US 5745681

Patent Overview — US5745681A ("Stateless shopping cart for the web")

• Patent number: US5745681A (application US08/583,877)
• Title: Stateless shopping cart for the web
• Inventors: Fredrick E. Levine; Bruce A. Carter
• Original assignee: Sun Microsystems, Inc.; current assignee listed as Oracle America Inc
• Priority/filing date: 1996-01-11; publication/grant date: 1998-04-28
• Legal status: Expired – Lifetime; anticipated expiration 2016-01-11
• Source: https://patents.google.com/patent/[US5745681](/patent/US5745681)/en

Purpose. The patent addresses problems in earlier network shopping carts, which required a user to log in or hold an account, stored user identity or account data on a remote server, and risked losing a session (e.g., overnight server cleanup deleting an in-progress item collection). The invention instead makes the server-side shopping process stateless: the server does not maintain a list of selected items.

Plain-language scope. A browser at a client station requests a shopping page; a "market" application at the server returns a shopping page file with selectable items. When the user adds items, the browser sends an add request containing the selected items. The server determines whether a previously selected-items list exists, creates/updates one, converts it to a hidden "cart" field, embeds that field in a new shopping page file, and returns it to the browser. On the next add request, the browser returns the cart field together with newly selected items, and the server rebuilds the cart list from both. Maintenance of the cart field is therefore under the user's control via the browser at the client station, so the server bears no cart-maintenance burden and needs no persistent per-user account. The described embodiment concerns items such as software files (e.g., device drivers) to be downloaded.

What the excerpt does not establish. It does not supply the full claim language or complete specification, the precise claims' scope, the full reference numerals/figure details (FIGS. 1–6 appear only as thumbnails/partial text), the complete list of cited prior art, or any litigation outcome, ownership-transfer detail, validity, or infringement analysis. Numerical identifiers are reported as listed; no corrections are made.

I'll search litigation databases for US patent 5745681. Let me run multiple targeted queries.

The general searches are returning citation artifacts. Let me query the litigation databases directly.

Key finding: the EDTX case is AML IP v. Hallmark Cards asserting US 6,876,979, not the '681 patent. Let me verify whether 5745681 appears in any docket directly, and check other litigation databases.

Litigation search results for US 5,745,681

I searched the Unified Patents litigation portal, RPX Empower/litigation trackers, the EDTX docket record, PTAB file histories, and general web/patent databases for the specific number 5745681 (and "5,745,681"). Below is what the record actually shows.

---

Bottom line

No litigation asserting U.S. Patent 5,745,681 has been verified. The single litigation link carried on the Google Patents record for this patent resolves to a case that asserts a different patent. Every other hit for the string "5745681" was a citation artifact (the number appearing on an IDS/Form PTO-1449 or a search-report reference list), not a case caption.

---

The one flagged case — and why it is NOT about the '681 patent

Google Patents' "Family has litigation" flag for US5745681 links to:

Critical point: the amended complaint in that case expressly pleads infringement of "the '979 patent" (U.S. 6,876,979), an AML IP e-commerce patent. US 5,745,681 is not the patent-in-suit. The Google Patents/Darts-IP "family has litigation" tag appears to be a dataset mis-association (a family-level or portfolio-level link), not evidence that '681 was ever asserted.

Related Federal Circuit appeal (again, NOT the '681 patent): AML IP's case against Bath & Body Works was dismissed for improper venue (Rule 12(b)(3)) and patent ineligibility (Rule 12(b)(6)); the Federal Circuit affirmed, holding a court may address a 12(b)(6) merits ground after finding venue improper. Reported via Knobbe Martens and Lexology briefs. The article is dated September 18, 2026, which post-dates the April 26, 2026 date you supplied — I flag that discrepancy rather than reconcile it. This appeal concerns the '979 patent, not '681.

---

Why the absence of '681 litigation is plausible

• Expired 2016-01-11. Google Patents lists the legal status as "Expired – Lifetime," with anticipated expiration 2016-01-11 (20 years from the 1996-01-11 filing). An expired patent supports only past-damages theories and cannot be the basis for injunctive relief — an unattractive NPE target.
• Ownership chain: Sun Microsystems, Inc. (original assignee, 1996) → Oracle America, Inc. (reassignment recorded 2015-12-11, merger/change of name). No later transfer to a litigating entity was verified.
• It is foundational prior art, not an assertion asset. Unified Patents shows 718 "Referenced By" documents; Google Patents shows 338+ "Cited By." It is heavily used against other patents (e.g., its number appears in the CBM2016-00047 file history for U.S. 8,041,626, and in IPR2018-01009 exhibits) — but being cited is the opposite of being asserted.

---

Search artifacts I ruled out (do not treat these as litigation)

• "5745681" inside CBM2016-00047 (U.S. 8,041,626, Kirwin) file-history exhibits — it is a listed reference, not the challenged patent.
• "5745681" inside IPR2018-01009 Exhibit 2005 — a patent-number list in a file history.
• Various EPO/INPADOC and PCT search reports listing US5745681 as a family/cited document.
• Justia/patent mirror pages listing "5745681" in "References Cited" tables.

None of these is a case caption identifying 5,745,681 as a patent-in-suit.

---

What I could not verify (stated explicitly, not asserted)

1. No parties, asserted claims, or outcome can be reported for any case with 5,745,681 as the patent-in-suit, because I found none. I will not manufacture a caption.
2. The Darts-IP family-litigation entry (family ID 24334960) resolves to the same AML IP v. Hallmark chain; I could not retrieve a case list from Darts-IP directly. It cannot be relied on as '681-specific without pulling the docket.
3. I could not confirm the final per-defendant disposition of EDTX 4:22-cv-00214 from the sources retrieved.
4. My searches were executed against public portals; a case filed and not yet ingested, or a state-court/settlement demand never docketed, would not appear. Absence of a hit is not proof of absence.

---

Answer to your question

• Known litigation involving US 5,745,681: none verified. The plaintiff/defendant/jurisdiction/case-number fields you asked for have no entries that identify 5,745,681 as the asserted patent.
• The one case linked to this patent's record — AML IP LLC v. Hallmark Cards, Inc. et al., EDTX No. 4:22-cv-00214, filed 2022-03-18 (Judge Sean D. Jordan) — asserts U.S. 6,876,979, not 5,745,681. Treat the '681 litigation flag as an unverified dataset association pending confirmation on the docket itself.
• If you need a definitive negative, pull the docket for 4:22-cv-00214 and confirm the patent-in-suit, and run a plaintiff-name/patent-number query on RPX Empower or Docket Alarm for "5,745,681" filtered to exclude citation tables.

I'll verify the ODP baseline against public sources before writing anything.

The ODP baseline says "none," but let me stress-test that against PTAB dockets, the EDTX case, and Unified Patents' tracker.

Proceedings overview

Total AIA trial proceedings on US 5,745,681: 0. The USPTO Open Data Portal returns no IPR, PGR, or CBM petition — instituted or otherwise — against this patent, and my independent searches of PTAB dockets, the Unified Patents patent record, and public litigation trackers surfaced no proceeding that I can verify by number (breakdown: 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution denials; all 21 claims are untested before the PTAB). Bottom line for a defendant: there is no IPR-based shortcut here. You cannot point to a canceled claim, and you cannot ride anyone else's estoppel — but you also face a patent whose entire asserted life is behind it, because US 5,745,681 expired on 2016-01-11 (Google Patents: "Anticipated expiration"; legal status "Expired - Lifetime") after running its full 20-year term from the 1996-01-11 filing.

No proceedings to report

There is no {PROCEEDING_NUMBER} — {Petitioner} v. {Patent Owner} section to write, because there is no proceeding. I will not manufacture one. What I can report precisely is what the record does show:

• PTAB (USPTO ODP, canonical): no AIA trial proceedings on file as of the most recent ingest.
• Independent search check: queries against PTAB petition documents, PTAB/CBM filing reports, the Unified Patents patent page for US-5745681-A, and CourtListener-style dockets returned only noise hits — i.e., other patents' IDS/Form PTO-1449 listings that cite "5745681" as a reference, not as the patent under review. For example, the hits in IPR2018-01354 and CBM2016-00047 are exhibits quoting IDS forms; the IPR2018-01354 exhibit addresses a "'601 Patent" with challenged claims running into the 60s (claims 1–3, 6–8, 12, 14–16, 19–21, 25, 27–29, 32–34, 38, 40–42, 45, 47, 49, 51–53, 56, 57, 60–62, 66). US 5,745,681 has 21 claims, so that document cannot be about this patent. Treat any search-engine hit pairing "5745681" with an IPR number as a citation artifact until you open the paper and see the patent number in the caption.
• Litigation flag (non-PTAB): the Google Patents record carries a "Family has litigation" entry pointing to a Texas Eastern District Court case, 4:22-cv-00214, via Unified Patents' litigation dataset (link), plus a Darts-IP worldwide-family litigation entry (link). I could not verify the parties, the asserted claims, or the outcome of that case, and the patent's 2016-01-11 expiry makes a 2022 assertion unusual (consistent only with a past-damages theory, a family member, or a mis-tagged docket). Do not rely on my silence about it — pull the docket before you respond to any demand.

Strategic summary

Claim status: everything is UNTESTED. No claim of US 5,745,681 has been canceled, confirmed, or construed by the PTAB in an AIA trial. The patent issued 1998-04-28 with 21 claims — independent claims 1 (method of supplying items), 7 (method of user shopping), 11 (stateless shopping cart apparatus in a server computer), 15 (apparatus for emulating a shopping cart metaphor across server and browser), and 19 (computer program storage medium) — and they stand exactly as granted. The prosecution history of record cited several non-patent-literature items now listed on the Unified Patents page (Hauser et al., Generic Extensions of WWW Browsers (First USENIX Workshop on Electronic Commerce, July 1995); Rowley, Retailing and Shopping on the Internet (1996); Fried-Cassorla, Successful Marketing on the Internet (Direct Marketing, Feb. 1995)). That is a § 325(d) double-edged sword: a petitioner re-using that art faces discretionary denial under the Advanced Bionics framework, while any new petition must bring art not previously before the Office. Note I am relying on the reference list; I have not confirmed which of those items were the basis of any Examiner rejection.

Estoppel landscape: empty. Because no IPR/PGR/CBM was ever instituted, there is no § 315(e)(2) estoppel binding anyone, and no petitioner is barred from asserting prior-art grounds in district court. Conversely, a defendant today has no free ride on someone else's invalidity win. Practically, the invalidity case is wide open — and unusually favorable in one respect that has nothing to do with the PTAB: with an effective filing date of 1996-01-11 and a 1998-04-28 grant, the entire world of 1994–1996 web-commerce art (e.g., pre-1996 CGI shopping-cart implementations, the Hauser USENIX paper) is available as § 102(b) prior art, and there is no AIA § 102 grace-period complication because this is a pre-AIA patent. Expired-patent status also means claims are construed under Phillips, and there is no possibility of claim amendment to escape art — a one-way ratchet for a challenger. Any IPR you filed now would still be institution-eligible (IPRs are available against expired patents) and would run on the statutory clock — institution decision within ~6 months of petition, FWD within 1 year of institution under § 316(a)(11) — but the strategic value is thin: with the patent expired, your real invalidity leverage is an § 282 district-court defense and, if you want a cancellation decree, an IPR whose practical payoff is limited to stopping past-damages theories.

Pattern signals: none. There is no serial petitioner (no second, third, or fourth IPR against this patent), no Patent Owner appeal activity at the Federal Circuit, and no defensive aggregator in the chain — the Unified Patents entry on this patent is a passive tracking page, not a filed challenge; Unified's CBM/IPR docket does not include a petition against 5,745,681. Ownership runs Sun Microsystems, Inc. (original assignee, filed 1996-01-11) → Oracle America, Inc. (reassignment recorded 2015-12-11, merger/change of name), which matches the Google Patents "Current Assignee" field; I have not independently verified whether any later transfer occurred. The foreign family (EP0784279B1, EP0807891B1, DE69600368T2, DE69608417T2, JPH09296819A) is a separate story — EP0784279/EP0807891 were the subject of the 2000-era European software-patent controversy (the FFII/Swpat pages reproduce the EPO claims) — but I have not verified any EPO opposition or revocation outcome, so do not carry that into a US defense.

Recommended next steps

• Do not plead a PTAB-based defense. There is no FWD to cite and no canceled claim to quote. If a demand letter or complaint asserts claims 1–21, the PTAB gives you zero argument; build your case on § 102/§ 103 against 1994–1996 art and on the expired-patent posture (no injunctive relief lever, past damages only).
• Confirm the litigation flag before anything else. Pull EDTX 4:22-cv-00214 (Unified Patents link) and the Darts-IP family entry (link) to identify the parties and compare the asserted claims against the 21 granted claims. I could not name the parties from public search and will not guess.
• Verify the correspondence address / current owner through USPTO PatentCenter and the assignment record (the 2015-12-11 Oracle America reassignment) before serving any notice or stipulation. Google's assignee field and USPTO's assignment chain are not always in sync post-2016.
• Re-run the PTAB check at the source if you file: USPTO PTAB E2E and the ODP API. My finding is a point-in-time check, and a petition filed after the last ingest would not appear here.
• Preserve the record. Because there is no PTAB proceeding, there is also no estoppel and no IPR-driven claim construction to inherit — a defendant's best artifact is a clean, non-cumulative prior-art chart, not a trial record.

Confidence and limits. High confidence that no AIA trial proceeding exists on US 5,745,681 (ODP is silent and multiple independent public-source checks agree). Low confidence on the EDTX 4:22-cv-00214 details — flagged, not asserted. I found no claim-level outcome, judge panel, settlement, or Federal Circuit appeal to report because there is no proceeding from which such facts could arise; anything more specific would be fabricated.

I'll research the assignment chain and litigation for US 5745681.

Key lead: the EDTX case is AML IP LLC v. Hallmark Cards, Inc. et al. Let me dig into whether this patent is in that NPE's portfolio.

AML IP LLC is a confirmed NPE. Now I need the assignment records showing how this patent got to it (if it did).

I'll search for authoritative information on US 5745681 specifically — its USPTO record and, critically, the prior-art references cited on its face.

I have the front-page "References Cited" data. Let me verify details and dates on the key cited references, especially the most relevant ones.

Prior Art Cited for US 5,745,681 — Analysis

Patent under analysis: US 5,745,681 (US5745681A), "Stateless shopping cart for the web," App. No. 08/583,877, filed 1996-01-11, granted 1998-04-28, Levine & Carter, originally Sun Microsystems, Inc.

---

0. Important scope caveat before the analysis

Two things must be flagged up front, because "the citations for 5745681" is ambiguous and the two sets mean opposite things:

1. Backward citations ("References Cited" / prior art) — this is what your task asks for. The patents cited against US 5,745,681 during prosecution. That set is ten U.S. patents plus three non-patent publications, listed below.
2. Forward citations ("Cited By") — NOT prior art to this patent. Google Patents shows 338+ "Cited By" documents and Unified Patents shows 718 "Referenced By." These all post-date 5,745,681 and cite it as prior art. They cannot anticipate it under § 102 and must not be confused with the references-cited set. (Note that the full patent text provided in the earlier section contains only the "Cited By" list — it does not contain the front-page "References Cited" list, which is why this task required going to the USPTO-derived mirrors. See § 5, confidence notes.)

Legal standard applied: Anticipation under 35 U.S.C. § 102 requires that a single reference disclose every element of a claim, arranged as claimed. A reference that discloses some elements but not all is § 103 material, not § 102 material. I apply that standard strictly rather than labelling every cited reference an "anticipation."

---

1. The ten U.S. patent references cited on the face of US 5,745,681

Source: USPTO front-page data as mirrored by uspto.report (https://uspto.report/patent/grant/5745681) and Unified Patents (https://portal.unifiedpatents.com/patents/patent/US-5745681-A), with per-reference corroboration from Google Patents. The USPTO front page lists these in reverse-chronological grant order; I have re-ordered by relevance to the claimed subject matter.

R1. US 5,623,656 (US5623656A) — the closest reference

• Full citation: US 5,623,656, "Script-based data communication system and method utilizing state memory," Kenneth B. Lyons, filed/priority 1994-12-15, granted 1997-04-22, assignee AT&T / Lucent Technologies Inc.
• Status relative to 5,745,681: Pre-dates the 1996-01-11 filing → § 102(a)/(b)/(e) art, depending on the publication event relied on.
• Description: A preprocessing script system for HTML Fill-In-Form ("HTML-D") communications that embeds information about the previous state of the system inside the script data itself, so that a client can have one script influence another. The specification expressly frames the problem as prior HTML FIF systems being "stateless" — "when a server or a client computer begins to process a received script or form, it has no information as to what that particular form is" — and states the invention "effectively imposes a state memory upon what would otherwise be a stateless system… without the need for storing state information on a data system server, thereby providing increased system security." Form field data is returned to the server as data blocks each with "an associated identifier," and the server's preprocessor reconstructs/redirects the next HTML-D script from that returned data. (https://patents.google.com/patent/US5623656A)
• § 102 assessment — claims potentially anticipated: None, in my assessment. R1 is the only cited reference that reaches the core insight of 5,745,681 (server is stateless; state travels with the client-returned page/script; no server-side account retention). That concept maps most directly onto the "circulating list / stateless server" limitations of independent claims 1, 7, 11, 15 and 19, and onto the name/value-pair and "cart" field parsing of dependent claims 5, 6, 8. But R1 discloses no shopping cart, no user selection of items to purchase/download, no "add request" containing a first field of new selections and a second field of previous selections, and no cart list accumulation across round-trips. Because at least those elements are missing, R1 cannot § 102-anticipate any claim as a whole. It is, however, the single most pertinent reference for a § 103 obviousness challenge to the "stateless" framing of claims 1/7/11/15/19 when combined with any online-shopping reference (R3/Rowley, R8/Chen, or R10).

R2. US 5,500,929 (US5500929A) — browsing and selecting networked resources

• Full citation: US 5,500,929, "System for browsing a network resource book with tabs attached to pages," Robert D. Dickinson, filed/priority 1993-08-30, granted 1996-03-19, assignee Taligent, Inc. / Apple.
• Status: Filed 1993-08-30, granted after the 1996-01-11 filing → § 102(e) art; its 1996-03-19 grant is after filing so it is not § 102(b) on its face.
• Description: A network-accessible "resource book" that lets a user browse a hierarchy of network resources, select one or more desired resources, and place the selected resources on the user's own workstation "for subsequent retrieval and usage." Resources may then be used to implement a task "at the direction of the user." (https://patents.google.com/patent/US5500929)
• § 102 assessment: The disclosure touches the "items selectable by a user using the browser" element of claims 7, 11 and 15, and the retrieve-and-deliver-selected-items concepts of dependent claims 4, 10, 14, 18 and 21. But it discloses neither a cart list, nor a request that carries previously selected items back to a server, nor a returned hidden cart field. Not anticipatory of any claim; relevant at most as a § 103 secondary reference or as general background for "select resources now, retrieve later."

R3. NPL — Rowley, "Retailing and Shopping on the Internet"

• Full citation: Jennifer Rowley, "Retailing and Shopping on the Internet," Internet Research, MCB University Press, UK, vol. 6, No. 1, 1996, ISSN 1066-2243, pp. 81–91.
• Status: 1996 journal issue. Caveat: whether this issue was actually published before 1996-01-11 is not established by the sources I retrieved; if it published after that date it is not § 102(b) art against this filing. Verify the issue date before relying on it.
• Description: A survey of retailing/shopping over the Internet — cited by the examiner for the general concept of online shopping.
• § 102 assessment: Supplies only the general "shopping on a network" context. It does not disclose the stateless cart architecture of claims 1, 7, 11, 15 or 19. Not anticipatory; contextual/§ 103 only.

R4. US 5,625,781 (US5625781A) — building a list without visiting the sites

• Full citation: US 5,625,781, "Itinerary list for interfaces," Cline et al., filed/priority 1995-10-30, granted 1997-04-29, assignee International Business Machines Corp.
• Status: Filed 1995-10-30, granted after filing → § 102(e) art.
• Description: Creating a list of links (an "itinerary") without first opening/visiting each site and without manually entering each address — i.e., accumulating a persistent list of user-selected network destinations for later use.
• § 102 assessment: Relevant to the abstract "creating a list" and "list of selected items" language in claims 1(d)–(f), 7 and 19. It does not disclose a browser carrying that list back to the server in a form field, nor a server returning an updated list in a page. Not anticipatory; possible § 103 support for the "list-building" element.

R5. US 5,572,643 (US5572643A) — browser-side storage/history of pages

• Full citation: US 5,572,643, "Web browser with dynamic display of information objects during linking," David H. Judson, filed/priority 1995-10-18, granted 1996-11-05.
• Status: Filed 1995-10-18, granted after filing → § 102(e) art.
• Description: A Web browser that locally stores, retrieves and outputs information objects (including previously fetched pages) to reduce download latency during linking. This is the reference underlying the browser-history/local-page-storage aspects of the specification's discussion of stored shopping page files.
• § 102 assessment: Touches the specification's observation that the browser stores page files and can return to a page containing the cart field — relevant to the "storing at least the shopping list by said browser" element of claims 7, 15 and 20. It discloses no cart field, no add request, no server-side cart reconstruction. Not anticipatory; § 103/background only.

R6. US 5,305,440 (US5305440A) — client-side maintenance of data returned by a server

• Full citation: US 5,305,440, "File extension by clients in a distributed data processing system," Morgan et al., filed/priority 1989-05-15, granted 1994-04-19, assignee International Business Machines Corp.
• Status: Granted more than one year before the 1996-01-11 filing → § 102(b) art.
• Description: In a distributed network, a client requests a range of bytes from a server; the server replies with the requested bytes plus an additional "nascent zeros" range that the client is permitted to write into and keep, with the server reserving (and later revoking) the space. The client modifies and maintains the returned data locally rather than the server retaining per-client state. (https://patents.google.com/patent/US5305440A)
• § 102 assessment: Conceptually the nearest § 102(b) reference on the "server hands data to the client, client retains/modifies it, server does not keep per-client state" theme — arguably pertinent to the circulation/return-of-updated-list limitation of claims 1(e)–(h) and to claim 20 (server releasing its storage). However it is a memory-mapped file-system invention with no browser, no page file, no item selection and no cart. Not anticipatory of any claim; strongest as a § 103 reference on the client-side-state-maintenance element.

R7. US 5,630,068 (US5630068A) — the "shopping cart" metaphor source

• Full citation: US 5,630,068, "Shoppers communication system and processes relating thereto," Vela et al., priority 1987-10-13, granted 1997-05-13, assignee Klever Marketing, Inc.
• Status: Priority 1987-10-13, granted after filing → § 102(e) art (very early priority).
• Description: A physical in-store shopping cart fitted with message relay units that deliver audio/video promotional messages to the shopper as the cart moves through the store, controlled by a master computer. (Cited, I infer, because the examiner was identifying the "shopping cart" metaphor itself.)
• § 102 assessment: Antedates the Internet entirely and discloses no browser, no network page files, no add request, no cart field. Not anticipatory of any claim; it is at most a § 103 reference for the metaphor/word choice, which is legally thin for the claimed architecture.

R8. US 5,590,197 (US5590197A) — client-held account/state (no server-side account retention)

• Full citation: US 5,590,197, "Electronic payment system and method," Chen et al., filed/priority 1995-04-04, granted 1996-12-31. (Unified Patents lists priority 1995-04-03; Google/others list 1995-04-04 — both recorded as found.)
• Status: Granted after filing → § 102(e) art.
• Description: A "cyber wallet" in which protected account information is carried on the customer's own device or alongside the browser/mosaic software, enabling payment to a merchant at a remote Internet site without the server storing the user's account information. (https://patents.google.com/patent/US5590197A)
• § 102 assessment: Directly relevant to the specification's stated motivation (avoiding server-stored user accounts that get purged overnight) and to the "information about who the user was is not retained" theme of claims 1, 7, 11, 15 and 19. It does not disclose the cart list/field circulation, and it is a payment scheme, not a selection/shopping-list scheme. Not anticipatory; a plausible § 103 companion piece.

R9. US 5,522,045 (US5522045A) — distributed shared memory

• Full citation: US 5,522,045, "Method for updating value in distributed shared virtual memory among interconnected computer nodes having page table with minimal processor involvement," Sandberg, filed/priority 1992-03-26, granted 1996-05-28, assignee Matsushita / Panasonic Corp. of North America.
• Status: Granted after filing → § 102(e) art.
• Description: Updating values in distributed shared virtual memory across interconnected nodes via page-table management. General distributed-systems infrastructure.
• § 102 assessment: Pertinent only to the generic client/server hardware recitation of claim 11 (processor, memory, I/O, storage, network). It discloses nothing about requests, pages, lists or carts. Not anticipatory; boilerplate § 103/background art at best.

R10. US 5,621,892 (US5621892A) — networked event/alert management

• Full citation: US 5,621,892, "Method and apparatus for managing alerts and events in a networked computer system," Cook, filed/priority 1995-10-09, granted 1997-04-15, assignee Intel Corp.
• Status: Granted after filing → § 102(e) art.
• Description: Managing alerts and events across a networked computer system. General network-management subject matter.
• § 102 assessment: No shopping, browser, page-file, cart or list disclosure bearing on claims 1, 7, 11, 15 or 19. Not anticipatory of anything; cited as general networked-computing background.

R11. US 5,623,589 (US5623589A) — incremental browsing of content levels

• Full citation: US 5,623,589, "Method and apparatus for incrementally browsing levels of stories," Needham et al., filed/priority 1995-03-30, granted 1997-04-22, assignee Intel Corp.
• Status: Granted after filing → § 102(e) art.
• Description: Incrementally browsing successive levels of a story/content hierarchy.
• § 102 assessment: At most general background for the specification's reference to a browser "history of pages through which the user may search." No cart/list/selection disclosure. Not anticipatory; no meaningful § 102 or § 103 weight against the independent claims.

---

2. The non-patent literature cited on the face (complete list)

(Unified Patents lists the NPL entries twice in slightly different punctuation/formatting — e.g., "Fried-Cassorla" vs. "Fried Cassorla" — the duplicates are listed as found, not as separate references.)

---

3. Overall § 102 conclusion for US 5,745,681

• No reference cited on the face of US 5,745,681 anticipates any of claims 1, 7, 11, 15 or 19 in my assessment. Each independent claim requires, as a combination: a browser-issued shopping-page request; a server-generated selectable-item page file; an add request carrying both newly selected items and previously selected items in a field; server-side creation or reconstruction of a cart list, addition of the new items, and return of the updated list embedded as a field in a returned page file; and repetition of that circulation with the client retaining the cart. No single cited reference discloses that combination — which is consistent with the fact that the patent issued with all 21 claims intact over these references.
• The only reference that reaches the distinguishing "stateless" concept is R1 (US 5,623,656, Lyons). It discloses embedded state inside scripts precisely to avoid server-side state storage — the same architectural idea — but in a forms/preprocessor context with no items, no selections and no cart. Its practical significance is as the lead § 103 reference in combination with an online-shopping reference (N3/Rowley or R8/Chen): together they would supply "stateless script with embedded state" + "shopping on the Internet." That combination is exactly the kind of argument the examiner evidently considered and rejected in allowing the claims.
• Ranked relevance for any future § 102/§ 103 challenge: R1 (5,623,656) > R6 (5,305,440) ≈ R8 (5,590,197) > R2 (5,500,929) > R5 (5,572,643) ≈ R4 (5,625,781) > R3/Rowley > R7 (5,630,068) > N1/Hauser > R9, R10, R11 (background only).

---

4. Notable observation: a § 102(e)-class reference not on the face

US 5,826,242 ("Method of on-line shopping utilizing persistent client state in a hypertext transfer protocol based client-server system," Netscape Communications Corp.), which appears in the forward-citation list for 5,745,681, carries a priority date of 1995-10-06 — i.e., before the 1996-01-11 filing of 5,745,681 — and therefore could qualify as § 102(e) prior art. It was not among the references cited on the face of 5,745,681. I flag this as a candidate for further investigation; I have not verified its cited filing/priority basis or whether it was before the examiner in a different form, so treat it as a lead, not a finding. (Its title — "persistent client state" — suggests cookie-based server-associated state, which is arguably the opposite architecture from the "stateless" cart claimed here.)

---

5. Sources, discrepancies, and confidence

Sources used: USPTO front-page data as mirrored by uspto.report (https://uspto.report/patent/grant/5745681) and Unified Patents (https://portal.unifiedpatents.com/patents/patent/US-5745681-A); Google Patents pages for each cited reference (US5623656A, US5305440A, US5500929, US5590197A); and the EPO/EP0784279B1 family document listing the same U.S. family member. I could not query the USPTO's own full-text/PatentCenter system directly in this session; the front-page "References Cited" list is therefore reproduced from authoritative-mirror sources, not from a primary USPTO API response.

Identifiers reported literally (no auto-correction applied):

• Patent number for the subject: 5745681 / US 5,745,681.
• Application number: mirrors and Google list 08/583,877; the earlier analysis flagged a mirror showing "US 5838770." Recorded as found.
• Inventor spelling: "Fredrick E. Levine" appears in the USPTO/Unified Patents record; "Frederick E. Levine" appears elsewhere. Both recorded as found.
• Priority/filing date: Google Patents and USPTO-derived sources say 1996-01-11; Unified Patents says 1996-01-10 (and grant 1998-04-27 vs. 1998-04-28). Recorded as found; this one-day spread does not change any § 102 date conclusion here except at the margins.
• US 5,623,656 filing/priority: Google says 1994-12-15; Unified Patents says 1994-12-14. Recorded as found.
• US 5,590,197 priority: 1995-04-04 (Google/searchlight) vs. 1995-04-03 (Unified Patents). Recorded as found.

Confidence: High for the identity of the ten cited U.S. patents and three NPL items and for their titles/assignees (cross-corroborated across three sources). Medium for exact grant/filing day values (mirror-sourced; may be off by a day). Medium-low for the per-reference § 102 mapping, because it is my element-by-element judgment of what each reference's summary/abstract-level disclosure supports — I did not perform a full-text, claim-chart analysis of each reference's complete specification, and several (especially R9–R11) I reviewed only at the summary level. Low for N1 (Hauser) content — I have not verified what that paper actually discloses. Explicitly uncertain: the actual publication date of N3 (Rowley, 1996) relative to the 1996-01-11 filing, and the provenance of the off-face candidate in § 4.

I'll research the prior art references associated with this patent before building the §103 analysis.

Now let me research the examiner-cited prior art references and any validity challenges to this patent.

Obviousness Analysis — U.S. Patent 5,745,681 under 35 U.S.C. § 103

Note on framing: 5,745,681 is a pre-AIA patent (filed/priority 1996‑01‑10 or 11, depending on source — see the discrepancy flagged in the prior section), so pre‑AIA § 103(a) and pre‑AIA § 102(e) apply. Per KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), and Graham v. John Deere Co., 383 U.S. 1 (1966), I analyze (i) the level of ordinary skill, (ii) the scope/content of the prior art, (iii) the differences, (iv) motivation with reasonable expectation of success, and (v) objective indicia. All claim text is from the authoritative full text at https://patents.google.com/patent/[US5745681](/patent/US5745681)/en.

---

1. Person of ordinary skill in the art (PHOSITA)

Given the 1996 priority date, the field, and the patent's own hardware/software description (SPARC/UNIX workstations, HTTP, HTML, cgi-bin scripts — see the "ds.cgi" example in the specification), the PHOSITA is someone with a B.S. in computer science/electrical engineering plus ~2 years' experience (or an M.S. plus ~1 year) in network client-server programming, HTTP/HTML form handling (including CGI), and on-line catalog/order-entry systems. This is the level at which a single artisan would have knowledge of: (a) HTTP's statelessness, (b) HTML FORM/POST/GET mechanics and hidden fields, (c) the newly introduced cookie mechanism, and (d) interactive teleshopping/catalog systems.

---

2. Scope and content of the prior art

(a) The patent's own admitted prior art (Background)

The specification makes several express admissions that are prior art against the claims:

• Shopping-cart systems that "require that the user be uniquely identified," via log-in/account stored on the server, with the identifier carried "either in the uniform resource locator (URL) or as hidden data passed back through a form submission." This is an admission that both URL-embedded state and hidden-form-field state were known mechanisms for associating server-side data with a user.
• Server-side retention causes "resource constrained" servers to purge carts ("nightly administrative clean up … has therefore deleted the previous evening's collection of items").
• The claimed "market program" is a CGI program (http://gemini.west//dx/…/cgi-bin/ds.cgi), and the cart is implemented as <INPUT TYPE=HIDDEN NAME="cart" VALUE="AT&T GIS NCR 53C815"> (Fig. 7C, entry 252).

Significance: the patent's point of novelty is where the cart lives (client-managed rather than server-managed) and how it is carried (page-embedded field rather than server account/ID). The Background concedes that the carrying mechanisms themselves (URL, hidden fields) were conventional.

(b) Montulli / Netscape — the closest prior art

• U.S. 5,774,670 (Montulli; Netscape Communications), filed Oct. 6, 1995 — "Persistent client state in a hypertext transfer protocol based client‑server system" (https://patentimages.storage.googleapis.com/8f/31/db/db36c823b1f2d3/[US5774670](/patent/US5774670).pdf).
• U.S. 5,826,242 (Montulli; Netscape), filed Aug. 27, 1997 as a divisional of Ser. No. 08/540,342 (Oct. 6, 1995) — "Method of on-line shopping utilizing persistent client state…" (https://patents.google.com/patent/[US5826242](/patent/US5826242); https://uspto.report/patent/grant/5826242).
• Corroborating publication: Netscape, Persistent Client State HTTP Cookies (the "preliminary specification," 1994–1995; archived at https://web.archive.org/web/19961027104920/http://www3.netscape.com/newsref/std/cookie_spec.html), and the account in http://governingwithcode.org/case_studies/pdf/Cookies.pdf (cookies were built at Netscape, in part, for an online shopping-cart application; shipped with Navigator from late 1994).

What Montulli teaches, verbatim from the specification/claims:

• "in a client-server system … where no state inherently exists in such a system," the server sends a state object with a document; "the client stores the state information" and "in a subsequent request … the client can send the stored state information to the server."
• The express commercial application: "an on-line shopping system… the customer can also select products to be placed in a virtual shopping basket. The server then sends state information related to the selected products to the browser on the client for storage. When the customer wants to purchase the products…, the browser sends the corresponding state information to a specified check-out Web page for processing."
• Merchant flow (Fig. 5): browser requests page → server returns HTML document → customer selects product → server returns synthetic page and a cookie describing the selected product → browser stores the cookie → repeat for more products → check-out sends the accumulated cookies back. The spec even states the cookies "include an identification of the contents of a virtual shopping basket."

This is, in substance, claim 1's core loop (server creates/updates a list; returns it to the browser; browser returns the list with the next selection; repeat until termination), with the list carried as a cookie rather than as a page-embedded hidden field.

(c) HTML hidden fields and URL-embedded state (the "carrier" art)

• HTML 2.0 (RFC 1866, Nov. 1995) defined INPUT TYPE=HIDDEN as a standard form control; CGI programs routinely parsed POST bodies into name/value pairs.
• The published literature records that, pre-cookies, "shopping cart information was appended to URL strings or stored in hidden HTML fields" (see the Kristol/Berghel discussion reproduced in http://governingwithcode.org/case_studies/pdf/Cookies.pdf), and Netscape's own engineers described the problem with URL-based state as the motivation for cookies.

(d) Prior art of record cited by the examiner (per https://portal.unifiedpatents.com/patents/patent/US-5745681-A)

(e) Non-patent literature of record

• Hauser & Steiner, "Generic Extensions of WWW Browsers," First USENIX Workshop on Electronic Commerce, New York, July 11–12, 1995 (https://www.usenix.org/conference/first-usenix-workshop-electronic-commerce/generic-extensions-www-browsers). Abstract: extensions so that "electronic payments ought to be seamlessly integrated into a customer's browser to enable him to shop via the Web," "without altering current browser technology." Filed as prior art against 5,745,681 — i.e., web shopping with browser-side support was known and published six months before the filing date.
• Fried-Cassorla, "Successful Marketing on the Internet: A User's Guide," Direct Marketing, Feb. 1995 — Internet marketing/selling.
• Rowley, "Retailing and Shopping on the Internet," Internet Research, vol. 6, no. 1 (1996), pp. 81–91 — web retailing practices.

Important negative point about the "Cited By (338)" list

Most of the 338 "Cited By" entries on the Google Patents page post-date the 1996‑01‑11 filing and therefore are not prior art (e.g., US 5,900,973 Hitachi/Hitachi basket presentation, priority 1996‑09‑30; US 5,960,411 Amazon 1‑Click, 1997‑09‑12; US 5,956,709 Xue, 1997‑07‑28). The only directly usable §102(a)/(e) item in that list is US 5,826,242 (Montulli, priority 1995‑10‑06) — which is exactly why it dominates the analysis. (Note also that pre-AIA §103(c) would have removed commonly-owned §102(e) art from consideration — but Netscape's art is not commonly owned with Sun, so no such exclusion applies.)

---

3. The differences between the prior art and the claims

Narrowed to essentials, the differences are:

1. Storage locus of the accumulated item list. Montulli stores the state in a cookie (HTTP response header → client cookie file → automatically re-sent on matching requests); claim 1/7/11/15/19 require the updated list to be returned as a field (the "cart" field) embedded in the served page file, with the browser re-submitting that field with the next selection. The patent's own Background already characterizes the page-embedded variant ("hidden data passed back through a form submission").
2. The request that carries the list forward. Claim 7's "first field"/"second field" and claim 11's "add request" are the standard HTML POST name/value pairs of a form submission; Montulli uses the cookie header for the same informational purpose.
3. Deletion (claims 3, 9, 16) — list editing, not described in Montulli's flow but trivially implementable.
4. Download of the selected items/software (claims 2, 4, 10, 14, 18, 21) — conventional file transfer (HTTP/FTP) of files named in the list.
5. Naming the field "cart" (claims 5–6, 8) — a non-functional label; labels without structural/functional consequence carry no patentable weight.
6. Module/apparatus recitations (claims 11–18) and the storage-medium claim (19–21) — functional and article-of-manufacture restatements of the same process. Note that, post-Williamson v. Citrix, 792 F.3d 1339 (Fed. Cir. 2015), "module" is a nonce term that may trigger § 112(f), in which case the corresponding structure is the Fig. 2 flowcharts — which Montulli's Fig. 5 flow discloses in equivalent form (and which narrows the gap rather than widening it).

There is no claimed technical effect in the differences beyond the admitted benefits of client-side storage (session survival across server cleanups, resumability, no account ID).

---

4. Combinations that render the claims obvious

Combination 1 (primary): Montulli + conventional page-embedded state (hidden field / URL query)

References: US 5,774,670 and/or US 5,826,242 (Montulli/Netscape), in view of HTML form hidden fields (RFC 1866) and the admitted use of "hidden data passed back through a form submission" and IDs "in the uniform resource locator" (5,745,681 Background).

Mapping: Montulli supplies every substantive step of claims 1, 7, 11 (§ "receiver/cart list module/add module/shopping page module" = the merchant CGI described in Montulli Fig. 5) and 15 (browser sends a request containing the selection and the previously stored state; server accumulates; client stores and re-sends). The hidden-field/URL art supplies the two differences: returning the list inside the page file and re-submitting it as form data with the next selection. Claims 5–6/8's name/value parsing is the ordinary CGI POST parser applied to cart=<value>.

Why the artisan would combine them:

• Same problem, same field. Both references address the statelessness of HTTP for on-line shopping. 5,745,681's Background states the problem (server-side carts, IDs, logins, cleanup, firewalls that "intentionally do not identify the sender"); Montulli states the identical problem ("the state of a client can be maintained… where no state inherently exists in such a system") and names on-line shopping as the application.
• Predictable use of a known alternative. In early 1996, cookies were supported essentially only by Netscape Navigator; Mosaic, Lynx, and the first Internet Explorer releases did not universally support them. A server deployed to a heterogeneous browser population had a concrete reason to carry the cart list in the one mechanism every HTML 2.0 browser supported — a hidden field submitted with the form — rather than in a Set-Cookie header. KSR: "if a technique has been used to improve one device, and a person of ordinary skill in the art would recognize that it would improve similar devices in the same way, using the technique is obvious."
• Additional design incentives present in the art itself: page-embedded state travels with the specific page, so the user can use the browser's history/stored page to resume (the very behavior the patent touts); server-side per-user storage is avoided, eliminating the cleanup/purge problem the Background identifies; and no log-in/account is required.
• Reasonable expectation of success: the combination is pure software, requires no new hardware, and both constituent techniques worked.

Combination 2 (alternative): URL/hidden-field state + interactive catalog/order-entry systems

References: the admitted URL/ID-in-URL and hidden-form-field art (5,745,681 Background) in view of US 5,721,832 (Westrope; catalog server, menu browsing, order entry and fulfillment over a telephone/terminal network), US 5,630,068 (Vela; cart-displayed item lists), US 5,625,781 (IBM; user-manipulable "itinerary list") and the NPL of record (Hauser & Steiner 1995; Fried‑Cassorla Feb. 1995; Rowley 1996).

Mapping/motivation: Westrope teaches the commerce half (a central processor serving selectable catalog items to remote terminals and processing orders); the admitted URL/hidden-field art teaches the state-carrying half (identification/state appended to the request so the server can associate prior selections without an account); Hauser teaches that web shoppers would be served with browser-side support. A single artisan combining on-line catalog shopping with HTTP form mechanics would arrive at claim 1 and claim 7 as a matter of routine design: put the accumulated selections into the request the browser already has to make, have the server rebuild the list from that request, update it, and hand it back inside the next page.

Motivation: avoiding the account/ID infrastructure (and its server-side cleanup) that Westrope and the admitted art required; the patent itself argues these are the benefits.

Combination 3: The download-type claims (2, 4, 10, 14, 18, 21) and deletion claims (3, 9, 16)

• Download: Montulli (or Combination 1) in view of conventional network file retrieval (HTTP/FTP) and Westrope's order-fulfillment teaching. The patent's specification admits the items are "software files such as device drivers" (Sun's Solaris Driver Express); retrieving files by identifier and sending them to the requester was the oldest function on the network. Obvious.
• Deletion: Montulli/Combination 1 in view of US 5,625,781 (list management in an interface) or merely ordinary list-editing. Removing a named element from a list before regenerating the page is a routine, predictable data-manipulation step. Obvious.
• Claim 20 (releasing server storage after the field is sent): inherent in the stateless approach — Montulli expressly states no state is retained on the server; releasing a temporary variable is an automatic consequence.

Combination 4: Apparatus claims 11–14 and 15–18

The recited hardware (processor, memory, keyboard, display, I/O control, storage devices, network adaptor) is the general-purpose computer of Montulli's Fig. 1B/6 disclosure. The "modules" are the logical operations the references describe; under Williamson the specification supplies the algorithms as the Fig. 2 flowcharts, which correspond to Montulli's Fig. 5 merchant flow and the browser's request/POST behavior. Obvious in view of Combinations 1–3.

Combination 5: Claim 19–21 (Beauregard-style storage medium)

A computer-readable-medium claim stands or falls with the process it encodes; see In re Beauregard, 53 F.3d 1583 (Fed. Cir. 1995), and MPEP 2114. Given Combinations 1–3, programming a conventional CD-ROM/diskette (the very media recited in 5,745,681's own Fig. 1 description — cd-rom 28, diskette 31) with that process is obvious.

---

5. Objective indicia (secondary considerations)

I found no verifiable evidence in the record supporting non-obviousness, and several circumstances cut the other way:

• No unexpected results or long-felt-but-unsolved need. The problem (HTTP statelessness for shopping) was being solved contemporaneously and publicly: Netscape's cookie work dates to 1994–95 and was disclosed in the Cookie specification and in the Oct. 6, 1995 Montulli application; Hauser & Steiner published on web shopping in July 1995; the 1996 filings cluster (Netscape Oct. 1995; Sun Jan. 1996; Hitachi Sept. 1996) indicates the solution was within ordinary skill, not beyond it.
• No verified nexus evidence. The high citation count (338+ citing documents) and the PTAB/CBM citation (e.g., CBM2016‑00047 re U.S. 8,041,626) show the patent is foundational prior art, which says nothing about whether it was non-obvious when filed.
• The 2022 Texas Eastern litigation (case 4:22-cv-00214, per Unified Patents data on the Google Patents page) is of unverified party-level detail and in any event cannot establish nexus; the prior section's caveat stands.
• If a patentee were to argue commercial success of the "Driver Express" cart, it would have to show the success flowed from the claimed features rather than from the general rise of web commerce — a difficult showing given the art above.

---

6. Bottom line

On the art visible on the page and in the record, claims 1–21 of US 5,745,681 are, in my assessment, invalid under § 103 as obvious over Montulli (US 5,774,670 / US 5,826,242) — which discloses client-stored, server-reconstructed shopping-basket state expressly for on-line shopping with "no state inherently exists" on the server — in view of the admitted and conventional page-embedded state techniques (HTML hidden fields / URL-encoded state, RFC 1866; 5,745,681 Background). Claim 1 in particular sits very close to Montulli and may also be anticipated under pre-AIA § 102(e), which strengthens the § 103 conclusion. The remaining claims add only (i) field naming, (ii) list deletion, (iii) file download of the listed items, (iv) generic computer apparatus, and (v) a generic storage medium — all predictable, routine, and supported by the secondary art discussed above.

Strongest counterarguments to expect from the patentee: (1) Montulli's cookie is a header-side object, not a page field, and the claims require the cart to travel inside the served page file (the "history/stored page" and browser-managed benefit); (2) in 1995–96, cookies were Netscape-only, so the artisan using hidden fields was not "modifying Montulli" but choosing a different (older) mechanism — which is a § 102 argument about which reference teaches page-embedded state, not a non-obviousness argument; (3) secondary considerations without clear nexus. Neither point 1 nor 2 defeats the combination, because the "carrier" mechanism was admitted prior art in the same patent.

---

7. Confidence and limitations

• High confidence in the claim text, the parties, and the content quoted from the full patent text and from Montulli/US 5,826,242 and US 5,774,670 (retrieved via Google Patents, uspto.report, and the USPTO-hosted PDF).
• Medium confidence in the precise dates of the Netscape cookie specification as a printed publication (sources variously cite 1994, Oct. 1995, and Oct. 1996); this does not affect the analysis because the Oct. 6, 1995 US application supplies a §102(e) date before the 1996‑01‑10/11 priority of 5,745,681 on either date.
• Lower confidence / unverified: the exact content of US 5,623,656 ("state memory") beyond its title and assignee as listed on the Unified Patents page; whether the examiner actually had Montulli before him (the Google "Cited By" asterisk identifies US 5,826,242 as examiner-cited, but I could not retrieve the file wrapper to confirm which application that citation belongs to); and any EPO opposition to EP 0 784 279 / EP 0 807 891 (my query on that point returned no usable results). EPO grant of the family members is noted (EP 0 784 279 granted 1998‑06‑17; EP 0 807 891 granted 2000‑05‑17, per http://data.epo.org/pise-server/rest/collections/lgpi/EP0784279B1.pdf and http://eupat.ffii.org/patents/samples/ep807891/), but EPO inventiveness findings are not controlling and rest on a different prior-art universe (notably, the Netscape cookie filings may not have been in the EPO search file).
• Contradiction to flag: the prior section reported the TXED case 4:22-cv-00214 as unverified; I found nothing further, and I do not treat it as probative of validity either way. Also note the persistent date discrepancy for this patent (1996‑01‑10/11; grant 1998‑04‑27/28) between Google Patents and the Unified Patents page — both dates precede and postdate Montulli the same way, so the analysis is unaffected.