to be aware of the '932 record.

Caveats on the record above: I retrieved the Unified denial language only as quoted by Patent Owner's public statement (not the primary Paper 13 PDF); I could not retrieve the full text or ground table of the IPR2023‑00593 denial; I could not retrieve Unified's specific prior-art combinations; and I could not confirm the Federal Circuit's disposition in 25‑1651. Each of those is marked above rather than filled in. The Google Patents record lists three PTAB cases for this patent, which conflicts with the ODP-derived "no proceedings" block in the prompt — the three are corroborated independently by PTAB/Unified/party filings, so I have reported them while flagging the discrepancy.

I'll research the assignment record, litigation history, and any entity details for US 8965932 before building the report.

Let me dig into the specific assignment records and Robocast's corporate/financing history.

Let me try to pin down the specific reel/frame and check for any additional recorded transfers.

Ownership & Assignment Analysis — US 8,965,932 B2 ("Automated content scheduler and displayer")

Verification starting points: USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (search "8965932" or app. 11/316,209) · Google Patents legal events — https://patents.google.com/patent/US8965932/en · Family: US 7,155,451 · US 8,606,819 · US 8,606,820 · US 8,738,655 · US 8,965,932 (all share the same chain).

Data limitation, stated up front. I was able to confirm the existence, type, parties, and recording date of the only post-filing assignment in this chain, but I could not retrieve the Assignment Center reel/frame number, execution date, or the recorded correspondent from the sources accessible to me. Those fields are marked "not retrieved" below rather than estimated. Everything I assert is sourced; everything I could not verify is flagged.

Inventors

• Sole inventor on all five family patents. No co-inventors. The Google Patents/PTAB records list "Damon Torres" / "Damon C. Torres" as the only named inventor for the '932 patent.
• "All inventors departing" pattern — not applicable. With a single inventor who is the assignee's founder/CEO, there is no inventor exodus. The unusual pattern here is the inverse: the inventor and the assignee are the same person, and the entity was built around the patent rather than the patent being assigned away from a product company.
• Chain-of-title oddity: the parent '451 patent shows "Original Assignee: Individual" in Unified Patents' record (https://portal.unifiedpatents.com/patents/patent/US-[7155451](/patent/7155451)-B1), consistent with the family being filed/applied for in Torres' individual name and only later papered over to Robocast, Inc. That is corroborated by the lone recorded assignment (below) being recorded in 2010 — long after the 1996/1997/1998 filings.

Original assignee

Robocast, Inc. — a Delaware corporation with a place of business in New York, NY (as pleaded in Robocast, Inc. v. Apple Inc., No. 1:11-cv-00235, D. Del., Complaint ¶2). Founded 1997 by Damon Torres.

• Products shipped: Robocast developed and marketed a software suite — RoboCast / RoboSurf / RoboSites / RoboGuide / RoboPublisher / RoboSearch / RoboStats (described in the '932 specification itself). Customer/partner claims from ~2001 include Hachette Filipacchi, ZDNet News.com, InfoSpace/MetaCrawler, and membership in the Microsoft TV Developer Program.
• Product revenue: material evidence says effectively none. In Robocast v. Netflix the transcript records that Torres "admitted… that they have not had a sale or revenue generated from the products that they have marketed" (D. Del. 1:22-cv-00305, Dkt. 342). Robocast's own investor letter (Aug. 24, 2016, filed as Dkt. 300-3) frames the business as a licensing campaign: "plan to aggressively accelerate our licensing campaign," "approach our biggest targets," referencing "investors… and litigation funders."
• Primary line of business today: patent licensing/enforcement plus an announced "relaunch" (RoboSearch, Robocast Social, RoboFeed). The company describes itself as providing "licensing, content and education."
• Current status: operating, not insolvent, but capital-dependent. Raised a reported ~$35M (Series B led by 2 NRP Managers LLC, with follow-ons from Brown Venture Group and vendor-investor Warecorp). No bankruptcy, Chapter 7, or Chapter 11 has surfaced. Historic: settled with Microsoft (2022 filing date? no — suits 2010/2011) and Apple (2011), and licensed Vevo (2016). Active litigant against Netflix and YouTube/Google, plus a Federal Circuit appeal (25-1651).

Assignment timeline

Recorded post-filing assignments found: exactly one. There is no evidence of a security agreement, merger, change-of-name, license recording, release, or correction in this chain.

• Executed: not retrieved / recorded 2010-12-17 — Reel not retrieved/frame not retrieved
  • Conveyance: Assignment (Assignment of Assignors' Interest)
  • Assignor: Torres, Damon C.
  • Assignee: ROBOCAST, INC.
  • Correspondent: not retrieved. (For context only — not a substitute for the recorded correspondent — prosecution counsel of record for the family was Ipsilon USA, New York, NY, per the '451 patent's "Attorney, Agent, or Firm" field; a 2012 terminal disclaimer in this family was signed "Joseph Sofer, Esq." dated 2012-12-13, app. 13/449,570, in USPTO PTAB papers. Whether the same firm filed the 2010 assignment recording is unconfirmed.)
  • Context: Chain-of-title cleanup around the first infringement suit. Recording date 2010-12-17 falls ~11 days after Robocast filed Robocast v. Microsoft (1:10-cv-01055-RGA, D. Del., filed 2010-12-06). It perfects Robocast's standing as owner of the family; it is not a transfer to a third party.

No later assignments. Per Google Patents legal events, after the 2010-12-17 recording the only remaining events are the 2015-02-24 grant, the 2020-12-10 adjusted expiration, and the litigation entries. Robocast, Inc. is, as far as the record shows, still the owner — which is itself the key finding: there is no shell-LLC conveyor belt to analyze.

Caveat: I could not open the Assignment Center record directly. If the Assignment Center returns additional entries (e.g., a security interest or a later assignment not surfaced in the Google Patents event list), the two "not retrieved" fields and the "no later assignments" statement should be updated.

Timeline diagram

timeline
    title Ownership of US 8965932
    1996 : Torres files provisional 60 025 360
    1997 : Parent application 08 922 063 filed
    1998 : CIP application 09 144 906 filed
    2005 : Application 11 316 209 filed
    2010 : Torres assigns family to Robocast Inc
         : Robocast sues Microsoft in Delaware
    2011 : Robocast sues Apple in Delaware
    2015 : Patent 8965932 issues
    2016 : Vevo takes non-litigated license
    2022 : Robocast sues Netflix and YouTube
    2023 : Unified and Google IPRs denied

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. No assignment from an operating assignee to a licensing-only LLC exists in this chain. The '932 patent was never moved off Robocast, Inc.; the sole recorded transfer (2010-12-17) runs to the operating entity from its founder. No "IP/Holdings/Ventures" successor, no registered-agent address shift, no single-purpose Delaware/Texas vehicle appears. (Robocast is itself licensing-focused, but the signal as defined — a transfer to a shell — is absent.)

2. Known asserter in the chain — NOT PRESENT (as to the enumerated lists) / UNCLEAR (as to function). Robocast, Inc. does not appear on the enumerated rosters (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation, Spangenberg entities). There is no reel/frame evidence of any of those entities in this chain. Robocast is, however, a repeat patent plaintiff against large operating companies — so it functions as an asserter even though it is not a listed NPE. Cite as context only: D. Del. 1:22-cv-00304 and 1:22-cv-00305; IPR2022-01125; IPR2023-00592/-00593.

3. Repeat correspondent across the chain — NOT PRESENT. With a single recorded link, there is no recurrence to test. No correspondent identity was retrievable, so no recurrence finding can be made either way (no fabricated attorney attribution).

4. Cascading transfers — NOT PRESENT. One recorded assignment over ~28 years, executed decades after filing. No chained LLCs, no <24-month cascade, no shared-principal transfers.

5. Pre-litigation transfer — UNCLEAR (leans not present). The only recorded assignment was recorded 2010-12-17, eleven days after the Microsoft suit was filed (2010-12-06). That is a post-filing chain cleanup, not a pre-suit transfer. Because the execution date is not retrieved, I cannot exclude an execution date in the six months before filing; hence "unclear." Either way it is a standing/venue housekeeping step, not a transfer to enable assertion by a new owner.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11, no §363 sale, no patent auction. Robocast has raised equity (~$35M Series B; Brown Venture Group follow-on) and litigates with funding rather than liquidating.

7. Privateering — NOT PRESENT. Privateering requires an operating company transferring patents to an NPE that asserts on its behalf. Here the asserting entity is the original assignee; no third-party operating sponsor appears in the record.

8. Defensive aggregator (anti-NPE) — NOT PRESENT. The chain does not terminate at RPX, AST, LOT, Unified, or OIN. Note the inverse relationship: Unified Patents, LLC is the adversary here — it filed the IPR2022-01125 petition against this patent (institution denied 2023-01-20). Unified's involvement is defensive challenge, not ownership.

Verdict

NPE — moderate confidence.

Justification. The ownership chain is clean and terminates at the original assignee: the only recorded assignment is Torres → Robocast, Inc., recorded 2010-12-17, ~11 days after Robocast filed its first infringement suit (D. Del. 1:10-cv-01055), and no shell-LLC transfer, cascade, bankruptcy sale, or defensive-aggregator terminus exists — so the chain-based signals (1, 3, 4, 6, 7, 8) are all absent, which caps confidence below "high." What drives the NPE call is entity conduct, not chain structure: the patent owner is the inventor's own licensing vehicle with documented no product revenue (Torres deposition in 1:22-cv-00305, Dkt. 342: "no sale or revenue generated from the products that they have marketed") and a self-described aggressive licensing/enforcement campaign funded by investors and litigation funders (Dkt. 300-3), serially asserting the family against Microsoft, Apple, Netflix, and Google/YouTube while operating companies repeatedly petition against it at the PTAB. That is two weak-but-documented signals — non-practicing monetization plus litigation-driven licensing — sufficient for moderate, not high, confidence.

Verify at: https://assignmentcenter.uspto.gov/ (search 8965932 / 11/316,209) — confirm the reel/frame, execution date, and recorded correspondent for the 2010-12-17 entry, and confirm no security-interest or later-assignment records exist.

I'll research the prior art citations for US 8,965,932 across USPTO and patent citation sources.

Let me retrieve the complete "References Cited" list from the patent's front page and the associated IPR prior art.

Let me confirm the remaining exhibit-list references and the "Wolfe" reference cited in the IPR grounds.

Prior Art Analysis — US Patent 8,965,932 B2

Method note (read first): The authoritative full text of US 8,965,932 B2 was supplied to me in the user message, so I have treated that as the primary source for the patent's own disclosure and claim structure. For the citations of record, I relied on the front-page "(56) References Cited" as reproduced in (a) the patent exhibit filed in the Robocast litigations (Docket Alarm, D. Del. 1:22-cv-00305, Doc. 96-3), (b) the Justia patent page for 8965932, and (c) the PTAB IPR filings. I did not have authenticated USPTO PatentCenter/PAIR access, so the list below reflects the patent's printed front page as it appears in those reproductions. Where the OCR is corrupt or the list was truncated in my sources, I say so explicitly. I do not auto-correct identifiers — e.g., the foreign document reads literally "JP 04-04205171 7/1992" on the face of the patent, which is almost certainly a mangled OCR of a Japanese publication number, but I report it as printed.

1. Where the citations come from

A patent's prior art comes from two distinct places, and they should not be conflated:

The patent issued notwithstanding all face-of-patent citations, and all three IPRs against it were denied institution — that framing matters for the §102 discussion in Part 4.

2. Face-of-patent citations — U.S. Patent Documents (56)

The following were confirmed from the printed front page. I list them in front-page order with title/assignee where I could confirm it, publication date, and a one-line description of the subject matter. Claims column gives the claims whose elements each reference most plausibly touches (see Part 4 for the anticipation reasoning).

2a. Early references (1971–1996)

Gap flag: My sources truncate the (56) U.S. list in mid-1998 (immediately after US 5,768,528). The next confirmed entry is US 5,809,247 (Sept. 1998). Entries such as US 5,793,980 (Glaser), US 5,778,187 (Monteiro), and US 5,802,292 (Mogul) appear in the reference list of the sibling family member US 8,606,819 and are very likely in this gap, but I could not confirm them on 8965932's own front page and therefore do not assert them as of-record citations here.

2b. Later references (1998–2001) — confirmed from Justia

3. Face-of-patent citations — Foreign documents and Non-Patent Literature (56)

3a. Foreign Patent Documents

3b. Other Publications (Non-Patent Literature) — the material set

This is, in substance, the most dangerous prior art for this patent, because it is all pre-1996 guided-tour/hypermedia literature:

4. § 102 analysis: which references could anticipate which claims

Critical framing: Anticipation under §102 requires a single reference to disclose every element and limitation of the claim as arranged. Because all of the above were cited of record and the patent still issued (and because PTAB denied institution on the IPR grounds), the honest conclusion is that none of the face-of-patent references was found to anticipate, and the strongest documented challenges were run entirely as §103 obviousness combinations, not §102 anticipation. With that caveat, here is the candidate-by-candidate assessment against the granted claims.

Granted independent claims (structure)

• Claim 1 (method): a show structure of nodes; nodes identify network-accessible resources; automatic access without user input; content auto-accessed/presented in accordance with the show structure; prefetching ("during presentation of some content, other resources are being accessed"); and interactively variable duration information enabling the user to vary display duration.
• Claim 22: search-driven variant — receive an online search request, return results + URLs, automatically create the show structure from those results, then auto-access/present.
• Claim 46: categorized search variant (predefined category list → results → auto-built show structure).
• Claim 63: authoring variant — select resources, drag-and-drop identifiers to generate the node list, then auto-access/present.

Anticipation candidates

Bottom line on §102: No single face-of-patent reference appears to disclose all of claim 1's limitations — in particular the combination of (i) a show structure of nodes that each carry interactively variable duration information, and (ii) prefetching of other network resources while current content is presented. The closest single-reference candidates are Richardson (5,809,247), Leak (6,182,072), Borman (5,890,172), and Zellweger/Hauck in the literature, and each is missing at least one core limitation.

5. The IPR prior art (the reference set the challengers actually bet on)

Because the face-of-patent list is not where the real fight was, the most relevant prior art for validity purposes is the Google/YouTube IPR set (IPR2023-00592 and -00593):

Grounds as pleaded (all §103, never §102):

• Ground 1: claims 1–11, 14, 16–21 obvious over Shimizu + Benedict + Fox + Wolfe
• Ground 2: search-based claims over Fox + Guinan (with Filepp for ad/profile claims)
• Ground 3: claims 46–62 over Shimizu + Fox + Benedict + Wolfe + Guinan + Miller
• Ground 4: claims 63–67 over Shimizu + Fox + Benedict + Wolfe + Sagman
• Ground 5: claims 11–16 (+ Glaser); Ground 6: claims 30–31 (+ Filepp); Ground 7: claims 52–53 (+ Filepp); Ground 8: claim 65 (+ Robertson)

All three IPRs were denied institution (Unified Patents IPR2022-01125, and Google IPR2023-00592/-00593), which is strong practical evidence that even the combined set did not render the claims unpatentable — much less that any one reference anticipated them.

6. Honest limitations of this analysis

1. I could not retrieve the complete (56) list verbatim. The U.S. patent portion is confirmed from ~1971 through mid-1998 and again through 2001, but there is a gap in mid-1998 (between US 5,768,528 and US 5,809,247) that my sources truncated. Items like Glaser 5,793,980 and Mogul 5,802,292 appear in the sibling patent's list and are probable but unconfirmed here.
2. The foreign document identifier "JP 04-04205171" is reported exactly as printed and is almost certainly an OCR artifact; I could not resolve the true Japanese publication number.
3. The "DAD486x" and "Wolfe" references are unidentified in my sources.
4. No authenticated USPTO PatentCenter access was available to me; a claim-chart-grade §102 opinion should pull the issued claims and the full "(56)" block directly from the patent PDF / PatentCenter.
5. §102 vs §103 distinction: my Part 4 assessment is a novelty screen. The actual litigation history shows the art was deployed as obviousness combinations, not anticipation, and was rejected at institution.

Sources: Google Patents (https://patents.google.com/patent/US8965932/en); Justia (https://patents.justia.com/patent/8965932); Docket Alarm D. Del. 1:22-cv-00305 Doc. 96-3 (front page of the '932 patent); PTAB IPR2023-00592 petition, Schmandt Declaration, and Petitioners' Corrected Exhibit List (https://www.docketalarm.com/cases/PTAB/IPR2023-00592/Google_LLC/); Unified Patents IPR2022-01125 / IPR2023-00593 records.

I'll research the prior art landscape for this patent, since the "Prior Art" section wasn't included in the fetched page text.

Let me get more detail on the specific IPR grounds and claims.

Obviousness Analysis — U.S. Patent No. 8,965,932 ("Automated content scheduler and displayer")

Analyst note / scope limitation (flag): The task instructs me to "use the results from the Prior Art section of this page." The page text I was given (the Google Patents fetch) did not include the "Citations"/"Cited By"/prior-art tab — it contained the bibliographic block, classifications, description, and Abstract only. I therefore could not read the patent's own front-page reference list from the supplied text. To avoid fabricating a "Prior Art section," I base the reference set below on (a) the prior art actually asserted in the three PTAB petitions against this exact patent, which I retrieved via search, and (b) references that appear on the face of this patent family's front page (visible in the sibling 13/449,570 document reproduced in the petitions). Any reference identification here that I could not verify should be checked against the patent's "References Cited" list and the issued claims.

Building on prior sections: The summary above correctly reports the bibliographic data, the "show structure" claim-1 core, the three IPRs, and the two Delaware cases. I do not repeat those. I add here the substantive prior-art grounds and the §103 combination analysis. One contradiction to flag: the task header states "Current Date: April 26, 2026," while the system date and fetch timestamp are 2026-09-30; nothing in this analysis turns on that, but the dates are inconsistent in the record.

1. Legal framework and the claim set that must be met

Because the family's earliest priority is September 3, 1996 (provisional 60/025,360) and the applications predate the AIA, the governing standard is pre-AIA 35 U.S.C. § 103(a) (as the petitions themselves framed it: "§ 103(a)"). The KSR/Graham framework applies: the question is whether the differences between the claimed subject matter and the prior art would have been obvious to a person of ordinary skill in the art (POSITA) at the time of invention, considering the scope and content of the prior art, the level of ordinary skill, and objective indicia of non-obviousness.

Independent claims (text drawn from the IPR excerpts of the issued patent):

Two constructions are decisive for §103 and were both litigated:

• "interactively variable duration information" = "a parameter specifying how long a content is to be displayed by default before a subsequent content is displayed, where the viewer of the content can change the parameter" (D. Del., Robocast v. Microsoft, June 28, 2013, Ex. 1011/EX1023), further clarified in Netflix v. Robocast, IPR2023-00081, Paper 19 at 17 (May 11, 2023): the parameter is provided during creating and capable of being varied during displaying.
• "nodes" = identifiers of separate and distinct resources (per the patentees' own specification, "Each node identifies an available resource").

A §103 case therefore lives or dies on (i) the interactively variable duration parameter and (ii) the concurrent look-ahead access, not on the auto-play idea in general — because auto-play/auto-advance over a sequence was pervasively known (the specification itself concedes PointCast "automatically provides updated information").

2. POSITA at the time of invention (Sept. 1996)

A POSITA would have had a bachelor's degree (or equivalent) in computer science/EE plus ~2 years' experience, or equivalent, with working knowledge of: HTML/HTTP and web browsers; client-server and CGI architectures; hypertext systems (including "guided tours"); GUI authoring tools (presentation/slide software); caching and prefetching; and information "push" systems. This is a low-to-moderate skill level for a field that was, by 1996, already crowded with auto-display and hypertext-navigation art — a factor that increases the likelihood of obviousness.

3. The prior art actually relied on (grounded in the petitions)

3.1 Unified Patents, IPR2022-01125 (challenged claims 1–9, 13–15, 17, 18, 20, 22–27, 33–43; denied Jan. 20, 2023)

Patent Owner's Preliminary Response shows the fault lines: Henley allegedly did not teach (i) the look-ahead "other resources are being accessed during presentation" step or (ii) "interactively variable duration information"; and Shoham/Richardson allegedly did not teach automatically creating a show structure from search results, "without requiring user input" access, or the concurrency/prefetch element.

3.2 Google/YouTube, IPR2023-00592 (claims 1–21, 63–67; denied Sept. 18, 2023)

3.3 Google/YouTube, IPR2023-00593 (claims 22–62; not instituted)

3.4 Family-face and specification-admitted art

• U.S. 5,740,549 (Reilly et al., PointCast, filed June 12, 1995) — "Information and advertising distribution system and method"; auto-selects categorized information + advertisements, displays on idleness, filters by subscriber profile.
• U.S. 6,807,558 (Hassett et al., PointCast) — "Utilization of information push technology"; CIP of the Reilly division, with 1995 lineage.
• EP 0 749 081 B1 — the Reilly/PointCast European counterpart.
• U.S. 5,778,367 (Wesinger) — automated on-line information service/directory served in HTML.
• U.S. 6,769,009 / 6,594,692 (Reisman) — on-line advertising/content placement.
• The specification itself admits PointCast "automatically provides updated information," that a user may access Netscape Netcaster / IE Active Desktop "channels" and "sit back," and that the invention's advance is reducing "clicks and decisions." These are admissions against interest usable as prior art.

4. Obviousness combinations and the motivation to combine

Below are the strongest §103 theories. For each I give the claim elements, the mapping, the KSR-type motivation, and the weakness.

Ground A — Auto-playing a sequence of Web resources (claims 1, 10, 37; and 22/46 in part)

Combination: Reilly (U.S. 5,740,549) + Guinan (guided tours) + Shimizu (U.S. 5,634,062) + Fox + Benedict.

• Reilly supplies a server that stores categorized information items + advertisements, distributes them to remote clients, stores them locally, and automatically displays them on a display device (idleness-triggered). That is "content derived from a plurality of resources … displayed automatically in accordance with [an arrangement]."
• Guinan supplies the "show structure of nodes" concept in the hypertext sense — dynamically planned guided tours through a hypertext graph, i.e., a path through nodes that is authored/selected for the user rather than requiring per-node clicks.
• Shimizu supplies automated sequencing/display of hypertext documents (the IPR characterized Shimizu as disclosing "hypertext nodes"), including automatic transition between documents.
• Benedict/Sagman (slide-presentation software) supply the known convention of per-item display timing with automatic advance, and a user-adjustable advance/speed control — the "duration information" element.

Motivation: Each reference is in the same field (hypertext/navigation, client-server content delivery, presentation software), the combination is of familiar elements performing their known functions (KSR), and there was strong market pressure to reduce the number of user clicks — a problem the patent itself frames. Guinan expressly motivates the tour approach to reduce disorientation and effort; Reilly motivates automatic display to reduce user action. A POSITA integrating them would arrive at "auto-play a path of nodes with timed advance" with a reasonable expectation of success (all components existed and were individually operable).

Weakness: The patent owner successfully argued in IPR2022-01125 that Henley/Shoham did not teach the concurrent look-ahead access and the interactively variable duration parameter (see Grounds B/C). Ground A must map those two elements or it fails — which is precisely why the Google grounds added Glaser and Robertson.

Ground B — The two "hard" limitations: concurrent look-ahead access and user-variable duration (claim 1)

Combination: Shimizu + Fox + Benedict + Glaser (U.S. 5,793,980) [+ Robertson (U.S. 6,486,895)].

• Glaser (asserted in Google Ground 3) is cited for the element "such that during said presentation of at least some of said content … at least some of said other network accessible resources are being accessed" — i.e., retrieving/preparing a next document while a current document is shown. Prefetch/background-retrieval of hypertext documents was a well-known browser/server optimization by 1996 (and the '932 specification itself describes "queue[ing] and cach[ing] the next resource … then load[ing] the next resource for viewing" as an implementation detail).
• Benedict (Harvard Graphics) is cited for the "interactively variable duration information" limitation: slide/presentation software conventionally let the presenter/user set the on-screen duration per item.
• Robertson (U.S. 6,486,895) was added for the "ordering of documents/content" aspect as applied to a dependent claim.

Motivation: Combining a document prefetch mechanism (Glaser) with an auto-advance presentation (Benedict/Shimizu) is the classic predictable combination: prefetching is a recognized way to hide latency in exactly the moment when the system is auto-advancing to the next resource (there is a direct functional relationship — the very reason to prefetch is that the next item will soon be needed). Supplying a user-adjustable dwell time is a routine GUI/presentation feature applied to the known auto-advance. KSR: "a combination of familiar elements according to known methods … yields no more than one would expect."

Weakness (real): Under the Microsoft/Netflix construction, "interactively variable duration information" must be provided at the creating step and be variable during playback — i.e., not merely a fixed timer, and not merely a prefetch. The patent owner argued Google's Benedict teaches only presenter-set timing in an authoring tool, not viewer-variable timing during playback of a node that is being auto-presented. This is the single most contestable element, and it defeated every petition to date. A stronger showing would combine Benedict (duration parameter) with Reilly or Filepp, which disclose an end-user-facing viewer whose display is controlled/personalized at runtime (Reilly's profile-based filtering; Filepp's user-side control), to close the "viewer can change it during playback" gap.

Ground C — Search-results-driven show-structure creation (claims 22 and 46)

Combination: Shoham (U.S. 5,855,015) + Richardson (U.S. 5,809,247) + Fox + Guinan + Benedict; alternately Shoham/Richardson + Filepp.

• Shoham is a network search/retrieval system returning result sets; Richardson (U.S. 5,809,247) is directed to organizing/retrieving network documents. Together they supply "receive a request for an online search … provide a plurality of search results and their corresponding resource locators."
• Guinan's "dynamically planned guided tours" is the natural bridge from search results → structured path through nodes: a guided tour is, definitionally, an automatically assembled ordering of hypertext nodes for a user.
• Fox supplies the HTML/publishing mechanics to render the tour as browsable resources; Miller (Using CompuServe) supplies category-directory (predefined-category) search behavior for claim 46.

Motivation: By 1996, search engines (Yahoo, Excite — named in the '932 specification itself) were returning ranked lists of URLs; "guided tour" research had already automated the ordering of hypertext nodes. A POSITA seeking to reduce the click burden of working through a search-result list would foreseeably combine the two: turn the result list into an auto-playing tour. This is an obvious design choice (KSR) driven by the same market need.

Weakness: The patent owner argued Shoham/Richardson did not teach "without requiring user input automatically accessing a plurality of said resources" nor the concurrent element, and that there was no articulated motivation to combine (IPR2022-01125 PO Response, §IX). The Board's discretionary §314(a) denial means this motivation argument was never fully resolved on the merits.

Ground D — Advertising/PSA insertion and personalization (dependent claims; claims tied to "persistent content window," dynamic content)

Combination: Filepp (U.S. 5,347,632) + Reilly (U.S. 5,740,549) + Reisman (U.S. 6,594,692 / 6,769,009) [+ Wesinger (U.S. 5,778,367)].

• Filepp (Prodigy) discloses retrieving network information for presentation with selective advertisement insertion; Reilly discloses categorized information + advertisements and subscriber-profile filtering; Reisman discloses on-line advertising placement; Wesinger discloses automated HTML service of user entries.
• The '932's interspersed advertising/PSA and profile-filtering features (Firefly-based) are thus squarely within the known art; the patent's own Summary concedes "interspers[ing] with advertising and public service announcements."

Motivation: Ad-supported "push"/auto-display was the prevailing business model (the specification admits as much), giving a strong economic motivation to combine content sequencing with targeted ad insertion.

Ground E — Drag-and-drop authoring (claims 10, 63)

Combination: Shimizu + Benedict + Fox + Wolfe + Sagman.

• Sagman (PowerPoint 4) and Benedict (Harvard Graphics) document the routine GUI paradigm of selecting items in one area and assembling them into an ordered presentation in another; Wolfe (U.S. 5,715,445) and Fox address hypertext document assembly/tour building.
• Mapping "drag icons from a first section to a second section to build an ordered list of resources" onto "create a show structure of nodes" is a predictable application of a known GUI technique to the claimed data structure — strong KSR footing.

Weakness: Patent owner argued the petition failed to show Shimizu's "nodes" qualify as the claimed nodes (a claim-construction fight about "nodes" = separate and distinct resources), and failed to show motivation/expectation of success for the combination (PO Prelim. Resp. §§VIII–X in IPR2023-00592/00593).

5. Summary judgment on obviousness

Overall: The concept of automatically sequencing and displaying a path of network/HTML resources was, by September 1996, an obvious combination of (i) push/auto-display systems (Reilly/PointCast), (ii) hypertext guided tours (Guinan, Shimizu, Wolfe), (iii) presentation-software auto-advance with user-set timing (Benedict/Sagman), and (iv) document prefetch (Glaser). The non-obvious core — if any survives — is narrow: a viewer-adjustable dwell parameter provided at authoring time and variable during auto-playback, coupled with look-ahead access of the next resource during display. Claims reciting that precise combination (notably claim 1) present the closest question; broad claims and the drag-and-drop/search claims present weaker non-obviousness positions.

6. Countervailing considerations (why three IPRs were denied)

1. §314(a) discretion, not pure merits. IPR2022-01125 and IPR2023-00592/00593 were all denied institution, several on Fintiv-type discretionary factors (parallel Delaware litigation, no stay, petitioner duplicating defendant efforts) layered on top of merits reasoning. Denial of institution is not a holding of validity.
2. Claim construction favored the patent owner on the critical terms. The "interactively variable duration information" and "nodes" constructions (from Microsoft, Ex. 1011, and Netflix-00081, Paper 19) required the petitions to prove a viewer-variable, playback-time duration parameter and separate-and-distinct resource nodes — which Henley, Benedict, and Shimizu were each found not to teach.
3. Motivation-to-combine was successfully attacked for the Shimizu + Guinan/Fox and Shoham + Richardson pairings (patent owner argued hindsight reconstruction).
4. Objective indicia would need weighing at trial (licensing to Vevo, industry recognition per Robocast's press release) — though §103 analysis here is on the merits of the art, and such evidence is often weak absent a nexus to the claims.

7. Confidence and verification

• High confidence: the identity and content of the IPR grounds and reference sets (Shimizu, Benedict, Fox, Wolfe, Guinan, Miller, Filepp, Robertson, Glaser, Sagman, Ackermann for Google; Henley, Shoham, Richardson, Reisman for Unified), the claim-element mapping, and the two controlling claim constructions — all drawn from PTAB and district-court documents.
• Lower confidence (flag): the precise titles/technical content of several reference patents (e.g., U.S. 5,634,062 Shimizu; U.S. 5,715,445 Wolfe; U.S. 5,793,980 Glaser; U.S. 6,486,895 Robertson; U.S. 5,761,417 Henley; U.S. 5,855,015 Shoham; U.S. 5,809,247 Richardson) — I could not verify their full text in this session and describe them per the petitions' characterizations. Verify against the patents themselves before relying on any specific teaching.
• Not verified: the full granted claim set of U.S. 8,965,932 (I have independent claims 1, 22, 37, 46, 63 and descriptions of claims 10/11–16, but not the complete dependent set), and the cited-by/prior-art tab of the Google Patents page, which was absent from the supplied text.
• Recommended next steps: pull the issued claims (PatentCenter / patent PDF) and the "References Cited" front page; obtain the four IPR petitions' claim charts (Ex. 1017/1018 in IPR2022-01125; the Ground charts in IPR2023-00592/00593); and confirm the Federal Circuit matter 25-1651 status, which the earlier summary flagged as unverified.