US 5025398

I'll search for the specific patent number and check for any litigation/docket records.

Let me check for any litigation or PTAB records involving this specific patent.

US 5,025,398 — "Data stream independent printer"

Search scope and result. I searched for the literal identifier 5025398 / US5025398A across patent databases (Google Patents, Justia, uspto.report, EveryPatent) and for court/PTAB records. I found no litigation, PTAB proceeding, or 2026 CAFC docket entry involving US 5,025,398. One caution: searching "5025398" also returns JP 5025398 B2 — an unrelated Japanese patent (Fujifilm, image-reproduction/slideshow apparatus, registered 2012). That is a different patent in a different jurisdiction and should not be conflated with the US number here. I also did not access USPTO PatentCenter/PAIR directly in this session; the bibliographic data below is grounded in the patent full text and the databases cited above, so treat the docket/status items as indicative rather than authoritative docket confirmations.

Bibliographic data

Abstract (as granted)

A nonimpact printer compatible with host processors programmed to operate line printers and with hosts programmed for nonimpact printers. Compatibility is achieved by conversion apparatus that emulates a line printer (e.g., a high-speed band printer) when the host is so programmed. The conversion apparatus stores character set definitions, overlays, copy modifications and setups on a disk drive in the printer, so band-definition information from the host can be translated into print-character-generation control signals. The host sends band identification data, which is used to retrieve the character set form/format from disk; the host then sends lines of print into a line buffer as character-identification-plus-position data; the conversion apparatus translates that into control signals that cause the nonimpact printer to generate the requested character in the requested position on a page basis, with any named overlays and copy modifications. Nonimpact printers can thus be substituted for high-speed line printers without reprogramming the front-end host software.

Plain-language overview of the independent claims

Claim 1 — Apparatus: line-printer-facing interface for a nonimpact printer. A host-side "line printer print driver" sends line-printer setup data and character definition data before it sends print character data. The interface has: (1) storage for the received setup and character-definition data; (2) memory holding multiple character set definitions, forms overlays, and copy modifications, each set made of individual print character definitions; (3) means that automatically — "independent of operator intervention" — convert the received character-definition data into character set selection data; (4) means to select the matching stored character set definition using that selection data; (5) means to convert the stored setup data into nonimpact printer control signals; and (6) means to activate the nonimpact printer to print each received character, plus the designated forms overlays and copy modifications. The core idea: the small host-supplied definition acts as a lookup key to a fuller set of character definitions kept inside the printer, with no human action needed.

Claim 4 — Method counterpart of claim 1. The same sequence as steps: store received setup and character-definition data; store multiple character set definitions, overlays and copy modifications; automatically translate the received character-definition data into character set selection data; select the corresponding stored character set; convert the setup data into nonimpact printer control signals; and activate the nonimpact printer to print a character for each received print character datum. (Note: the overlay/copy-modification language appears in the storing step, but the recited activating step is narrower than claim 1's.)

Claim 7 — Apparatus: printer interface framed as an "emulating" structure. Directed to a printer system in which the host's line printer print driver sends setup data, forms overlays, copy modifications and character-definition data before any print character data. The interface includes (a) means for emulating a line printer — storing the received data, translating the character-definition data (again, automatically, without operator intervention) into selection data, selecting the matching character set, and selecting the applicable forms overlay and copy modification — and (b) means, connected to both the nonimpact printer and the emulating means, for converting line printer control signals into nonimpact printer print control signals. This claim is notable for expressly reciting the emulation framing plus selection of overlays and copy modifications.

Claim 12 — Apparatus: page-printing nonimpact printer, line-at-a-time input. Emphasizes that the nonimpact printer prints page-by-page while the driver feeds print character data one line at a time. Adds explicit physical elements to the claim-1 concept: a line buffer buffering received characters line by line; means for translating each stored line into a corresponding line of nonimpact print data using the selected character set definition; a page buffer holding several such translated lines; and means for printing the page once an entire page is assembled.

Claim 14 — Method counterpart of claim 12. The step-by-step version of the same pipeline: store setup and character-definition data; store multiple character set definitions; translate (independent of operator intervention) into selection data; select the matching set; convert setup data into nonimpact control signals; buffer incoming characters line by line; translate each line into nonimpact print data; accumulate lines in a page buffer; and print once a full page is stored.

Claim 16 — Apparatus: host-to-printer over a data channel with a band printer driver. The most specific independent claim. The host includes a band printer print driver that puts line printer setup data and print band definition data on the data channel before sending character data line by line. The interface stores that setup and band-definition data, keeps a library of print band character set definitions, selects the one matching the received band definition, buffers incoming print character data in a line buffer, translates each character into corresponding nonimpact print data using the selected definition, converts the setup data into nonimpact printer control signals, formats each line of nonimpact print data using those control signals, holds a full page in a page buffer, and prints the page. This is the claim that most tightly ties the invention to band-printer (e.g., IBM 3211-class) hosts.

Notes, caveats, and things a reviewer should know

• Dependent claims 2–3, 5–6, 8–11, 13, 15, 17 add: line buffer/page buffer/forms-overlay buffer/copy-modification buffer details and page assembly (2, 9); formatting using the converted control signals (3, 6, 10, 13, 15); activation means (8); and host-resident generation plus transmission of character set definitions, overlays and copy modifications (11, 17).
• Means-plus-function exposure: the "means for …" recitations in claims 1, 7, 12 and 16 would today be construed under 35 U.S.C. § 112(f) to cover the corresponding structure disclosed in the specification and equivalents. The specification's disclosed structures are the control apparatus 113, page buffer 132, forms overlay buffer 133, copy modification buffer 134, forms control buffer 135, UCSB 136, channel buffer 137, and disk drive memory 114.
• Source-text artifacts: the granted claim 1 text in the databases reads "transmits lien printer setup data" (apparent typo for "line"), and the specification includes other OCR-level artifacts (e.g., "speCified," "copy modification: data," "for the purpose of simplicity"). I have not corrected them, per instruction — but they are transcription defects, not substantive claim language to be relied on.
• Technical context in the specification (useful for claim construction): the interface purports to conform to IBM's System/360 and System/370 I/O Interface Channel to Control Unit OEMI (GA22-6984-08), supporting block multiplexer, single/double-tag interlock, channel command retry, static two-channel switch, and data streaming; and the line-printer emulation follows the IBM 3211 Printer, 3216 Interchangeable Train Cartridge, and 3811 Printer Control Unit description (GA24-3543-9). The claimed solution turns on the observation that a 512-byte UCSB cannot carry 8K–52K of nonimpact character definitions — so the UCSB entries are repurposed as pointers into a disk-resident resource library loaded via the Diagnostic Write (X'05') channel command in 8K blocks.
• Uncertainty: I cannot confirm from authoritative USPTO records accessed live in this session whether any post-2003 litigation or 2026 CAFC activity exists; none surfaced. Given the patent lapsed in 2003 for non-payment of maintenance fees, any 2026 appellate activity would be unusual, and I would treat the absence of results as "no record found" rather than proof of none.

I'll research litigation involving US Patent 5,025,398 (the "Data stream independent printer" patent, filed 1990-05-21, granted 1991-06-18, originally assigned to Storage Technology Corp, now expired).

Initial searches returned only unrelated matches (a Japanese patent JP 5025398 B2, a Brazilian corporate registration number, and a journal DOI). Let me try litigation-specific sources.

Let me try a few more targeted litigation sources to be thorough.

Litigation Search Results: U.S. Patent No. 5,025,398

Bottom line

I found no record of any litigation involving U.S. Patent No. 5,025,398. Based on the searches I was able to run, there is no known district court action, ITC investigation, PTAB proceeding, or Federal Circuit appeal asserting or challenging this patent. I cannot report a plaintiff, defendant, jurisdiction, case number, filing date, or outcome because I found no case to report.

What I searched and what came back

False positives I explicitly excluded (numeric look-alikes)

Per the instruction not to auto-correct or conflate identifiers, these surfaced during search but are not US 5,025,398:

• JP 5025398 B2 — a Japanese patent (画像再生装置、及びプログラム / image reproduction apparatus), published 2012-09-12. Different jurisdiction, different number series.
• Brazilian court case No. 5025398-76.2018.4.03.6100 — Citigroup Global Markets Assessoria Ltda., TRF-3 (12ª Vara Cível Federal de São Paulo), a tax/mandado de segurança matter. Contains the digits but is a Brazilian judicial docket number, not a US patent case.
• Brazilian corporate registration No. 5025398 (Everton Seli Stre Ramos Pellenz) and DOI 10.1155/2019/5025398 — unrelated.
• UK FOI reference 5025398 (IOPC) — unrelated.

StorageTek litigation I found — none of it on the '398 patent

Because the patent was originally assigned to Storage Technology Corporation, I checked StorageTek's known patent disputes. The results concerned other patents and are not attributable to the '398 patent:

• Storage Technology Corp. v. Custom Hardware Engineering & Consulting, Inc., No. 02-12102-RWZ (D. Mass.), Fed. Cir. No. 04-1462 — copyright/DMCA/trade secret (GetKey circumvention); the opinion notes a patent infringement count was asserted but "not at issue in this appeal," and no patent number is tied to the '398.
• Odetics, Inc. v. Storage Technology Corp. (E.D. Va.) — U.S. Patent No. 5,???'151 ("rotary means" / pass-thru ports).
• Storage Technology Corp. v. Cisco Systems, Inc., 329 F.3d 823 (Fed. Cir. 2003) — claim construction appeal.

None of these identify U.S. 5,025,398 as an asserted patent.

Why the absence of litigation is plausible

This is context, not a substitute for a records search, but it is consistent with the null result:

1. The patent lapsed early. The Google Patents legal-events record shows a maintenance-fee reminder in January 2003, lapse for failure to pay maintenance fees on 2003-06-18, and expiration "due to nonpayment of maintenance fees under 37 CFR 1.362" (effective 2003-06-18). Although the "anticipated expiration" date is listed as 2010-05-21, the patent actually ceased to be enforceable in June 2003. Infringement suits generally cannot be maintained on an expired-for-nonpayment patent for ongoing conduct, which materially narrows the window (1991–2003) in which an action could have been brought.
2. Single citation, no reverse-citation trail to litigation. The prosecution cited one reference (US 4,779,105, Printer interface, Delphax Systems). The 16/25 "Cited By" documents are later patents/publications (Canon, Brother, Apple, Taligent, Lexmark, IBM, Seiko Epson), not litigation records.
3. No assignment/license trail indicating enforcement. The assignment record shows only the inventor→StorageTek assignment (reel 005314/0011, effective 1990-05-14). There is no recorded transfer to a patent assertion entity, which is the usual profile for an asserted patent of this vintage.

Confidence and caveats — please read

• Confidence in the negative finding: moderate, not high. I was unable to complete the two most probative searches you specified (site:portal.unifiedpatents.com for this patent number, and CourtListener/numeric full-text searches) because my search budget was exhausted. Unified Patents' litigation portal, Docket Navigator, PACER, Lex Machina, and CourtListener's numeric patent search are the authoritative sources for this question, and I could not fully query them.
• PTAB/AIA proceedings: Because this patent lapsed in 2003 — 9 years before the AIA's IPR/PGR provisions took effect (Sept. 2012) and well before IPR practice matured — IPR/CBM/PGR challenges to it are highly unlikely. I found none. (Pre-AIA reexamination remains theoretically possible but I found no record of one.)
• Recommendation: Before relying on this as a definitive "no litigation" answer, run a numeric search for 5,025,398 in Docket Navigator, Lex Machina, or CourtListener's RECAP docket search, and check the Unified Patents portal entry for the patent. If those return nothing, the negative finding can be treated as reliable.

I do not know, with high confidence, that no litigation exists. What I can state affirmatively is that my searches returned no evidence of any, and that the patent's early lapse and lack of reassignment make litigation unlikely.

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

PTAB Activity Report — US Patent 5,025,398 ("Data stream independent printer")

Proceedings overview

Total AIA trial proceedings on file: 0. The structured "PTAB proceedings on file" block (USPTO Open Data Portal, most recent ingest) returns no IPR, PGR, or CBM proceedings for US 5,025,398, and independent web searching surfaced none either (no institution decisions, no Final Written Decisions, no Federal Circuit appeals). There is therefore no breakdown by status to report — no claims have been canceled or sustained in an AIA trial, because no trial was ever instituted. Bottom line for a defendant: this is not a "hardened IPR survivor" story and it is not a "claims canceled" story — it is a dead-patent story. US 5,025,398 lapsed for failure to pay maintenance fees, with the lapse effective 2003-06-18 (USPTO legal events: maintenance-fee reminder mailed 2003-01-02; "Lapse for failure to pay maintenance fees" 2003-06-18; "Patent expired due to nonpayment of maintenance fees under 37 CFR 1.362" 2003-07-16). Its 20-year term would have run to 2010-05-21 in any event. Any assertion of this patent today fails at the threshold on status, before invalidity is even reached.

No proceedings to detail

I cannot populate the per-proceeding template (type, panel, grounds, institution decision, FWD, settlement, appeal) because there are no proceeding numbers to report, and the instruction is explicit not to invent them. What I verified:

• Google Patents lists the patent as "Expired – Fee Related," with no PTAB "Cited By"/proceeding entries of the AIA-trial type: https://patents.google.com/patent/US5025398/en
• USPTO PatentCenter for application 07/526,505 shows the maintenance-fee lapse chain ending 2003-08-12 ("Lapsed due to failure to pay maintenance fee," effective 2003-06-18).
• PTAB E2E / PTAB decisions searches and CourtListener docket searches for the patent number returned no AIA trial or CAFC records.

Two structural reasons no IPR ever realistically existed here:

1. Timing/status. IPR practice is against live patents (the Board does entertain IPRs on expired patents, but only where there is a live dispute worth funding). This patent lapsed in 2003, seven years before its statutory term would have ended, and 5,025,398 is 35 years old as of today.
2. Vehicle unavailability. A PGR was legally impossible — PGR applies only to patents subject to the first-inventor-to-file regime (effective 2013-03-16), and this patent was filed 1990-05-21. CBM review was limited to patents covering a financial product or service and the transitional program sunset on 2020-09-16; a band-printer emulation patent is a poor CBM candidate on its face. IPR was the only conceivable vehicle, and it was never used.

Defensive value: any demand letter citing US 5,025,398 is facially defective. The patent was unenforceable-after-lapse as of 2003-06-18 and expired no later than 2010-05-21. There is no PTAB record to litigate about, and no § 315(e)(2) estoppel has ever attached to anyone because estoppel requires a final written decision.

Strategic summary

Claim status. US 5,025,398 issued with 17 claims — independent claims 1, 4, 7, 12, 14, and 16 (apparatus and method claims directed to emulating a line/band printer driver to drive an all-points-addressable engine, using UCSB/FCB data as pointers into a disk-resident character-set, overlay, and copy-modification library), plus dependent claims 2–3, 5–6, 8–11, 13, 15, and 17. Routed through the PTAB taxonomy: CANCELED: none. SUSTAINED: none. UNTESTED: all 17 — and, critically, all 17 are expired and unenforceable. "Untested" here is marginally less important than "extinguished": there is nothing left to challenge.

Estoppel landscape. There is no estoppel landscape. Section 315(e)(2) estoppel flows from a final written decision; no IPR was instituted, so no petitioner or privy is barred from raising anything. Conversely, a defendant gets no benefit from an estoppel record and no benefit from a cancellation record — the defense here is status/expiration, not prior art. If you nonetheless want an invalidity fallback position, the only art the patent itself cites is US 4,779,105 (Delphax Systems, "Printer interface," granted 1988-10-18) — a single reference, which is a thin prosecution record, and the surrounding art (printer-emulation and driver-translation patents from Canon, Brother, HP, IBM, Lexmark, Seiko Epson, Taligent) is dense, per the forward-citation list on Google Patents. Treat that as a residual § 102/§ 103 sketch, not a formed ground.

Pattern signals. No repeat-petitioner pattern exists (there has never been a petitioner). No patent-owner appeal pattern exists (no FWD to appeal). No defensive aggregator (Unified Patents or similar) appears anywhere in the chain — expected, given the patent lapsed in 2003, before the AIA even existed. The only post-grant "event" of note is the fee lapse itself, a self-inflicted termination. Ownership: originally Storage Technology Corporation (inventor Thomas C. Nelson; assignment recorded 1990-05-21, effective 1990-05-14), later part of the StorageTek → Sun Microsystems → Oracle lineage, though I did not verify a current chain of title and would not represent one without a recorded-assignment check.

Recommended next steps

1. Lead with status, not invalidity. A patent that lapsed under 37 CFR 1.362 on 2003-06-18 and whose term expired 2010-05-21 cannot be infringed. Practically, this means a pre-suit demand response letter, a declaratory-judgment posture, or a motion directed at failure to state a claim — not an IPR petition. There is no live patent to petition against, and the Board would have no reason to institute.
2. Verify current fee/status and title yourself before relying on this. Pull the PatentCenter maintenance-fee page and the assignment record for 07/526,505. Any revival window under 37 CFR 1.362 has long since closed (the lapse is now over two decades stale), but confirm no later petition-to-revive or reissue appears on the record rather than assuming it.
3. Watch for a patent-number mismatch. The string "5025398" also corresponds to a wholly unrelated Japanese patent (JP 5025398 B2, Fujifilm, image reproduction apparatus) and to unrelated literature. If a demand letter or claim chart cites "5025398," confirm it means US 5,025,398 B1 and not one of these — a misidentified number is a fast, cheap way to attack the assertion's credibility.
4. Preserve the record of the search. If this comes up in litigation, document that as of 2026-10-01 there are zero AIA trial proceedings on the patent and zero CAFC appeals, and that the structured USPTO ODP feed confirms it. Re-run the ODP and PTAB E2E check at the time of any assertion in case a filing appears after the most recent ingest — but given the patent's status, a new filing is effectively inconceivable.

Confidence note: My high-confidence findings are (a) zero PTAB proceedings per the ODP structured block and (b) the 2003-06-18 lapse and 2010-05-21 anticipated expiration per Google Patents' legal-events table. I could not confirm (and therefore do not assert) any district-court litigation involving this patent, any current chain of title detail beyond the original assignment, or any non-AIA post-grant activity.

I'll verify the assignment record and original-assignee history before reconstructing the chain.

Let me check the Justia assignee record and look for any post-issuance assignment activity.

Assignment & Ownership Reconstruction — US 5,025,398

Scope note. I worked from the authoritative full text (Google Patents legal events block), the Justia/uspto.report mirrors, and StorageTek corporate history. I did not access USPTO Assignment Center interactively in this session (no live API hit), so the reel/frame below is the one recorded in the patent's own legal-events record rather than an independently re-queried Assignment Center extract. Where I could not verify a datum, I say so rather than infer it. No contradiction with the previously generated sections was found; this section extends, and does not restate, the bibliographic table already produced.

Search-hygiene caution (repeated because it recurs). Querying the bare string "5025398" returns several unrelated records that must not be conflated with this patent: JP 5025398 B2 (Fujifilm slide-show apparatus), a Brazilian court docket number, a UK FOI reference, and IRS/state registry entries. None are this patent. This matters for the "repeat correspondent" step below — a naive string match would manufacture a false hit.

Inventors

Pattern check — flag: unclear / no evidence either way. The fact pattern the task asks about (all inventors departing within 12 months of filing, preceding a portfolio fire-sale) cannot be assessed here: there is one inventor, and I found no record of his departure, no subsequent assignment, and no change-of-name or reissue naming a different owner. A sole-inventor patent filed and assigned pre-issuance to the employer is the ordinary corporate case, not an unusual pattern. I will not call it one.

Original assignee

Storage Technology Corporation (Louisville, Colorado; NYSE: STK; renamed "StorageTek" c. 1983). Public company, ~$2.2B revenue and ~7,000 employees as of 2004.

• Did they ship a product embodying the claims? The patent is directed to a nonimpact (all-points-addressable) printer that emulates an IBM 3211-class band printer — character set, forms overlay, and copy modification emulation driven by the host's UCSB/FCB loads. Whether StorageTek commercialized this specific printer is not established by anything I found. StorageTek's documented commercial identity in this era is tape automation and disk arrays (the 1987-onward automated tape library line), not electrophotographic/all-points-addressable print engines. Caveat stated plainly: the specification is written as a printer-architecture disclosure (print engine 112 and control apparatus 113 are described as "well known in the art"), which reads more like an OEM/interface product than a StorageTek-built engine. I cannot confirm a product and will not assert one.
• Primary line of business: data storage hardware and software (tape drives, tape libraries, disk arrays, HBAs) plus professional services.
• Current status: Dissolved as an independent entity. StorageTek was acquired by Sun Microsystems in August 2005 (announced June 2005; $4.1B cash, $37.00/share) and Sun was in turn acquired by Oracle on 2010-01-27 ($7.4B). The remaining product line is marketed as Oracle StorageTek. Separately, StorageTek had filed Chapter 11 in 1984 — but that is six years before this application was filed (1990-05-21), so it has no bearing on this patent's chain (see signal 6).

Assignment timeline

Chronological, every recorded assignment:

• 1990-05-14 (executed) / recorded 1990-05-21 — Reel 005314/0011
  • Conveyance: Assignment (ASSIGNMENT OF ASSIGNORS INTEREST)
  • Assignor: Thomas C. Nelson
  • Assignee: Storage Technology Corporation (Colorado)
  • Correspondent: Not determinable from the sources accessed. The legal-events record identifies the assignee and the reel/frame and states "ASSIGNMENT OF ASSIGNORS INTEREST," but does not expose the recording correspondent's name/firm in the material I retrieved. I am not going to guess an attorney name from a Colorado patent boutique of that era; doing so would be fabrication. If a full Assignment Center pull is taken, the correspondent field is the one datum still to be filled in.
  • Context: Ordinary pre-issuance employment assignment — inventor to his corporate employer, executed one week before the application was filed.

That is the entire assignment record. Beyond this single entry, the legal-events block shows only maintenance-fee and fee-payment-procedure events (FEPP 1990-12-08; FPAY 4-year 1994-09-26; FEPP 1998-12-01; FPAY 8-year 1998-12-07; REMI 2003-01-02; LAPS 2003-06-18; STCH 2003-07-16; FP 2003-08-12). No post-issuance assignment, security agreement, merger recording, change of name, license, or release was recorded.

Important structural observation on the Sun/Oracle acquisitions. StorageTek's 2005 acquisition by Sun and Sun's 2010 acquisition by Oracle were asset-level successions that would have swept in StorageTek's patent portfolio — but no merger/assignment recording for this patent appears in the record, and more decisively, this patent lapsed on 2003-06-18 for non-payment of maintenance fees (37 CFR 1.362), two years before the Sun deal and seven years before Oracle. The patent had already expired as an enforceable right (anticipated expiration 2010-05-21) before either acquirer could have asserted it. Any notional "current owner" is therefore Oracle via the described M&A, but the acquired asset was already a dead right. I flag this as the single most important fact in the chain.

Timeline diagram

timeline
    title Ownership of US 5025398
    1990 : Inventor Nelson executes assignment
         : Recorded Reel 005314 0011
         : Application filed same month
    1991 : Patent issues June 18
    1994 : Four year maintenance fee paid
    1998 : Eight year maintenance fee paid
    2003 : Lapsed for unpaid maintenance fee
    2005 : StorageTek acquired by Sun
    2010 : Sun acquired by Oracle

(Event text kept short and punctuation-free for parser safety; the "Reel 005314 0011" rendering drops the slash deliberately.)

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. The only assignee in history is an operating public company (Storage Technology Corporation, Reel 005314/0011). No "IP/Holdings/Ventures" transferee, no registered-agent address, no single-purpose LLC anywhere in the record.

2. Known asserter in the chain — Not present. Neither StorageTek, Sun Microsystems, nor Oracle appears on the RPX / Unified Patents / public NPE directories in connection with this patent. No assignment to Acacia, Marathon, IV, Wi-LAN, Conversant, Round Rock, Document Generation Corp, or any Spangenberg entity exists in the record. I found no litigation of any kind naming US 5,025,398 (consistent with the finding in the prior section).

3. Repeat correspondent across the chain — Not present / not assessable. There is exactly one recorded assignment, so there is no "chain" in which a correspondent could recur; recurrence is structurally impossible here. Additionally, the correspondent identity itself was not retrievable this session. Two honest limits, not two findings. (And note the search-hygiene trap above: string-matching "5025398" yields unrelated correspondents in JP/BR/UK records that could be mistaken for repeat players.)

4. Cascading transfers — Not present. Zero consecutive transfers, let alone multiple LLC hops within 24 months. The single assignment is dated 1990-05-14 and nothing follows it.

5. Pre-litigation transfer — Not present. There is no infringement suit to precede. No transfer within 6 months (or 6 years) of any suit.

6. Bankruptcy fire-sale — Not present. StorageTek's Chapter 11 (1984) pre-dates this patent's 1990 filing by six years and therefore cannot have involved this patent. StorageTek emerged from that proceeding and was a going concern through the 1990 filing, the 1991 grant, and the 1994/1998 fee payments. No Kodak/Nortel/Polaroid-style patent auction touched this asset; it instead simply lapsed in 2003 for fee non-payment — a much weaker signal than a fire-sale (it indicates the owner judged the patent not worth $X in maintenance fees, not that an NPE bought it).

7. Privateering — Not present. No transfer from an operating company to an NPE asserting on the operating company's behalf; no SEC 10-K/8-K disclosure, EFF, or Patent Progress coverage surfaced. Again, no assertion exists to be a privateering backend. (Limitation: I did not pull Sun's or StorageTek's SEC filings live this session; the negative rests on no assertion surfacing and on the patent's 2003 lapse.)

8. Defensive aggregator (anti-NPE) — Not present. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. Termination point is fee lapse (2003), with nominal M&A succession to Oracle thereafter.

Verdict

Insufficient data — but for a sharper reason than an empty record: the Assignment Center chain contains only the original employment assignment (Thomas C. Nelson → Storage Technology Corporation, executed 1990-05-14, recorded 1990-05-21, Reel 005314/0011), with no post-issuance transfer of any kind. The patent then lapsed 2003-06-18 for non-payment of maintenance fees, roughly two years before StorageTek was acquired by Sun (2005) and seven before Sun was acquired by Oracle (2010). Every NPE signal in the checklist is not present, and most are structurally impossible (no chain, no suit, no LLC, no aggregator). This is a textbook operating-company-owned, never-transferred, quietly-lapsed patent; the correct NPE conclusion is no NPE activity, and the residual uncertainty is limited to records I could not re-query live.

Two items a reviewer should close out with a live Assignment Center pull:

1. The correspondent of record on Reel 005314/0011 (the one datum missing above).
2. Confirmation that no Sun/Oracle merger recording ever touched US 5,025,398 (I expect none, since the patent had lapsed, but that is an expectation, not a verified negative).

Verification link: https://assignmentcenter.uspto.gov/ (search 5025398; the equivalent legacy search is https://assignment.uspto.gov/patent/index.html).

I'll research the prior art for US 5,025,398, grounding the analysis in the authoritative full text and verifying the cited reference. Let me run targeted searches.

Let me verify the cited reference's bibliographic details and check the USPTO record directly.

Prior-Art Analysis — US 5,025,398 ("Data stream independent printer")

Scope note and one flag before the analysis

Search scope. I queried Google Patents, Justia, uspto.report, and EveryPatent for the literal identifier 5,025,398 / US5025398A and for its cited references. The full patent text supplied in the record is authoritative and I treated it as such. I did not access USPTO PatentCenter/PAFT live this session, so the bibliographic items below are grounded in the patent record and the database mirrors above rather than a live USPTO query.

Date flag. The task header states "Current Date: April 26, 2026," while the operating environment states today's date is 2026-10-01. This doesn't change the analysis (all events here are historic), but I'm noting it because the two statements conflict and I was told not to silently resolve identifiers/dates.

Critical framing point — read this before the tables. US 5,025,398 has exactly one patent citation of record: US 4,779,105 A. This is confirmed by the patent's own front-page-style data ("Patent Citations (1)") and the full text's citation section. Everything else appearing in a typical Google Patents page for this number — the "Cited By (16/25)" entries and the "Similar Documents" list — is not prior art to the '398 and cannot anticipate it:

• "Cited By" documents (Canon, Brother, Apple, Taligent, Lexmark, IBM, Seiko Epson, etc.) cite the '398; they post-date it and therefore have later effective dates. A reference cannot be § 102 prior art to a patent it post-dates.
• "Similar Documents" (e.g., US 3,803,558, US 4,860,219, the DTIC report AD-A140587) are algorithmically surfaced by the database. Except where separately established as pre-1990 prior art, they are not citations in the '398 record. I address one of them below only as supplementary context, clearly flagged as not of record.

So the honest answer to "look at each patent citation for 5025398" is: there is one, and only one. I analyze it fully rather than padding the list with non-citations.

Citations of record

Source: https://patents.google.com/patent/US4779105 and the '398 record's citation section.

Reference 1 — US 4,779,105 A ("Printer interface," Delphax Systems)

Full citation and dates

• Citation: U.S. Patent No. 4,779,105 A, "Printer interface," assigned to Delphax Systems.
• Filing date: 1987-10-06. Issue date: 1988-10-18.
• Prior-art date relative to '398: The '398 was filed 1990-05-21. With a 1987 filing and 1988 publication, US 4,779,105 qualifies as prior art under § 102(a)/(b) (and under pre-AIA § 102(e) as of its filing date if relied on for its disclosure). It is the sole reference the examiner cited.

Brief description of the disclosure (grounded in the reference text)

US 4,779,105 is a printer interface / control module for a nonimpact (ion-deposition, "cell image engine") printing system that:

• Receives parallel data from a host computer via a channel (IBM-style channel 46).
• Explicitly states the control module "can be modified to emulate a peripheral printer to any host computer," and that "the parallel data in the format received from the host computer is translated into a 'native' format used internally to create serial data in raster scan arrangement for the printing subsystems."
• Uses a layered architecture: a first layer (channel driver/receiver card + channel interface controller with an Intel 80186) emulates the host-facing printer; a second layer (cell image engines with Intel 80286) drives the print engine.
• Stores "the fonts and formats to be used by the cell image engines" on a 3.5-inch floppy disk via a floppy drive controller, along with control programs.
• Notes the page format is "either selected at the first layer or is part of the output from the host computer," and that the format "defines a layout of rows and columns of cells which after translation into native code is stored in immediate memory in the second layer as a list," after which "character data is translated and written into the cells from the originating data coming from the host computer."
• Feeds the print engine with serial raster scan data to be converted to dot imaging on the page.

Sources: https://patents.google.com/patent/US4779105 and the issued PDF at patentimages.storage.googleapis.com (US4779105.pdf).

Which claim(s) it potentially anticipates under § 102

Bottom line: US 4,779,105 is the closest prior art of record, but on its face it does not appear to disclose every element of any independent claim — so it reads better as a § 103 obviousness base than as a § 102 anticipation reference. The examiner evidently reached the same practical result (cited it, then allowed the claims).

Element-by-element against the independent claims (1, 4, 7, 12, 14, 16):

Specific § 102 conclusions:

• Claims 1, 4, 7, 12, 14, 16 (all independent claims): US 4,779,105 does not appear to anticipate. The missing elements are principally (i) the host-supplied-definition-as-pointer into a stored library of character set definitions, and (ii) forms overlays and copy modifications merged into a page buffer. Under § 102, anticipation requires all elements in a single reference; these gaps defeat anticipation on the face of the record.
• Dependent claims 2–3, 5–6, 8–11, 13, 15, 17: No anticipation, for the same reasons (they add the buffer/overlay/copy-mod and host-generation structures the reference does not clearly show).
• § 103 relevance (why it still matters): US 4,779,105 is nonetheless the strongest starting point for an obviousness attack — it teaches the core architectural insight (host-facing emulation layer + translation to native raster for an all-points-addressable nonimpact engine + disk-stored fonts/formats + host-selected page format). A challenger would combine it with an art-recognized band-printer/UCSB pointer scheme to reach the '398 claims.

Why the examiner's allowance is consistent with this read: the '398's stated inventive insight is that a 512-byte UCSB cannot carry the 8K–52K of nonimpact character-definition data, so the UCSB entries are repurposed as pointers into a disk-resident resource library (loaded via the Diagnostic Write X'05' channel command in 8K blocks). US 4,779,105 stores fonts on a floppy disk, but it does not disclose that pointer-based indirection from a size-constrained host buffer — that is the gap that distinguishes the claims.

Explicitly excluded (not prior art / not citations) — to prevent conflation

Supplementary context — not of record (flagged): a U.S. Postal Service technical report, DTIC AD-A140587, describes a formatter/controller for a Delphax printer that accumulates plot data in page memories and "or's" stored graphics/logo data from disk with the text page data — conceptually adjacent to the '398's forms-overlay merging. This document is not a citation in the '398 file, I have not confirmed its public-availability date as of 1990-05-21, and it should not be treated as prior art of record without an independent prior-art date investigation. I mention it only because it surfaced in search and would be worth checking if a validity challenge were mounted.

Confidence and caveats

• High confidence that US 4,779,105 A is the only patent citation of record for US 5,025,398 (confirmed by the "Patent Citations (1)" listing and the full text).
• Moderate-to-high confidence in the non-anticipation conclusion, based on the reference disclosure as retrieved. I did not obtain the full text of every claim of US 4,779,105, so a genuinely complete § 102 element-mapping would require pulling its full claim set and figures from the issued PDF/PATFT.
• I did not query USPTO PAFT/PatentCenter live this session; if the goal is a formal file-wrapper antecedent basis (e.g., examiner's stated reasons for allowance, or any references cited during prosecution but not printed on the face), that requires a PatentCenter/IFW pull. I cannot confirm whether the examiner's allowance was supported by grounds beyond the single printed citation — formally, I do not know that US 4,779,105 is the only reference ever considered; it is the only cited reference printed in the record.
• Per the standing instruction, I have not auto-corrected identifiers: the granted claim 1 text's "lien printer setup data" is left as-is, the JP 5025398 B2 is kept distinct, and US 4,779,105 is kept distinct from the '398.

I'll ground the § 103 analysis in the actual prior-art references, starting with the one reference cited on the face of the patent and the "Similar Documents" set.

Good — US 4,779,105 is directly on point. Let me pin down its disclosure more precisely and check for secondary references on forms-overlay/copy-modification and printer-emulation art.

Obviousness Analysis — U.S. 5,025,398 Under 35 U.S.C. § 103

0. Framing note (builds on the prior sections)

The bibliographic and litigation sections established that the '398 patent has a single priority/filing date of 1990-05-21, issued 1991-06-18, and lapsed for non-payment of maintenance fees effective 2003-06-18. Two consequences shape everything below:

1. Because filing and priority coincide, all prior art must predate 1990-05-21. The "Cited By" lists on the Google Patents page (16- and 25-entry versions — note the source page shows both counts inconsistently) are not usable § 103 art. Their earliest priority date is US 5,542,050 (1990-09-28, Fuji Xerox), which post-dates the '398 filing by four months. Every other citing document is 1991 or later. They are evidence of the field's later trajectory, not of obviousness.
2. The patent is unenforceable, so § 103 is now a defensive/validity-interest question (portfolio diligence, prior-art dating, freedom-to-operate history) rather than an infringement issue — this is consistent with the litigation section's finding of no asserted litigation.

Legal standard applied: pre-AIA § 103(a) (the AIA first-inventor-to-file provisions do not apply to an application filed 1990-05-21). The controlling framework is Graham v. John Deere as refined by KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) — a combination is obvious where the references are from the same field, address the same problem, and the combination yields no more than predictable results; an explicit teaching, suggestion, or motivation is sufficient but not necessary. I apply the KSR rationales (design incentive, known-technique substitution, predictable variation, "obvious to try") rather than a strict TSM test.

PHOSITA (my construction): a B.S.-level engineer or equivalent with 2-4 years in high-speed printer controller/I-O channel design circa 1989-1990, familiar with IBM System/360-370 OEMI channel protocols and with the IBM 3211/3800-class host print drivers, and with raster/bitmap page-memory print engines.

1. The prior-art set actually available

Flagged uncertainty: I could not verify the DTIC report's publication date or whether it was publicly accessible before 1990-05-21; treat it as a candidate § 102(b) printed publication requiring confirmation, not as settled art. Similarly, I was unable to retrieve a confirmed pre-1990 reference specifically directed to IBM 3800-class "forms overlay"/"copy modification" in this session (search budget exhausted). That gap matters to § 4 below.

2. Element-by-element mapping of the independent claims

Read-out: US 4,779,105 alone discloses or suggests the substance of claims 1, 4, 7, and 16 at the level of generality in which those claims are drafted — same field, same problem, same architecture (channel-facing emulation layer + printer-resident disk font/format library + raster translation to a nonimpact engine). The features US 4,779,105 does not clearly supply are (i) displacement of operator console font/format selection by an automatic host-driven selection path (the "independent of operator intervention" limitation), (ii) line-buffer/page-buffer assembly details, and (iii) copy modification.

3. Combination A — US 4,779,105 alone (single-reference obviousness)

Claims 7 and 16 are the most exposed. Claim 7's central limitation is an interface with (a) means for emulating a line printer that stores host setup/character-definition data, selects a stored character set, and selects a stored forms overlay and copy modification, and (b) means for converting line printer control signals into nonimpact print control signals. Delphax states the emulation purpose in haec verba and implements the conversion (host parallel data → native → raster scan) for an ion-deposition (nonimpact) engine. A POSITA reading 4,779,105 would find the claimed architecture, minus the enumerated forms-overlay/copy-modification selection, and would understand that Delphax's disk-resident "fonts and formats" already includes format/overlay-like resources.

Where the single-reference case is weakest: claim 7's express "means for selecting the one of said forms overlays… and… copy modifications" has no clean antecedent in 4,779,105, and claim 16's "print band definition data" is band-printer-specific while Delphax speaks generically of "any host computer." A single-reference rejection of claims 7/16 would likely be sustained only with a second reference or with reliance on the APA.

4. Combination B — US 4,779,105 + US 4,860,219 (primary combination)

This is the combination I would expect a validity challenger (or a reexamination request) to lead with, and it maps cleanly onto every independent claim.

4.1 Motivation to combine (articulated)

1. Same field, same problem. Both are high-speed printer controller architectures handling host-formatted data and generating page rasters for nonimpact engines. KSR makes same-field/same-problem correspondence a strong rationales anchor.
2. Express design incentive in the primary reference. US 4,779,105 itself states the control module "can be modified to emulate a peripheral printer to any host computer" and that the first layer is replaceable "to emulate a different printer for a different host computer." A POSITA seeking to improve that emulation front end has an express reason to look to better rasterizer/page-composition back ends — exactly what 4,860,219 supplies.
3. Complementary, non-overlapping teachings. 4,779,105 is strong on the channel-facing emulation layer and disk-resident font/format storage; 4,860,219 is strong on the back end — page memory, per-character font identification with pixel-matrix fetch, dual page memories, and raster forms storage merged into page readout, at 350 ppm. Combining an emulation front end with a proven rasterization back end is a predictable assembly of known components with expected results.
4. Substitution of a known technique to improve a known method. Replacing Delphax's console-driven font/format selection (console serial controller 76 "permits selection of font and format") with 4,860,219's automatic font identification and fetch satisfies the "independent of operator intervention" limitation using a known technique (KSR rationale: known technique, predictable result, no change in principle of operation).
5. Elimination of a manual step. Operator console selection is a human-in-the-loop bottleneck in an otherwise fully channel-driven subsystem; automating resource selection on the basis of host-supplied data is an ordinary engineering objective, not a leap.

4.2 Resulting claim conclusions

Dependents. 2, 3, 5, 6, 8-11, 13, 15 and 17 add only conventional implementation detail — buffer hierarchies (2, 9), formatting with the converted control signals (3, 6, 10, 13, 15), generic "activating means" (8), and host-side resource generation/transmission (11, 17). None of these has independent patentable weight under KSR; each is a predictable mechanical arrangement of the elements already in the independent claims, and the host-side generation utilities in claims 11/17 are admitted in the specification as the "resource creation utility 126" and "download utility 127."

5. Combination C — the "admitted prior art" problem (and the real soft spot)

The most damaging aspect of the '398 disclosure for § 103 purposes is not the cited reference — it is the specification's own admissions:

• The Background and Detailed Description describe, as existing technology, the IBM 3211-class host behavior of downloading FCB (256 bytes) and UCSB (512 bytes), transmitting band-identification data, and sending character-ID-plus-position data line by line per CCW. That is the entire "line printer setup data + line printer character definition data prior to print character data" requirement of claims 1, 4, 7, 12, 14 and 16.
• The specification further admits that "other parameters can be transmitted via entries into the universal character set buffer data, such as: forms overlay names, copy modification segment names and copy counts" — i.e., the overlay/copy-modification naming scheme the claims recite as part of the invention's environment.

Post-KSR, an applicant's own admitted art can be combined with a single reference. A POSITA given (a) the admitted line-printer protocol with UCSB/FCB overlays and copy modification, and (b) US 4,779,105's "emulate a peripheral printer to any host computer" + disk-resident "fonts and formats," arrives at the claimed subject matter with only routine engineering. The only genuinely inventive-looking step in the whole disclosure is the recognition that the 512-byte UCSB cannot carry 8K-52K of nonimpact character definitions, and therefore must be repurposed as a pointer/index into a disk-resident resource library ("the data stream independent printer 102 uses the universal character set buffer data as pointers to reference a set of nonimpact printer character definition data… stored… on disk drive memory 114").

That is the limitation a patentee would litigate. Two responses for a challenger:

• US 4,779,105 already teaches the functional equivalent at the system level: resources (fonts/formats) live on printer disk, and the host-supplied page format/"part of the output from the host computer" selects them. Using a small host-supplied identifier to index a larger printer-resident resource library is exactly the relationship the reference discloses; the size-mismatch rationale is a design consideration a POSITA would recognize and solve, not an unpredictable result.
• US 4,860,219 teaches the small-key-to-large-data-array relationship explicitly: the page memory stores characters "in encoded form," and the processing unit "identifies the font to which the retrieved character belongs and identifies which character has been retrieved" before fetching the pixel matrix from font storage. A short code indexing into a full character-definition store is precisely the claimed "character set selection data" step.

If the DTIC report (ADA140587) can be authenticated as a pre-1990 printed publication, it independently reinforces this — it describes a Delphax-oriented formatter/controller converting text and form data into plot lines, accumulating them in a page memory, merging graphics overlay data into a second page memory, and storing character sets in different fonts in memory modules.

6. Secondary considerations and rebuttal considerations

There is no evidence of record supporting objective indicia: nothing in the file history, the assignment record (single inventor→StorageTek assignment, reel 005314/0011), or the citation record suggests commercial success, long-felt need, failure of others, copying, or industry praise attributable to the claims. The single examiner citation (US 4,779,105) and the absence of any post-issue challenge mean there is no adversarial record developing secondary considerations either way. A patentee would have to build a nexus argument from scratch, and the patent's lapse in 2003 for non-payment works against any commercial-success theory.

Genuine nonobviousness hooks (where a challenger is exposed):

1. "Independent of operator intervention." US 4,779,105's disclosed font/format selection is at least partly console-driven ("console serial controller 76… permits selection of font and format"). A patentee will argue Delphax teaches away from fully automatic host-driven resource selection. The answer is US 4,860,219's automatic font identification, plus the admitted art in which UCSB/FCB loading is already automatic — so the limitation is met in the combination.
2. The pointer/index insight. No single reference retrieved here states the 512-byte-vs-8K-52K size-mismatch rationale in those terms. This is the strongest candidate for a nonobviousness argument, and it should be met with the KSR "design incentive" rationale plus 4,860,219's encoded-character/font-identification teaching.
3. Copy modification. Claim 1's memory means must store copy modifications, and claims 2/9 write copy modification data into the page buffer. Neither 4,779,105 nor 4,860,219 clearly discloses a copy modification (per-copy variable segment) as opposed to a form; I could not retrieve a confirmed pre-1990 copy-modification reference this session. Absent such a reference, this element is the hardest part of claim 1 to reach, though it is arguably satisfied by the specification's own admission that copy modification segment names were already UCSB-transmitted parameters, and by the KSR rationale that merging a second stored image segment is the same known technique as merging a form.

7. Bottom line

• Most likely outcome: Claims 1, 4, 7, 12, 14, 16 and their dependents 2, 3, 5, 6, 8-11, 13, 15, 17 are invalid under § 103(a) over US 4,779,105 in view of US 4,860,219, further in view of the line-printer protocol admitted in the '398 specification itself. The combination is a same-field, same-problem assembly with an express emulation motivation in the primary reference, complementary non-overlapping teachings, and predictable results.
• Strongest invalidity pressures, ranked: (1) the specification's own admissions of the UCSB/FCB/overlay/copy-modification environment; (2) US 4,779,105's express "emulate a peripheral printer to any host computer" + disk-stored fonts and formats; (3) US 4,860,219's encoded-character page memory, automatic font identification, raster forms storage, and dual page memories.
• The only plausible survivability argument rests on the narrow "UCSB-as-pointer to a disk-resident 8K-52K character library" concept and on the "independent of operator intervention" language — and even those are reachable through KSR rationales plus the 4,860,219 font-identification teaching, because the claims are drafted in broad means-plus-function form (which, if anything, widens them toward the prior art).

8. Caveats and verification still required

• I did not retrieve a confirmed pre-1990 reference specifically claiming IBM 3800-class forms overlay/copy modification. That retrieval is the single highest-value next step for a complete § 103 record; it would close the one element gap in claim 1.
• The DTIC report (ADA140587) is potentially powerful prior art but its publication date and public accessibility before 1990-05-21 are unverified; do not rely on it until authenticated.
• US 4,860,219 appears on Google's "Similar Documents" list — an algorithmic similarity flag, not an examiner citation. I have not verified its full claim set or its filing date (only its 1989-08-22 grant date) from a primary source in this session; a filing date more than one year before 1990-05-21 is what would make it § 102(b) rather than § 102(a)/§ 103 art. Its grant date alone, however, still places it within § 102(a)/§ 103 (invented before the '398 invention).
• Prior-art dates above are drawn from the fetch text at https://patents.google.com/patent/[US5025398](/patent/US5025398)/en and https://patents.google.com/patent/US4779105, plus https://patents.google.com/patent/[US4860219A](/patent/US4860219A)/en and https://www.freepatentsonline.com/4860219.html. Where a date or disclosure was not visible in those sources, I have labeled it as an assumption rather than stating it as fact.