US 6587858

I'll verify current docket/litigation status before summarizing.

US Patent 6,587,858 B1 — Analyst Summary

Verification note: I could not reach the USPTO PatentCenter/Public PAIR or CAFC docket systems directly in this session. The bibliographic and claim data below are taken from the authoritative full-text record of US6587858B1 at Google Patents (https://patents.google.com/patent/[US6587858](/patent/US6587858)/en). My searches for 2026 activity specific to this patent returned nothing, so I have no authoritative information about any 2026 USPTO proceeding or CAFC docket involving 6,587,858. Treat all 2026 activity as unknown. I did confirm the historical Federal Circuit matter below.

Bibliographic Data

Abstract (verbatim, from the record)

"Systems and methods for controlling the dissemination of data from a repository based on the substitution of content contained within a generated response. The systems and methods are used by the repository to allow for the transfer of data that is not known or anticipated until the time of an access. For each access whose intent is to provide data to the repository or to retrieve variable data held by the repository, the repository uses the systems and methods to manage novel information structures whose purpose is to facilitate the substitution of the variable portions of content and to apply a privilege value to the portions that are variable. Using the concept of a 'notational fragment', the systems and methods are capable of assembling a variable result that has interaction with a data source. In addition, using a file structure known as a 'Query Template Construct (QTC)' the systems and methods provide for the language syntax of a data source to have variable portions."

Plain-Language Overview of the Independent Claims

Claim 1 — System claim ("comprising in combination"), a means-plus-function apparatus claim

A system inside a controlled/monitored data repository that exchanges data with a client. Its components:

1. Repository source means — the external data store/process from which data is obtained (specification identifies ODBC as the preferred implementation).
2. Extraction/parsing means — parses data in transmissions between repository and client, where that same data is also usable as data the repository source understands.
3. Mnemonic notation for transmission placement — a designer-defined set of finite character sequences that marks where, by position, data goes within a transmission; the notation must be distinct from the transmission content itself and must itself be usable as content of a "notational fragment."
4. Mnemonic notation for source-created constructs — same idea, but describing positional placement within a data construct created by the repository source.
5. Mnemonic notation for source-accepted constructs — same, for a construct the source accepts, and additionally required to be distinct from the commands the repository source understands.
6. Mnemonic notation for future transmissions — describes positional placement of current data (portions of source-provided content) within a future transmission (i.e., deferred/bind-later templates).
7. Privilege-based denial means — denies substitution into transmission content, denies providing content to the source, and denies retrieving content from the source, based on comparing a privilege value specific to the transmission against a predetermined privilege value of the repository source.

In short: Claim 1 is a template/interpreter architecture in which symbolic markers denote positional value-substitution slots in both client-facing transmissions and database-facing statements, combined with a privilege check that gates substitution, reads, and writes.

Claim 6 — Method claim

A method of describing placement order of data by position within (a) a transmission and (b) a construct understood by a repository source:

1. Use mnemonic notations (designer-specified, distinct from the data, usable in a notational fragment) to describe positional placement of data in transmissions and in source-created constructs.
2. Provide a user interface letting the designer select which transmissions and repository sources will use those notations.
3. Parse the format of a transmission and of repository-source data into symbolic notation describing the portions to be replaced.
4. Recognize the symbolic notation both in a transmission and in a repository-source construct.
5. Assign a correspondence between mnemonics used in a transmission and values accepted/returned by the repository source.
6. Determine and compare privilege values assigned to specific transmissions and source data against the privilege value assigned to the symbolic notations.
7. Reduce multiple mnemonic references to a single reference when they identify the same data object from the source (corresponds to the specification's SET/REF query/group consolidation optimization).
8. Use mnemonic notations to describe positional placement of current source-provided data within a future transmission.

In short: Claim 6 is the authoring-plus-runtime method counterpart to claim 1 — parse-and-tokenize formats into replaceable notation, map notation to database values, enforce privileges by comparison, and deduplicate references to the same underlying data object.

Dependent Claims (brief)

• Claim 2 — Adds data structures, expressed with the mnemonic means, that identify transmissions and repository-source content; those portions may or may not carry a privilege value, and privileged portions can have that value compared to a notational fragment's privilege value.
• Claim 3 — Interpreting the mnemonics as an interface symbol set: at least one base set, at least one finite character sequence per set, each sequence being a mnemonic symbol that references a software instruction sequence; the interface set is usable as notational-fragment content.
• Claim 4 — Data field markers: fixed syntax, unique from the content they substitute for, used to identify the position of modifiable data within transmissions, within content returned by the source, and within content provided to the source.
• Claim 5 — Means to associate data field markers used in a transmission with data field markers used in a structure understood by the repository source.
• Claim 7 — Incorporating a notational fragment into a sequence of fragment files, that sequence serving as instructions to an interpreter to assemble data into a transmission or into a source-understood format.

Litigation History (confirmed, historical — not 2026)

Google Patents' litigation metadata lists the following for this family (family ID 23606942):

• N.D. Tex. 3:15-cv-03745; W.D. Tex. 6:15-cv-00249; W.D. Tex. 1:15-cv-01056; D. Del. 1:15-cv-00777; E.D. Tex. 2:16-cv-00291
• CAFC No. 2016-2700 — Jack Henry & Associates, Inc. v. Plano Encryption Technologies LLC, 910 F.3d 1199 (Fed. Cir. Dec. 7, 2018) (Newman, Wallach, Stoll, JJ.), decided Dec. 7, 2018; originating case N.D. Tex. 3:15-cv-03745; outcome: reversed and remanded on personal jurisdiction/venue. The court held there is no per se rule barring patent demand letters from creating specific personal jurisdiction, so a declaratory-judgment action could proceed in N.D. Tex. where PET sent enforcement letters (Judge Stoll, joined by Judge Wallach, wrote "additional views" questioning Red Wing Shoe).

Important scoping caveat: the CAFC appeal was about venue/personal jurisdiction, not about the validity or infringement of 6,587,858. The patent number appears in that record only as an exhibit of PET's pre-suit licensing correspondence (Exhibit C, an infringement analysis of claim 11 of Patent No. 6,587,858). Note that 6,587,858 as issued has only 7 claims, so a reference to "claim 11" cannot refer to claim 11 of this patent as granted — it most likely refers to a different patent in the PET portfolio (the letter also referenced US 5,991,399 and US 5,974,550), or to a foreign/reissue counterpart. I cannot resolve that from the available record.

Source URLs for the CAFC matter:

• https://www.cafc.uscourts.gov/12-07-2017-2016-2700-jack-henry-associates-inc-v-plano-encryption-technologies-audio-uploaded/
• https://www.courtlistener.com/opinion/[4571118](/patent/4571118)/jack-henry-associates-inc-v-plano-encryption-technologies/

Stated Uncertainties

1. Assignee chain. Google Patents lists "Individual" as original assignee, with two recorded assignments (→ Plano Encryption Technologies LLC, 2015-05-07; → Cuica LLC, 2017-01-17). I have not independently verified the current 2026 chain of title; intervening transfers are possible.
2. 2026 activity. No search returned any 2026 USPTO (reissue, reexam, IPR/PGR, terminal disclaimer) or CAFC docket entry for 6,587,858. I have no authoritative information on 2026 status. Given the 2019-09-30 expiration, post-expiration proceedings are unlikely but I cannot rule them out.
3. Internal document inconsistencies. The published text contains apparent transcription/OCR errors that matter for claim-construction work — e.g., the specification refers to a "RESULT symbol 211" and a "FETCH symbol 206" where the numbering should be 201 and 205; and it inconsistently uses "SET parameter 215" vs. "REF parameter 215/224." These inconsistencies are present in the source document, not introduced by me, and any strict claim reading should be checked against the USPTO-certified copy rather than the HTML rendering.
4. Means-plus-function exposure. Claims 1–5 are drafted in means-plus-function form (35 U.S.C. § 112 ¶ 6 / now § 112(f)). Their scope is therefore tied to the structures disclosed in the 09/406,196 cross-referenced application — which is not reproduced in the 6,587,858 text I have — plus equivalents. I cannot state the full corresponding structure without that document.

I'll research litigation involving US 6,587,858. Let me start with the litigation portals and the case numbers already referenced in the patent record.

Let me pull the Stanford NPE database entry and search additional case records.

Let me gather outcomes/status for the key cases and any PTAB proceedings.

Litigation Report — U.S. Patent No. 6,587,858

Patent: U.S. 6,587,858 B1, "Systems and methods for the control of dynamic data and request criteria in a data repository"
Inventor: Steven Paul Strazza · Appl. No. 09/406,197 · Filed: 1999‑09‑30 · Issued: 2003‑07‑01 · Status: Expired – Lifetime (anticipated expiration 2019‑09‑30)

Ownership timeline (per the USPTO assignment record in the patent file):

• Original assignee: Individual (inventor Steven Paul Strazza)
• 2015‑05‑07 — assigned to Plano Encryption Technologies LLC (assignor: Strazza)
• 2017‑01‑17 — assigned to Cuica LLC (assignor: Plano Encryption Technologies LLC); Cuica LLC is the listed current assignee

Bottom line: Yes — 6,587,858 has been litigated. It was asserted/enforced almost exclusively by Plano Encryption Technologies, LLC ("PET") — a Texas patent-assertion entity whose "sole business is to enforce its intellectual property" — as part of a 2015–2017 campaign that aggregated this patent with two related patents, U.S. 5,974,550 ('550) and U.S. 5,991,399 ('399). The matters the patent record itself links to '858 are set out below; a broader set of PET campaign cases that asserted the same patent family follows.

1. Cases directly associated with '858 in the patent record

The Google Patents litigation panel (https://patents.google.com/patent/US6587858/en) links six matters to this patent:

Key appellate holding (case #2): Jack Henry & Assocs., Inc. v. Plano Encryption Techs., LLC, 910 F.3d 1199 (Fed. Cir. Dec. 7, 2018) (Newman, J., with Wallach and Stoll; additional views by Stoll, J.). The Federal Circuit held there is no per se rule that a patentee's infringement/assertion letters cannot create personal jurisdiction, and that PET was subject to personal jurisdiction in the Northern District of Texas based on its licensing-letters campaign directed at banks there. The dismissal was reversed and the case remanded. The three patents in suit were 5,974,550; 5,991,399; and 6,587,858. PET's demand letters to the banks cited "at least claim 6" (per the Broadway complaint) and, in the Jack Henry record, attached an infringement chart addressing claim 11 of the '858 patent.

2. Broader PET campaign (same patent family; '858 reported as added to the campaign)

PET filed two waves of suits in the Eastern District of Texas, Marshall Division (Judge Rodney Gilstrap). The following cases involve the same three-patent family and are part of the same enforcement campaign; the '858 patent is reported by RPX as being "later added to the PET campaign." Reported filing dates and dispositions:

Wave 1 (banks / banking-software providers, 2015):

Wave 2 ("general-purpose mobile apps," 2016):

Sources: Stanford NPE Litigation Database entry for 6,587,858 and PET party page (https://npe.law.stanford.edu/patent/6587858, https://npe.law.stanford.edu/party/plano-encryption-technologies-llc); PET's mandatory notices in IPR2017‑01783 (ptacts.uspto.gov); Patexia related-cases list for 2:16‑cv‑01073; RPX Insight, "Honeyman Cipher Files New Litigation over Former Liddle Patent…" (Aug. 23, 2019).

3. PTAB (IPR) proceedings

• IPR2017‑01721 and IPR2017‑01783, State Farm Mutual Automobile Insurance Company v. Plano Encryption Technologies LLC (filed July 13–14, 2017). These challenges were directed to U.S. 5,974,550 ('550), not to 6,587,858.
• No IPR specifically challenging 6,587,858 was identified in the records reviewed.

4. Post-litigation ownership / later transfers

• The '858 patent was recorded as assigned to Cuica LLC on 2017‑01‑17 (assignor PET). Cuica LLC is the current assignee of record.
• Separately, RPX reported that PET transferred the '399/'550 family (and characterized the '858 as part of the campaign) and that the patents moved among related Delaware entities (e.g., Honeyman Cipher LLC) in 2019; I could not independently confirm from the sources reviewed that 6,587,858 itself was among the 2019 transfers — the recorded chain for '858 per the patent file runs Strazza → PET → Cuica.
• The patent reached its 20-year term and expired 2019‑09‑30 ("Expired – Lifetime"), so the asserted claims are no longer enforceable.

Caveats on completeness and certainty

• I relied on the fetched Google Patents file, the Stanford NPE database, the Federal Circuit opinion (910 F.3d 1199), the D. Del. docket (1:15‑cv‑00777), the PTAB mandatory notices, and press/analyst summaries (RPX, law‑firm alerts, Patexia). Several of these sources list cases simply as "CLOSED" without a stated disposition, and I could not verify the precise final order in every case (e.g., whether dismissal/transfer/settlement), so I have flagged dispositions as "CLOSED" rather than characterizing them.
• The final disposition of the Jack Henry case on remand to the N.D. Tex. after the Dec. 7, 2018 reversal is not confirmed in my sources.
• I found no litigation in which 6,587,858 was asserted outside the PET campaign, and I did not locate a Cuica LLC‑ or Honeyman Cipher‑filed suit asserting '858 specifically.
• Minor docket‑date inconsistencies exist across sources (e.g., an AO‑120 form in the PTAB file shows "2:16‑cv‑791" with a 2015 date, while Patexia lists 2016‑07‑18); I used the dates reflected in the NPE/Patexia dockets.

If you need PACER‑verified docket sheets and final judgments for each case (and confirmation of which dockets formally named the '858 patent versus only the sister patents), I can pull those next.

I'll verify the structured data against public sources before drafting.

Proceedings overview

Total AIA trial proceedings on file for US 6,587,858: zero (0). Breakdown: 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution denied. The structured USPTO ODP block is the canonical list and it returns no AIA trials, and my independent searches for an IPR, PGR, or CBM against the ’858 patent likewise returned nothing — the only hits for the string "6587858" in PTAB-related documents were prior-art reference tables inside other patents' prosecution histories (e.g., a Google IPR2024-00750 exhibit listing ’858 in a Derwent citation table).

Bottom line for a defendant: this is not a "hardened patent" signal and it is not a "claims 1–7 already canceled" signal. It is a third state — no PTAB adjudication exists, so nobody is estopped, and every prior-art ground is still on the table. That said, the reason there is no PTAB history is almost certainly that the patent expired on 2019-09-30 at the end of its 20-year term (Google Patents: "2019-09-30 Anticipated expiration … Expired - Lifetime"). Your strongest defense is therefore not an IPR at all — it is the § 286 six-year damages window, which as of today leaves no compensable period for a newly filed suit. Details in the Strategic summary.

One caution on the source block: it contains no proceeding identifiers at all, so I have enumerated none. I did not invent any proceeding numbers, and I am flagging below the adjacent proceedings my search did surface and the one data point (a CAFC docket entry) I could not verify.

No proceedings on US 6,587,858 to enumerate

There are no entries to list in the requested per-proceeding format. The proceeding numbers that exist in the wild are not against this patent, which is the single most important thing to get right in a demand-letter response. The search results that look like "IPRs on Plano Encryption patents" are all on its siblings:

I deliberately have not filled in institution decisions, FWDs, or panel composition for those cases — I have no verified data on them, and misattributing an IPR outcome to the wrong patent is exactly the error that kills a defense brief. What matters for your purposes: the ’858 patent was never drawn into the PTAB wave that hit its siblings and the rest of the Plano Encryption portfolio, even though Unified Patents and State Farm were both actively petitioning against that portfolio in 2016–2017. Sources: Unified Patents, 2016-08-31 and the National Law Review PTO Litigation Center report for 2016-09-01.

The ’858 enforcement history you actually need (district court, not PTAB)

The structured block records five district court matters and one Federal Circuit entry tied to this family. These are the proceedings that explain the PTAB silence:

• E.D. Tex. 2:16-cv-00291 — https://portal.unifiedpatents.com/litigation/Texas%20Eastern%20District%20Court/case/2%3A16-cv-00291
• W.D. Tex. 6:15-cv-00249 — https://portal.unifiedpatents.com/litigation/Texas%20Western%20District%20Court/case/6%3A15-cv-00249
• W.D. Tex. 1:15-cv-01056 — https://portal.unifiedpatents.com/litigation/Texas%20Western%20District%20Court/case/1%3A15-cv-01056
• D. Del. 1:15-cv-00777 — https://portal.unifiedpatents.com/litigation/Delaware%20District%20Court/case/1%3A15-cv-00777
• N.D. Tex. 3:15-cv-03745 — https://portal.unifiedpatents.com/litigation/Texas%20Northern%20District%20Court/case/3%3A15-cv-03745
• CAFC 16-2700 — https://portal.unifiedpatents.com/litigation/Court%20of%20Appeals%20for%20the%20Federal%20Circuit/case/16-2700

Assignment chain per the structured block: Steven Paul Strazza → Plano Encryption Technologies LLC (recorded 2015-05-07) → Cuica LLC (recorded 2017-01-17). Note the sequencing: the suits were filed in 2015–2016 while Plano held title; the portfolio moved to Cuica in 2017. Cuica LLC is the current assignee of record.

Strategic summary

Claim status: all claims 1–7 are UNTESTED at the PTAB, and all are expired. There is no IPR certificate canceling anything, and there is no FWD sustaining anything. Claim 1 (the "repository source means / mnemonic notation to describe placement order" means-plus-function apparatus claim), claims 2–5 (dependent), claim 6 (the method of describing placement order), and claim 7 (incorporating a notational fragment into a sequence of fragment files) all stand with their original scope intact — but only as of the 2019-09-30 expiration date. Nothing has been narrowed, and nothing has been blessed. If a demand letter cites claims 1–7, the claims are real, but the enforceable term is over.

Estoppel landscape: a clean slate, which cuts your way. Because no IPR was ever instituted, § 315(e)(2) estoppel attaches to no one — not to any would-be petitioner, not to any privy, not to any defendant who considered filing and didn't. Every § 102 and § 103 ground, including art that "reasonably could have been raised" in some hypothetical IPR, remains fully available to a current defendant. The countervailing trap is § 315(b): if your client was served with a complaint asserting the ’858 patent more than one year ago — and the 2015–2016 filings suggest that is plausible for some parties — that client is time-barred from filing its own IPR irrespective of the estoppel analysis. Conversely, a defendant served only recently is not barred, and the patent remains technically IPR-eligible despite expiration (the Board construes expired claims under Phillips, and the patent owner cannot amend).

Pattern signals. (1) No repeat petitioner — the same petitioners (Unified Patents, State Farm) went after sibling patents but skipped ’858, which is consistent with ’858 being dropped from the 2015–2016 infringement contentions, settled early, or simply not worth the filing fee against a then-19-year-old patent. (2) No PTAB appeal aggressiveness — I found no ’858 FWD to appeal. The CAFC entry at 16-2700 is listed in the structured data as a Federal Circuit matter tied to this family; its subject matter and disposition are unverified — it may be an appeal from one of the district court cases or from a sibling-patent PTAB outcome, but I could not confirm which, and I am not going to assert one. (3) No defensive aggregator on this patent specifically — Unified Patents is in the chain only as the petitioner against sibling US 6,466,983 and as the source of the litigation data links, not as a challenger of ’858.

The dispositive fact. The patent reached the end of its 20-year term on 2019-09-30. Under 35 U.S.C. § 286, a complaint filed today can reach back only six years — to roughly 2020-09-29 — and the patent was already expired by then. There is no compensable infringement period left for a newly filed action. This is a stronger and cheaper defense than any IPR, and it does not depend on prior art at all.

Recommended next steps

1. Lead with expiration, not invalidity. For any assertion received on or after 2020-09-30, compute the § 286 window: the patent expired 2019-09-30 (Google Patents status: "Expired - Lifetime," anticipated expiration 2019-09-30; https://patents.google.com/patent/US6587858/en). A demand letter citing claims 1–7 as a basis for a forward-looking royalty or an injunction is not merely weak — the patent cannot be infringed after that date, and eBay injunctive relief is off the table.

2. Confirm the PTAB negative directly, in writing. Before relying on the absence of IPRs, run the patent number through PTAB E2E (https://ptacts.uspto.gov/ptacts/) and the USPTO Patent Trial and Appeal Board Decisions search, and preserve the null result. The ODP-derived block in this file is the canonical list, but a screenshot of the null search is what you put in a brief.

3. If the patent is still asserted for pre-expiration conduct, file the IPR only if you are inside the § 315(b) window. There is no estoppel against you, and an expired patent can be IPR'd, but weigh that against the near-certainty that the damages case is time-barred anyway. In most postures the § 286 argument disposes of the case without the $400K+ cost of an IPR.

4. Serve a Hygiene-proof record. Because laches was eliminated in SCA Hygiene Prods. AG v. First Quality Baby Prods., LLC, 580 U.S. 328 (2017), don't plan on delay-based defenses; run the § 286 and exhaustion-of-term arguments instead.

5. Resolve the CAFC 16-2700 open item. Pull the docket via the Federal Circuit's PACER/CM-ECF or CourtListener to determine whether that appeal touched the ’858 patent and what it held. If it was an appeal from a district court judgment on this patent, a preclusion or law-of-the-case argument may already exist in your favor — or a prior construction may bind you. Do not brief this patent without checking it.

6. Check whether your client is a privy of a past defendant. The 2015–2016 cases against E.D. Tex./W.D. Tex./D. Del./N.D. Tex. defendants may have produced settlements, licenses, or consent judgments whose scope runs to successors and affiliates. A prior license is a complete defense and is cheaper to find than to litigate.

Verification gaps, stated plainly. (a) I did not locate any PTAB proceeding number for US 6,587,858, and I am confident the count is zero based on the structured data plus two searches returning no such proceeding — but my search tooling was budget-limited before I could exhaustively query PTAB E2E and Docket Alarm. (b) I could not verify the nature or disposition of CAFC appeal 16-2700, and I have not characterized it. (c) I did not verify the institution decisions, FWDs, or judge panels for IPR2016-01644, IPR2017-01783, or IPR2017-01721, because those proceedings are on different patents and any detail I supplied would risk being attributed to ’858. If you want those filled in for portfolio-wide leverage analysis, say so and I'll pull them separately, clearly keyed to their own patent numbers.

I'll search for the USPTO assignment records, reel/frame numbers, and correspondent data for this patent.

Ownership Analysis — US 6,587,858 B1

Access limitation, stated up front (read before the timeline). I could not reach the USPTO Assignment Center or the legacy Assignment Search index in this session, and my searches did not surface any page exposing this patent's reel/frame numbers, execution dates, or correspondents of record. Rather than invent reel/frame identifiers, I have reconstructed the two recorded transfer events from Google Patents' legal-events metadata, which is itself derived from USPTO assignment records, and I have marked the fields I could not verify as NOT RETRIEVED. The Assignment Center does appear to have records for this patent (Google Patents shows two recorded reassignment events), so this is a data-access gap, not a finding of "no assignments."

Verify directly at: https://assignment.uspto.gov/patent/index.html (or https://assignmentcenter.uspto.gov/) — search by patent number 6587858 and read the reel/frame, conveyance, and correspondent fields.

Inventors

Pattern notes:

• Single inventor, self-owned, self-prosecuted. Both this patent (US 09/406,197) and the cross-referenced sibling (US 09/406,196) were filed by the same sole inventor on the same day (1999-09-30) with no corporate assignee. This is the signature of an independent developer, not a corporate R&D program.
• ~15.5-year dormancy before any recorded transfer. The patent sat with the inventor from filing (1999-09-30) to the first recorded reassignment (2015-05-07) — the entire life of the patent up to that point. There is no evidence the inventor ever commercialized it in that window.
• Technology-context inference (not an employment finding). The examiner's non-patent citations are Microsoft-centric web/database references (Blaszczak, "New ISAPI Classes in MFC 4.1," Microsoft Systems Journal (1996); Microsoft Data Access Components 2.5 SDK Beta (Jun. 1999); Raggett, HTML 3.2 Reference Specification (W3C, Jan. 14, 1997)). This tells you the inventor was working in the ODBC/HTML/ISAPI era of web-database middleware. It does not establish a Microsoft or any other employer, and I will not assert one.

Original assignee

Entity named on the issued patent: none — the patent issued to the inventor individually. Google Patents lists the original assignee as "Individual" (Steven Paul Strazza), and the record shows no assignment recorded before issuance. This differs from every scenario in the task template in which an "original assignee" is an operating company.

• Product embodying the claims: No evidence of any product. The patent discloses a template/interpreter architecture (notational fragments, the "QTC" file structure, an interface symbol set) intended to sit inside a repository that already enforces access control via the cross-referenced application's "DAC." Nothing in the record shows this shipped as a product.
• Primary line of business: None as an entity — this was an individual inventor's filing.
• Current status of the original holder: Unknown. Steven Paul Strazza is recorded as the assignor on the 2015-05-07 transfer (per Google Patents), so he divested. No public record of his later activity was located.

Contradiction flag (carried forward from the prior summary): the CAFC record in Jack Henry reportedly contains an infringement analysis of "claim 11 of Patent No. 6,587,858." This patent has 7 claims as issued. That reference cannot be to claim 11 of this patent; treat it as a portfolio/typo artifact in that exhibit rather than as claim-11 of '858. See the prior summary's caveat.

Assignment timeline

Two recorded transfer events. Execution dates, reel/frame numbers, and correspondents were not retrievable this session — do not treat the fields below as fully evidenced.

1. Recorded 2015-05-07 (execution date NOT RETRIEVED) — Reel NOT RETRIEVED / Frame NOT RETRIEVED

• Conveyance: Assignment — Google Patents states "ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)"; the underlying instrument's terms were not retrievable.
• Assignor: Strazza, Steven Paul
• Assignee: Plano Encryption Technologies LLC ("PET"), principal place of business Plano, Texas; PET's 2015 corporate disclosure statement in E.D. Tex. identifies its corporate parent as Koba Holdings, LLC. (Sources: PET's Rule 42.10(b) power of attorney and the docket entries in PET v. American Bank of Texas, reproduced at https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1494394](/patent/1494394)/download-documents; https://civilprocedureopinions.justia.com/2018/12/07/jack-henry-associates-inc-v-plano-encryption-technologies-llc/)
• Correspondent: NOT RETRIEVED. I cannot state the recording attorney/firm. (Caution: the litigation counsel of record for PET — Papool Chaudhari and Jeremy S. Pitcock, and IPR counsel Slayden Grubert Beard PLLC of Austin — are litigation counsel, not assignment correspondents. Do not conflate them without verifying the reel.)
• Context: Transfer-to-asserter. The inventor sold to a Texas LLC two months before that LLC's first infringement suits.

2. Recorded 2017-01-17 (execution date NOT RETRIEVED) — Reel NOT RETRIEVED / Frame NOT RETRIEVED

• Conveyance: Assignment
• Assignor: Plano Encryption Technologies LLC
• Assignee: Cuica LLC (Google Patents lists Cuica LLC as current assignee). Cuica LLC also holds unrelated patents (e.g., "Integrated Communication Security," "Inventory Management And Monitoring") per Justia's assignee page (https://patents.justia.com/assignee/cuica-llc) and PlainSite's entity table showing Cuica LLC as assignee on 2016, 2021 and 2022-era applications (https://www.plainsite.org/profiles/black-alistair/table-sec).
• Correspondent: NOT RETRIEVED.
• Context: Portfolio movement after the assertion campaign. Unclear whether this was a sale, settlement-driven transfer, or an affiliate reorganization — with no reel/frame or instrument, I will not characterize it further. It is not evidenced as a bankruptcy sale.

Bottom line for this section: exactly two recorded transfers are visible, both post-issuance, dated 2015-05-07 and 2017-01-17. That is ~20 months apart, from an individual to an LLC, then to a second LLC.

Timeline diagram

timeline
    title Ownership of US 6587858
    1999 : Filed by inventor Steven Paul Strazza
         : Sibling application 09/406196 filed same day
    2003 : Patent US 6587858 B1 issued
    2015 : Assigned to Plano Encryption Technologies
         : First PET infringement suits filed
    2017 : Assigned to Cuica LLC
    2019 : Patent term expires

NPE / troll-pattern signals

1. Shell-entity transfer — PRESENT (with a caveat on the assignor side).
The template's tell is an operating company → licensing-only LLC. Here the assignor was an individual inventor, not an operating company, so the classic fact pattern is only half met. What is present: the patent moved in May 2015 to Plano Encryption Technologies LLC, a Plano, Texas LLC whose own corporate disclosure names Koba Holdings, LLC as its parent — a two-tier holding structure. I have not located product evidence for PET, but "no products in commerce" is asserted here on the strength of PET's litigation record (below), not on a registry search I completed. Treat the "no products" leg as not independently verified.

2. Known asserter in the chain — PRESENT.
Plano Encryption Technologies, LLC is listed by the Stanford NPE Litigation Database as a patent asserter, categorized as "Acquired patents" (https://npe.law.stanford.edu/party/plano-encryption-technologies-llc). The database's page for this very patent ties it to Jack Henry & Associates Inc et al v. Plano Encryption Technologies LLC, N.D. Tex. 3:15-cv-03745 (https://npe.law.stanford.edu/patent/6587858). PET's suit record is high-frequency and spans two years: PET v. American Bank of Texas, E.D. Tex. 2:15-cv-01273; PET v. Q2 Holdings, 2:16-cv-00791; PET v. Best Buy, 2:16-cv-01049; PET v. Etsy, 2:16-cv-01050; PET v. State Farm Mutual Automobile Insurance Co., 2:16-cv-01072; PET v. Alkami Technology, 2:16-cv-01032 (with an order dismissing Groupon). PET is not on the named list in the prompt (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities), so this call rests on the "any entity surfaced by Unified Patents or RPX as a high-frequency plaintiff" prong — supported by the Stanford asserter classification plus the volume and venue pattern (E.D. Tex. Marshall, plus W.D. Tex., D. Del., N.D. Tex. matters in the Google Patents litigation metadata for family ID 23606942).

3. Repeat correspondent across the chain — UNCLEAR (data gap).
I retrieved no correspondent of record for either recording. I therefore cannot test the recurrence signal, which the task correctly identifies as the strongest tell. This is the single highest-value item to pull from the Assignment Center reel for both entries; if the same attorney/firm recorded both the 2015-05-07 and 2017-01-17 instruments, that, combined with signal 2, would move this to a very clean NPE finding.

4. Cascading transfers — WEAK / UNCLEAR.
Two consecutive LLC-to-LLC-adjacent transfers (2015-05-07 and 2017-01-17, ~20 months) meet the "<24 months" timing test, and both assignees are LLCs. But there are only two links, and I have no evidence that PET and Cuica share a correspondent address, a common principal, or a common registered agent. On the current record this is a two-step chain, not a cascade. Notable but not dispositive.

5. Pre-litigation transfer — PRESENT (strong).
The patent was assigned to PET on 2015-05-07. PET's earliest recorded assertion activity in this family follows almost immediately: E.D. Tex. litigation docketed from July 2015 (PET v. American Bank of Texas, 2:15-cv-01273, first docket entries 07/16/2015), with the N.D. Tex. declaratory-judgment action 3:15-cv-03745 brought by Jack Henry and a group of banks in the same year. Acquisition on 2015-05-07 → suits within roughly two months is well inside the 6-month window and is the classic clean-standing/venue-setup pattern. This is the strongest single signal on the page.

6. Bankruptcy fire-sale — NOT PRESENT.
No Chapter 7/11 proceeding, no court-supervised sale, and no evidence the original holder (an individual) was insolvent. There is no Kodak/Nortel/Polaroid-style fact here.

7. Privateering — NOT PRESENT / UNVERIFIED.
Privateering requires an operating-company backer using the NPE against its own competitors. PET's targets are community banks, a bank-software vendor (Jack Henry), a bank-technology vendor (Q2 Holdings, Alkami), and general retailers (Best Buy, Etsy, Groupon) — a scattershot monetization pattern, not a focused campaign on one sponsor's competitors. No SEC filing or Patent Progress/EFF coverage was located tying PET or Cuica to an operating sponsor. I found nothing to support this signal, but absence of found evidence is not affirmative disproof.

8. Defensive aggregator — NOT PRESENT.
The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. It terminates at Cuica LLC, which the Google Patents record identifies as the current assignee and which appears on unrelated patent applications as assignee. No evidence of neutralization.

Verdict

NPE — high confidence.

Justification: (a) the patent was transferred from the individual inventor to Plano Encryption Technologies LLC on 2015-05-07, and PET filed its first suits in this family within roughly two months (E.D. Tex. 2:15-cv-01273 docketed from July 2015; N.D. Tex. 3:15-cv-03745), satisfying the pre-litigation-transfer signal squarely; (b) PET is classified as an "Acquired patents" patent asserter in the Stanford NPE Litigation Database and is named in the Google Patents litigation metadata across six district-court matters plus CAFC No. 2016-2700, with its own 2015 corporate disclosure naming Koba Holdings, LLC as its parent — a holding structure with no product evidence in the record. Two independent strong signals therefore carry the verdict, with a third (pre-litigation timing) essentially inseparable from the first. The counterweight is honest: the correspondent field — the tell that would confirm a professionally managed assertion program rather than a one-off — was not retrieved, and the assignor here was an individual rather than an operating company, so the chain is a two-step LLC sequence rather than a cascade.

The next three data pulls, in priority order: (1) the correspondent of record on both the 2015-05-07 and 2017-01-17 recordings (Assignment Center search by patent 6587858) — recurrence is the deciding test; (2) the execution dates and conveying-party signatures on those instruments, to compute the true filing-to-suit interval; (3) Cuica LLC's formation records and registered agent, to test for common principals with PET.

Verification link: https://assignment.uspto.gov/patent/index.html (search 6587858).

I'll verify each reference cited on the face of US 6,587,858 against the record and live sources.

Scope and Method

Target ID, taken literally: US 6,587,858 B1 — Systems and methods for the control of dynamic data and request criteria in a data repository — Application No. 09/406,197, filed 1999‑09‑30, issued 2003‑07‑01, inventor Steven Paul Strazza.

Exclusions: I discarded hits for numerically similar but distinct patents that surfaced during searching — e.g. US 6,587,919 B2 (Douglis et al., System and method for disk mapping and data retrieval), which is a different patent and is not cited on 6,587,858. Nothing in the analysis below pertains to it.

Source limitation (stated honestly): My live query against USPTO PatentCenter did not complete (the search session was terminated before returning a result), and I do not have a certified copy in hand. The citation, date, and claim data below are therefore taken from the authoritative full-text record of US6587858B1 at Google Patents (https://patents.google.com/patent/US6587858/en), whose "Patent Citations (5)" and "Non-Patent Citations (3)" tables reproduce the USPTO front page. Where I state a § 102 subsection, the reasoning is mine, derived from the dates in that record.

Governing law: Because 09/406,197 was filed 1999‑09‑30, the pre‑AIA version of 35 U.S.C. § 102 governs. This matters materially, because three of the five cited references issued after the 6,587,858 filing date and therefore are available only as § 102(e) art.

The Five Examiner-Cited Patent References

Reference-by-Reference Analysis

1. US 5,644,764 — Unisys, "Method for supporting object modeling in a repository"

• Description: A repository (Unisys UREP) in which a user defines models, types, and features, and the repository catalogs, browses, and manages the resulting software components. It is a modelling/metadata repository — objects, meta-model relations, type definitions — not a request-processing environment. It says nothing about parsing client transmissions, substituting values into positional slots, or gating substitution on a privilege comparison.
• § 102 potential: Its only genuine overlap with 6,587,858 is the bare word "repository" and general object/data modelling. It does not anticipate claim 1 (no repository-source means in the sense claimed, no mnemonic-notation elements 3–6, no privilege-comparison denial means) and does not anticipate claim 6. At most it is § 103 background. No claim anticipated.

2. US 5,978,811 — Texas Instruments, "Information repository system and method for modeling data"

• Description: Models enterprise data as OBJECTs, inter-object RELATIONSHIPs, METHOD entities, and ACCESS INFORMATION entities; a navigator lets users browse the object/relationship network; FIG. 4 addresses "accessing data from a number of data sources located on diverse computer platforms." Critically, the specification states that a METHOD entity 16 may be a software routine that imposes a security clearance requirement before the data in OBJECT 12 may be displayed or otherwise made available to the user, and claim 2 / claims 12–13 recite "defining required clearance for accessing said data objects."
• § 102 potential (this is the closest prior art on the privilege element):
  • Claim 1, last element — the "means to deny ... derived from the comparison of a privilege value ... with said repository source means predetermined privilege value" has a real doctrinal neighbor in the 5,978,811 clearance/accessibility model. A § 102 challenge to that element alone is plausible. But § 102 requires all elements in a single reference, and 5,978,811 discloses no positional mnemonic notation, no parsing of client transmissions, and no substitution mechanism. Claim 1 as a whole is not anticipated.
  • Same for claim 2 (data structures with optional privilege values) — partial overlap on the privilege-value concept, no mnemonic notation.
  • Claims 3, 4, 5, 6, 7 — not anticipated.
• Note: This reference is the parent (07/921,826) of US 6,263,341 (Smiley, filed 1995‑06‑07, issued 2001‑07‑17, "Information repository system and method including data objects and a relationship object"). If a §102(e) attack were built on this family, 6,263,341 shares the disclosure.

3. US 6,003,039 — Platinum Technology, "Data repository with user accessible and modifiable reuse criteria"

• Description: A metadata repository in which reuse criteria used to detect duplicate/reusable repository objects are themselves modeled as repository objects, exposed through a Repository Services ("RS") API, and made user-accessible and user-modifiable. Implementation options surveyed include 3GL-with-embedded-SQL, RS calls, and stored procedures, and the spec discusses the API's select by NAME attribute, INS_ID and VERSION.
• § 102 potential: Superficial thematic resonance with 6,587,858's claim 6, step 7 ("reducing a multiple of mnemonic references to a single reference when said multiple references identify the same data object") — 6,003,039 is about deduplicating/reuse testing across repository objects. But 6,003,039 deduplicates authored repository objects by declared reuse criteria; 6,587,858 deduplicates QUERY/GROUP references by SET/REF parameter at run time. Different actors, different mechanism, and 6,003,039 has no mnemonic notation, no transmission parsing, no privilege-gated substitution. No claim anticipated. Useful only as § 103 background on "repository with user-modifiable criteria."

4. US 6,101,556 — New Era Of Networks, "Method for content-based dynamic formatting for interoperation of computing and EDI systems"

• Description (this is the most substantively relevant reference): An input message is parsed, per an input message format description, into a canonical form held in a rapidly searchable Field Group Array (FGA); an output message is then generated from a user-provided output message format description. Common ID values correlate the input field, output field, and FGA entry so that related fields map automatically. Format types are stored in a repository ("format dictionary"), and — significantly — "the user may add a new field ID value to the format dictionary." The stated purpose is to replace N² hand-written conversion procedures with declarative configuration. A counterpart was published as WO 98/30962 on 1998‑07‑16.
• § 102 potential: This is the reference whose architecture is genuinely close to the "notational fragment"/"QTC" idea — a designer-authored declarative artifact describing the structure and position of data that is substituted at run time, held in a repository, extensible by the designer.
  • Claim 1: 6,101,556 discloses designer-defined format descriptions and positional field mapping, but it does not disclose (a) the repository-source/client transmission context of 6,587,858, (b) a mnemonic notation required to be distinct from transmission content, or (c) the privilege-comparison denial means. Claim 1 not anticipated.
  • Claim 4 / claim 5 (data field markers with fixed syntax identifying the position of modifiable data, and means to associate a marker in a transmission with a marker in a source-understood structure): 6,101,556's field ID values correlating input/output/FGA entries are a fair §102(e) analogue of the association concept of claim 5. This is the strongest single-reference § 102 candidate in the set — but it is still directed to message-to-message reformatting rather than marker-to-datasource-construct association, so I would characterize it as a § 103 combination reference, not a clean anticipation, even for claim 5.
  • Claim 6: steps 3–5 (parse form into symbolic notation; recognize it; assign correspondence to source values) find their closest analogue here. Steps 6 (privilege comparison) and 7 (reference reduction) are absent. Claim 6 not anticipated.
• Cross-reference caveat: the WO 98/30962 publication date (1998‑07‑16) would make the publication itself available under § 102(b) as of that date, which is worth flagging if anyone ever tried to convert this from a § 102(e) to a § 102(b) reference.

5. US 6,285,997 B1 — IBM, "Query optimization with deferred update and autonomous sources"

• Description: An object-oriented query is transformed into subqueries, at least one against a database and one against an object cache; the database-directed subqueries populate the cache, and the cache-directed subquery then assembles the result so that it incorporates both cache updates and database updates. Filed 1998‑11‑16; a continuation issued as US 6,574,639 B2 (subject to a terminal disclaimer).
• § 102 potential: The query-transformation-and-consolidation theme is the only point of contact. 6,587,858's claim 6, step 7 ("reducing a multiple of mnemonic references to a single reference when said multiple references identify the same data object") describes a run-time optimization over references to source data; 6,285,997 optimizes queries over a caching tier. But 6,285,997 uses no mnemonic notation, no transmission parsing, no positional substitution, and no privilege comparison. No claim anticipated. It is a § 103 background reference on query optimization at best.

Non-Patent Literature Cited (all three bear on the enabling/prior-art implementation details, not on the claim combination)

None of these anticipates a claim; they are cited to establish that the disclosed implementation platform (ODBC, HTML forms) was conventional.

Conclusions

1. Most relevant prior art in the set: US 6,101,556 (content-based dynamic formatting via designer-authored, repository-stored format descriptions with correlating field IDs) is the closest overall, and US 5,978,811 (repository with METHOD-entity security-clearance gating and multi-platform datasource access) is the closest on the privilege limitation. If a validity challenge were mounted, the natural construction is 6,101,556 in view of 5,978,811 under § 103 — not § 102.
2. No cited reference anticipates any claim of 6,587,858 on its own. Claims 1 and 6 each require, in a single reference, the combination of (i) positional mnemonic notation usable in a notational fragment, (ii) parsing of client/repository transmissions, (iii) correspondence to datasource constructs, and (iv) privilege-comparison-gated denial of substitution/read/write. None of the five supplies all four. The examiner apparently agreed — the patent issued 2003‑07‑01 with all 7 claims intact over this art.
3. Subsection accounting worth preserving: only US 5,644,764 is a § 102(a)/(b) reference. US 5,978,811, US 6,003,039, US 6,101,556, and US 6,285,997 are all § 102(e) references, each granted on an application filed before 1999‑09‑30 but issuing after it. Any future challenge resting on those four must plead and prove the § 102(e) filing-date date, and must contend with the pre‑AIA § 102(e) proviso (the "before the invention by the applicant" prong), not merely the reference's issue date.
4. Do not confuse the "Cited By" list. The 27 (short table) / 85 (full table) forward citations — Citrix partial-page transmission, Voicbox/Nuance natural-language patents, etc. — are references that cite 6,587,858; they are not prior art against it.

Remaining Uncertainties

• USPTO PatentCenter not reached. My live query to PatentCenter terminated before returning; I cannot confirm from a USPTO primary source whether the front page carries any later-identified reference, a supplemental IDS citation, or a reexamination citation. The five-reference set above is what the front page of record shows.
• Certificate of correction / reexamination: I have no 1999–2026 record of a Certificate of Correction, reissue, or ex parte reexamination altering the claim set of 6,587,858. Given the 2019‑09‑30 expiration, none is likely, but I cannot rule it out.
• "Claim 11" anomaly (carried forward): A prior section flagged that court records referencing "claim 11 of Patent No. 6,587,858" cannot be correct, since the patent as issued has only 7 claims. I have no new information resolving that discrepancy and continue to flag it as unresolved.
• Means-plus-function scope: claims 1–5 are in means-plus-function form (pre‑AIA § 112 ¶ 6). The corresponding structure lives largely in the co-pending application 09/406,196, which is not reproduced in the 6,587,858 text and which I have not retrieved. Any § 102 analysis of claims 1–5 is therefore limited by not knowing the full disclosed structure.

Obviousness Analysis under 35 U.S.C. § 103 — US 6,587,858 B1

Prepared: April 26, 2026 · Basis: the "Prior Art / Citations" material reproduced on the US6587858B1 page (https://patents.google.com/patent/US6587858/en) plus the reference documents retrieved live.

Cross-reference check against the previously generated sections: No contradictions found. The prior art set I was instructed to use (US 5,644,764; US 5,978,811; US 6,003,039; US 6,101,556; US 6,285,997 + Blaszczak/ISAPI; MDAC 2.5 SDK Beta; Raggett HTML 3.2) matches the "Previously generated" summary. One item in the record is internally inconsistent and I flag it here rather than silently reconciling: the page shows both "Cited By (27)" and "Cited By (85)" headings for the same family, and both "Non-Patent Citations (3)" and "Citations (5)". Those cited-by documents are all later publications and therefore not § 102/§ 103 prior art against this patent; only the five examiner-cited patents and the three NPL items are used below.

1. Which references are actually available as prior art

This matters more than usual, because three of the five cited patents issued after 6,587,858's 1999-09-30 filing date. Under pre-AIA § 102(e) (6,587,858 is a pre-AIA application), a U.S. patent is prior art as of its U.S. filing date if that date precedes the applicant's invention. All five qualify:

The "cross-referenced application," US Ser. No. 09/406,196, is not § 103 prior art: it was filed the same day (1999-09-30) by the same inventor. It is the source of the corresponding structure for the means-plus-function claims and, for that reason, is a gap in any rejection rather than a weapon (see § 6).

2. Claim scope as it bears on § 103

• Claims 1–5 are means-plus-function apparatus claims (pre-AIA § 112 ¶ 6). For § 103 purposes, each "means" reads on the structure disclosed in the 09/406,196 application (DAC, form module, fragment types, build codes, response page entries, jump table, substitution list) plus equivalents. That is a double-edged fact: equivalents broaden the reach of the prior art, but the specification-locus requirement narrows it.
• Claim 6 is a method claim; claim 7 depends from it.
• The specification makes several admissions that are themselves prior art (MPEP § 2129): ODBC is "a well known, de facto standard method"; SQL "exists as prior art and is a well-known method"; HTML is governed by the cited W3C HTML 3.2 spec. These admissions strip the non-obviousness weight from the transport, query-language, and markup layers, leaving as the candidate points of novelty (a) the interface-set / notational-fragment architecture, (b) the data field marker positional-substitution scheme, and (c) the privilege-comparison denial.

3. Element-by-element mapping (lead reference: US 6,101,556)

US 6,101,556 is the closest single reference and is the natural primary reference. It is a declarative, format-description-driven engine that parses an input transmission into a canonical internal form and generates an output transmission, using user-authored format descriptions that are "devoid of message data," with field ID values correlating fields across input format, output format, and the internal Field Group Array (FGA) — and with "callouts to user supplied functions" and format "operations" that execute to derive field values.

US 6,285,997 is the best secondary reference, and one of its disclosures maps almost verbatim onto claim 6's most distinctive step: DECAF "rewrites groups of two or more SQL queries destined for the same source table into a single query." That is precisely claim 6's "reducing a multiple of mnemonic references to a single reference when said multiple references identify the same data object provided by said repository source means" — and it maps to the SET/REF consolidation the patent itself describes. 6,285,997 also discloses deferred updates, autonomous (independently updatable) sources, ODBC-accessible backends, and a three-tier query-service architecture.

US 6,003,039 supplies the repository-metadata + user-modifiable-criteria layer, with explicit ODBC/ANSI-SQL treatment of reuse routines. US 5,978,811 supplies the repository model (OBJECT/METHOD/ACCESS INFORMATION; navigation across "a number of data sources located on diverse computer platforms"; legacy-data incorporation irrespective of "technology, platform or format"). US 5,644,764 supplies the designer-facing repository modeling UI (model → types → features/operations/attributes/references).

NPL: Raggett HTML 3.2 documents the fixed-syntax FORM/INPUT field construct that claim 4's "data field marker" reads on. MDAC 2.5 SDK Beta documents ODBC/ADO/OLE DB — corroborating that the "repository source means" was standard, and pre-dating the filing by three months. Blaszczak documents ISAPI server extensions generating dynamic HTML/ISAPI filter parsing of HTTP transmissions — i.e., server-side dynamic page assembly.

4. Combinations

Combination A (primary): 6,101,556 + 6,285,997 + 6,003,039 (+ Raggett / MDAC as evidentiary NPL)

• 6,101,556 → the declarative, position/ID-based mnemonic notation engine (claims 1 ¶¶ 3–6; claims 3, 4, 5; claim 6 ¶¶ 1, 3, 4, 5).
• 6,285,997 → query consolidation (claim 6 ¶ 7), deferred/future binding (claim 1 ¶ 6), autonomous sources / ODBC backend (claim 1 ¶ 1).
• 6,003,039 → repository-side, designer/requestor-modifiable criteria and the natural situs of access control (claim 1 ¶ 7 in combination).
• Raggett/MDAC → the fixed-syntax marker and transport facts admitted in the specification.

Combination B: 6,101,556 + 5,978,811 + 6,285,997

Substitutes 5,978,811 for 6,003,039 for the "repository" side; 5,978,811's ACCESS INFORMATION entity ("access parameters," SQL access) is closer to the claim's repository-source data structure, and its explicit cross-platform/heterogeneous-source teaching supplies the architectural motivation.

Combination C: 6,101,556 + 5,644,764 + 6,003,039

Uses 5,644,764 for the claim-6 "user interface that allows the designer to select transmissions and repository source means" step and 6,003,039 for repository metadata/modifiable criteria; weaker on the query-consolidation limitation, which would then rest on the ordinary skill of a database artisan.

5. Motivation to combine (KSR / MPEP § 2143)

1. Same field of endeavor, common design pressures. All five patents address heterogeneous data interchange in distributed/client-server systems; 6,101,556 itself frames the problem as avoiding an N×(N−1)/2 explosion of procedural format converters, and 6,285,997 frames its goal as reducing data transferred between DBMS and client. Declarative, position/ID-keyed description (6,101,556) and query pushdown/consolidation (6,285,997) are two solutions to the same stated inefficiency. Reference in one would reasonably have led to the other.
2. KSR "predictable variation." Substituting a relational/ODBC datasource in place of 6,101,556's EDI/message formats is a substitution of one known element for another, yielding the predictable result of a web/DB template engine. The specification's own admissions (ODBC standard, SQL prior art, HTML 3.2) make this substitution more obvious, not less.
3. Market forces of 1997–1999. Blaszczak (ISAPI dynamic page generation, 1996) and the MDAC 2.5 SDK Beta (June 1999) evidence an industry-wide push to assemble database-backed HTML dynamically on the server. A POSITA faced with 6,101,556's format-description engine had strong market incentive to apply it to HTTP/HTML transmissions.
4. "Obvious to try" with a finite, predicable set. Once the goal is "let a repository designer describe substitution slots in a page and map them to database fields, without pre-building every page," the identified options were: positional markers, named markers, or a symbol/instruction set — a small, predictable set. 6,101,556 chose positional/ID markers; 6,285,997 taught collapsing multiple references to one source; combining them is not inventive.
5. Optimization rationale for claim 6 ¶ 7. Recognizing that multiple QUERY symbols hitting the same origin should collapse into one set-flagged query is a classic optimization; 6,285,997 claims exactly that consolidation. A POSITA optimizing a template engine would adopt it for the same reason 6,285,997 adopted it (fewer round trips, less data moved).
6. Access control. Repositories are, by definition, multi-user stores with per-object permissions (5,978,811's ACCESS INFORMATION entity). Adding a privilege gate to a template-interpreter's substitution/read/write paths — where the privilege of the accessing transmission is compared to the required level — is the application of a known repository-security technique to a known assembly engine, with predictable results. (I note the evidentiary weakness below.)

6. Weaknesses in the § 103 case (stated plainly)

1. Privilege-comparison denial (claim 1 ¶ 7; claim 6 ¶ 6). This is the weakest element. From the retrieved text, none of the five references discloses comparing a transmission-specific privilege value against a predetermined privilege value of the repository source to gate three distinct operations (deny substitution / deny write / deny read). 6,003,039 and 5,978,811 show repository access-control infrastructure, not this comparison. A well-supported rejection needs either a fifth reference or a § 103 "known-technique" rationale tied to concrete evidence.
2. "Eligible to be used as content of a notational fragment" (claims 1, 2, 3; claim 6 ¶ 1). This phrase ties the mnemonics to 6,587,858's own fragment type, defined only in the 09/406,196 application. If the PTAB/district court construes it as a structural requirement (the notation must itself be storable as a fragment processed by the DAC), the cited art was not shown to disclose it — because the DAC/fragment architecture is unique to the cross-referenced application, which is not prior art at all.
3. Claim 7 + the claim-6 "sequence of fragment files" step. Sequencing a notational fragment into a build-code sequence (with the 09/406,196 jump table/substitution list) is not disclosed by any cited reference. This is the most likely surviving limitation.
4. Means-plus-function specification locus. For claims 1–5 the corresponding structure must exist in 09/406,196. A § 103 rejection must map the reference disclosure to that structure or an equivalent; I cannot complete that mapping because 09/406,196 is not in the record I hold. This is a genuine evidentiary hole, not a formality.
5. No column/line cites. My element mapping above is at the disclosure level derived from the references' abstracts, claim text, and description excerpts retrieved in this session. A filing-grade § 103 analysis must cite specific columns/lines (and, for 6,101,556, the WO 98/30962 passages) to withstand an In re Kahn / KSR "articulated reasoning" challenge.

7. Claim-by-claim bottom line

Overall: US 6,587,858 is, on the cited art, largely obvious — the combination of 6,101,556's declarative, position/ID-keyed formatting engine with 6,285,997's query-consolidation and deferred-update techniques, set in the repository-access context of 6,003,039/5,978,811, discloses or renders obvious the substance of claims 1–6 with the single exception of the transmission-specific privilege-comparison denial. Claims 2, 4 and 5 are the most vulnerable; claim 7 (and, to a lesser extent, the "notational fragment"-eligibility language and the "base set" enumeration in claim 3) is the most defensible because it is anchored in structures described only in the co-pending 09/406,196 application — which, being a same-day, same-inventor filing, is not available as § 103 prior art.

Confidence: moderate. I confirmed reference identities, dates, and the specific text quoted above by live retrieval; I did not read the full specifications of all five references, and the strongest single non-obviousness argument (claims 1–5's corresponding structure in 09/406,196) cannot be evaluated without that document. Treat the privilege-denial and fragment-sequence conclusions as the two items that would decide a real § 103 challenge.