US 6519648

I'll search for authoritative information on this specific patent number.

Let me search for the full claim set and any recent (2026) litigation activity.

US Patent 6,519,648 — Analyst Summary

Scope note / sourcing: I was able to retrieve bibliographic data (Google Patents' mirror of USPTO data), the patent's own text, the N.D. Cal. claim-construction and summary-judgment opinions in Friskit v. RealNetworks, and the Federal Circuit opinion in Appeal No. 2007-1583. I did not obtain a live USPTO PatentCenter/Patent Trial and Appeal Board entry or a CAFC 2026 docket entry for this patent, and I found no 2026 CAFC docket activity referencing 6,519,648. I flag uncertainties explicitly below rather than filling gaps.

1. Bibliographic data (as listed)

Priority chain (as listed): The '648 patent is a continuation-in-part of U.S. App. Ser. No. 09/563,250, filed May 2, 2000 (that application issued as US 6,389,467), and claims benefit of provisional application 60/177,786, filed January 24, 2000, naming A. Eyal and G. Asporopos as inventors. The same-January-24-2000 priority date is shared across the Friskit family (see §5).

2. Abstract (as published)

"A playback system is provided to locate and playback streaming media from network resources. The playback system includes a search module that signals a query to a network site, and receives in return a search result. The search result identifies one or more links that are selectable to open media files. A media player couples to the search module to automatically play back streaming media contained in media files located by the search result."

3. Plain-language overview of the disclosure

The specification describes a client-side media playback system that (a) accepts a user query (or a preference/action-generated query), (b) translates it into search-engine-specific network queries (URLs) using a rule module, (c) submits them to multiple third-party Internet search engines (e.g., Yahoo®, Lycos®, streambox.com, gigabeat.com) and/or media directories, (d) parses the returned search-result pages with engine-specific logic to isolate links that open media files (filtering out ads, internal links, and broken/duplicate links), (e) optionally verifies/date-checks links via a hidden media-player instance, arranges them into a playlist, and (f) causes a media player to play them back continuously and automatically — effectively a "network radio/jukebox" operating in the background.

Notable implementation details: use of a controlled/hidden web browser instance as the parser component via an API; conversion of "HTTP" to "PNM" in URLs to work around a RealNetworks media-player bug; multi-page search-result traversal; background operation while the user works in other applications; and server-side alternative embodiments.

4. Independent claims — plain-language overview

Important limitation on the following: the claim text supplied to me truncates mid-way through claim 17, and claims 18–63 are not available in the retrieved text. I therefore describe only the independent claims I can ground, and I flag the one independent claim I can identify only by inference.

Claim 1 — "A media playback system" (client-side, multi-site)

Elements:

1. A search module that signals a search request to a first network site and receives a search result;
2. In response, the search module identifies a plurality of media resource locators, including:
   • a first media resource locator locating a first media resource at a second network site, and
   • a second media resource locator locating a second media resource at a third network site; and
3. A media player coupleable to the search module that plays back the first media resource using the first locator, and then "substantially automatically" plays back the second media resource using the second locator.

Plain language: search one place (a search engine), get back locators that point to media hosted on different sites from each other, and have a player automatically play them back-to-back without the user clicking each link. The claim draws the line at the client, not a server.

Claim 17 — "A media playback system operated on a terminal coupled to a network"

Elements (as far as the retrieved text goes):

1. A search module that signals a user-defined search request to one or more network sites to receive one or more search results, the result(s) identifying a plurality of media resource locators, each locating at least one media resource at one or more network sites; and
2. A media player coupled to the search module to automatically access and load at least some of the media resources loc… [text truncated in the retrieved document]

Plain language: the same automatic-search-then-auto-load-and-play concept, framed around a terminal and a user-defined request. The precise wording of the final limitation is uncertain because the retrieved claim text is cut off.

Claim 37 — independent claim (identified by inference, text not available)

The N.D. Cal. summary-judgment opinion refers to "the '648 Patent claim 49 (dependent on claim 37)" and to asserted claims 49 and 52. This establishes that claim 37 is an independent claim in the '648 patent, but I do not have its text and will not paraphrase it. Whether claim 52 is independent or dependent on claim 49 is also not established by the materials I retrieved.

If claim-numbering accuracy for claims 37+ is important for your purpose, I'd verify against the USPTO full-text/PatentCenter image rather than rely on this summary.

5. Family context (helpful for claim scope)

Related Friskit filings sharing the Jan. 24, 2000 priority date:

• US 6,389,467 B1 — Streaming media search and continuous playback system of media resources located by multiple network addresses (App. Ser. No. 09/563,250) — the parent CIP.
• US 6,484,199 B2 — Streaming media search and playback system for continuous playback of media resources through a network.
• US 6,725,275 B2 — Streaming media search and continuous playback of multiple media resources located on a network.
• US 6,735,628 B2 — Media search and continuous playback of multiple media resources distributed on a network.
• US 7,228,305 B1 — Rating system for streaming media playback system.

A meaningful distinction inside the family: the siblings emphasize network-server-module control (server-side "programmatic control"), while the '648 claims as issued are drawn to the search module / client-side architecture.

6. Enforcement and validity history

• Litigation listed on Google Patents: a US case filed in the California Eastern District Court, case 3:03-cv-05085; a "first worldwide family litigation" entry is also listed via Darts-IP (family 32110744).
  • Caution: the court documents I retrieved for this dispute are captioned FRISKIT, INC. v. REALNETWORKS, INC. and LISTEN.COM, No. C 03-5085 (FMS/WWS), U.S. District Court, Northern District of California (Friskit, Inc. v. RealNetworks, Inc., 2007 WL 1871910 / 541 F. Supp. 2d 1130 (N.D. Cal. July 26, 2007)). The docket number is the same (03-05085) but the district differs from the Google Patents listing. I report both as they literally appear rather than reconciling them.
• Five patents asserted in that case: '467, '199, '648, '275, and '628. The court limited the case to eight claims, including '648 claims 49 and 52.
• Claim construction (May 20, 2005): the court construed, inter alia, "media resource" = "At least one media file including audio, video or combinations thereof, capable of being loaded into a user accessible playback component"; "client module" = "A module that is installed and executed on a user-terminal"; "network server module" = a server-side module in a client-server network. The court expressly declined to read "streaming media" into "media resource."
• Summary judgment of invalidity (July 26, 2007): Judge Schwarzer granted RealNetworks' § 103 motion, holding the asserted claims obvious under KSR — the "glue together existing technologies" theory (media players, search engines, streaming media, playlists, and Internet radio, incl. mp3.com, IUMA.com's "Radio IUMA," and Winamp) did not supply a nonobvious difference.
• Federal Circuit: Friskit, Inc. v. RealNetworks, Inc., Appeal No. 2007-1583 (nonprecedential disposition) — the court affirmed, holding that "programmatic control" and "direct control" of the media player by network-based modules would have been obvious in light of IUMA Radio and the state of the art, and that Friskit's secondary-considerations evidence did not overcome the prima facie case. Because it affirmed on obviousness, the court did not reach indefiniteness of "substantially automatic" or non-infringement.
• Commercial/press record: RealNetworks issued a press release (dated July 30/31, 2007) announcing a "complete victory" and that Judge Schwarzer invalidated the asserted Friskit claims, citing KSR.

7. Explicit uncertainty flags

1. No 2026 CAFC docket activity found for 6,519,648, and no 2026 litigation found; the patent is listed Expired – Fee Related (anticipated expiration May 2, 2020). If the '648 patent were the subject of a new 2026 appeal, it was not surfaced in my searches, and I cannot confirm one exists.
2. I could not verify a live USPTO PatentCenter record (e.g., current fee/reinstatement status or any post-2020 maintenance events) in this session; the status field above is Google Patents' rendering of USPTO data.
3. Claims 18–63 were not available in full text to me (the supplied document truncates inside claim 17). My claim analysis is confined to claims 1 and 17, plus the inference that claim 37 is independent.
4. Claim 49 and 52 were at issue in litigation, but I did not retrieve their full text; I cannot state their exact scope.
5. The district-court identity for case 3:03-cv-05085 is internally inconsistent across sources (Google Patents says California Eastern District; the opinions say N.D. Cal.). I have not resolved which is correct.

Primary sources used:

• https://patents.google.com/patent/US6519648/en (bibliographic data, abstract, description, claims 1–17, litigation/family entries)
• https://storage.courtlistener.com/harvard_pdf/[1871910](/patent/1871910).pdf and https://www.courtlistener.com/opinion/1871910/friskit-inc-v-realnetworks-inc/ (N.D. Cal. SJ order, July 26, 2007)
• https://ipmall.info/sites/default/files/hosted_resources/Markman/pdfFiles/2005.05.20_FRISKIT_INC_v._REALNETWORKS.pdf (Claim Construction Order, May 20, 2005)
• http://cafc.uscourts.gov/sites/default/files/opinions-orders/07-1583.pdf (Fed. Cir. Appeal No. 2007-1583, nonprecedential)
• https://cn.realnetworks.com/press/releases/2007/realnetworks-wins-complete-victory-patent-suit (RealNetworks press release, July 2007)

I'll research litigation involving US patent 6519648 using the sources you specified.

The first pass pulled in unrelated JP6519648 (a Japanese Nike patent). Let me search more precisely for the US patent and its owner Friskit.

I found the key case. Let me verify details and check for any additional litigation involving this specific patent.

Research Summary — Litigation Involving US Patent 6,519,648

Patent: US 6,519,648 B1 — "Streaming media search and continuous playback of multiple media resources located on a network"
Inventor: Aviv Eyal · Assignee: Friskit, Inc. · Filed: July 11, 2000 (App. 09/613,433) · Granted: Feb. 11, 2003 · Priority: Jan. 24, 2000 (provisional 60/177,786) via CIP of Ser. No. 09/563,250 (filed May 2, 2000)

Based on the sources searched (Google Patents litigation metadata, the Unified Patents litigation portal, CourtListener/FJC IDB docket records, CAFC opinions, and Darts-ip family data), one litigation is known to involve US 6,519,648 specifically. I found no other suits naming this patent.

Case 1 — Friskit, Inc. v. RealNetworks, Inc. and Listen.com

Procedural history

• May 20, 2005 — Claim construction (Markman) order issued after the April 12, 2005 hearing (tentative order April 29, 2005). ipmall Markman PDF; CourtListener Doc. 181
• Through 2005–2006, extensive discovery and motion practice (interrogatories on written description/enablement, appointment of an independent expert, protective orders, motions to dismiss) — see docket docs at CourtListener docket 4165499 and Docket Alarm

Outcome

• July 26, 2007 — The district court granted RealNetworks' second motion for summary judgment of invalidity for obviousness (35 U.S.C. §103), holding the asserted claims — including claims 49 and 52 of the '648 patent — invalid. The decision is reported at 499 F. Supp. 2d 1146 (N.D. Cal. 2007) (matches the *1148/*1153 star pagination seen in the opinions). See district court summary-judgment opinion text and CourtListener opinion. A "Judgment in a Civil Case" was entered the same date (Jul. 26, 2007). The FJC IDB lists the case as terminated July 26, 2007 (some aggregators show closure 01/29/2008). FJC IDB record

Appeal

• Friskit appealed. The U.S. Court of Appeals for the Federal Circuit, in a nonprecedential decision, No. 2007-1583, affirmed on January 12, 2009, applying KSR Int'l Co. v. Teleflex Inc. and holding the asserted claims obvious. CAFC opinion PDF

Current status: Final judgment of patent invalidity against Friskit, affirmed on appeal; no injunction against RealNetworks. The '648 patent's legal status is now "Expired – Fee Related" (Google Patents lists anticipated expiration 2020-05-02).

Important caveats and discrepancies

1. Venue discrepancy. The Google Patents litigation metadata labels this suit as filed in the "California Eastern District Court" and links to portal.unifiedpatents.com/litigation/California%20Eastern%20District%20Court/case/3:03-cv-05085. However, the court filings, the Markman opinion, and the FJC IDB all show the case was in the Northern District of California, Case No. C 03-05085. Notably, the FJC IDB records the case Origin as "(5) Transferred from another district (pursuant to 28 U.S.C. §1404)", which may explain the Eastern District reference (i.e., a possible original E.D. Cal. filing later transferred to N.D. Cal.). I could not independently confirm an E.D. Cal. docket number, so treat the "Eastern District" label with caution.

2. This is the only case I could confirm involving US 6,519,648. The Google Patents page shows a single "Family has litigation" entry and a Darts-ip "First worldwide family litigation filed" link (family 32110744) — consistent with this one case being the only identified litigation for the patent/family. I did not find additional defendants (e.g., Napster, Musicmatch, Apple) sued on the '648 patent, but I was unable to complete exhaustive checks for every potential case before reaching the tool limit — so I cannot rule out other assertions with complete certainty.

3. Unrelated hits on similar numbers. Searches for "6519648" also return an unrelated Japanese patent JP 6519648 B2 (a Nike yarn/knitting patent, grant 2019) — this is a different patent with a coincidentally matching number and is not related to US 6,519,648. Per your instruction to interpret identifiers literally, these are distinct documents and I have excluded them.

Sources: Google Patents US6519648B1 · Unified Patents litigation portal · CAFC No. 2007-1583 · CourtListener docket 4165499 · FJC IDB

I'll start by checking the canonical structured source and then search for any PTAB activity on this patent.

PTAB Proceedings — US 6,519,648

Proceedings overview

The USPTO Open Data Portal reports zero AIA trial proceedings (IPR, PGR, or CBM) on US 6,519,648. The patent therefore has no PTAB claims-invalidated, claims-sustained, settled, or institution-denied record to work from — but that absence is not the good news you might assume, because the patent's invalidity history lives in the district courts and at the Federal Circuit. The bottom line for a defendant is: there is no PTAB cancellation to point to, but a district court invalidated the asserted claims of this patent for obviousness and the Federal Circuit affirmed — and the patent expired in 2020.

Sourcing note: Per the operating rules, I use the structured "PTAB proceedings on file" block as the canonical list. That block states the ODP API returns no AIA trial proceedings. My independent web search found no IPR/PGR/CBM naming US 6,519,648 as the challenged patent. I found no proceeding numbers to report — and I have not invented any.

No AIA trial proceeding is on file for US 6,519,648

• Type: N/A — no IPR, PGR, or CBM petition identified in the ODP dataset or in public web sources.
• Filed: N/A
• Status: N/A
• Defensive value: You cannot build an IPR-estoppel or FWD-based defense. You can point to a final, affirmed district court invalidity judgment on the asserted claims (below) and to the patent's 2020 expiration.

Why this is plausible rather than a data gap. The patent (priority 2000-05-02, filed 2000-07-11, granted 2003-02-11) was thoroughly stress-tested in Friskit, Inc. v. RealNetworks, Inc. and Listen.com, No. C 03-5085 FMS (N.D. Cal.) — the litigation the structured data flags as 3:03-cv-05085 (Google Patents labels the jurisdiction "California Eastern District Court," but the docket records I found are N.D. Cal., gov.uscourts.cand.15671; treat the "Eastern" label as suspect). Friskit never commercialized at scale, the patent went Expired – Fee Related with an anticipated expiration of 2020-05-02, and no defensive aggregator (Unified Patents, RPX, etc.) appears in the chain. A patent that is expired, was already invalidated in district court, and is held by a dormant entity is exactly the profile that attracts no IPR filings.

C 03-5085 FMS — Friskit, Inc. v. RealNetworks, Inc. and Listen.com (NOT a PTAB proceeding — included because it is the substantive invalidity record)

I flag this explicitly as district court litigation, not an AIA trial, so it is not double-counted in the overview above.

• Type: Federal district court patent infringement action (invalidity adjudicated on summary judgment).
• Filed: 2003 (N.D. Cal., No. C 03-5085 FMS).
• Status: District court granted summary judgment of invalidity; affirmed on appeal.
• Judge panel: Hon. Fern M. Smith (district court). Federal Circuit panel not confirmed from the source I retrieved.
• Claims at issue for the '648 patent: The court's obviousness analysis addressed '648 Patent claims 49 and 52 — claim 49 depends from claim 37, and claim 52 adds a priority-order limitation. (These are the litigation-era claims; they are distinct from the claims 1–17 that appear at the top of the patent's claim set.)
• Grounds: § 103 obviousness over prior-art media players and streaming sites (Winamp, RealJukebox, mp3.com, IUMA.com "Radio IUMA," Java/JavaScript server-to-client control, M3U playlist handling), decided post-KSR.
• Disposition: The court found all individual features existed in the prior art and that combining them was the "predictable use of prior art elements according to their established functions." For the '648 claims specifically, the court rejected Friskit's argument that the streaming-media and priority-order limitations saved claim 52: "The streaming media limitation does not render this claim nonobvious," and "streaming media playlists were put in predesignated order and then were played back by the user's media player in that order."
• Appeal: Friskit, Inc. v. RealNetworks, Inc., No. 2007-1583 (Fed. Cir.) — nonprecedential opinion affirming the district court's invalidity judgment. The court held that "a system employing 'programmatic control' and 'direct control' of the media player by the network-based server would have been obvious to one of ordinary skill in the art."
  • Opinion: http://cafc.uscourts.gov/sites/default/files/opinions-orders/07-1583.pdf
• Defensive value: High, but with a scope caveat. The asserted '648 claims were held invalid and that holding was affirmed, and the reasoning (Winamp + mp3.com/IUMA + M3U playlists + server-side script control) is a ready-made § 103 roadmap against the same subject matter. However, it is a district court judgment, not a PTAB cancellation — it binds parties and privies, and its preclusive effect on a new defendant depends on claim-scope mapping.

Supporting primary sources:

• Claim construction order (2005-05-20): https://ipmall.info/sites/default/files/hosted_resources/Markman/pdfFiles/2005.05.20_FRISKIT_INC_v._REALNETWORKS.pdf and https://storage.courtlistener.com/recap/gov.uscourts.cand.15671/gov.uscourts.cand.15671.181.0.pdf
• Summary-judgment invalidity opinion (incl. '648 claims 49 & 52): http://online.wsj.com/public/resources/documents/friskit.pdf
• Federal Circuit affirmance: http://cafc.uscourts.gov/sites/default/files/opinions-orders/07-1583.pdf
• Google Patents family/docket page (structured source): https://patents.google.com/patent/US6519648/en

Strategic summary

Claim status: no claims canceled, but the key asserted claims were invalidated outside the PTAB. US 6,519,648 has 63 claims. Claims 1 and 17 are the independent claims visible at the top of the set; claims 37, 49, and 52 appear in the later portion relied on in litigation. Because there was no IPR, no claim stands "confirmed" by the Board. Conversely, because the Board never canceled anything, no claim is formally dead. What is firmly established is that the '648 claims litigated against RealNetworks (including claims 49 and 52) were held obvious and that holding was affirmed. Treat claims 1, 17, and all other untested claims as UNTESTED at the PTAB — nobody has been estopped, and nobody has been confirmed. Verify the exact claim numbers the judgment covered by pulling the full summary-judgment order; my source showed claims 49 and 52 expressly addressed, and I will not assert it covered more.

Estoppel landscape is unusually favorable — because there is nothing to estop. Because no IPR/PGR was ever instituted, § 315(e)(2) estoppel is a non-issue: there is no petitioner or privy laboring under an estoppel bar, and there is no IPR record constraining claim construction. Your prior-art runway is essentially unlimited — you may raise any § 102/§ 103 ground, including art that would have been barred had an IPR run, subject only to ordinary district court rules and § 325(d)-style discretion if you choose to file an IPR now. The countervailing consideration is that Friskit v. RealNetworks gives you a strong, judicially validated obviousness narrative for free; you should adopt and extend that record rather than reinvent it.

Pattern signals — none of the usual ones. There is no repeat petitioner (no IPRs at all). There is no aggressive PTAB-appeals posture by the patent owner — Friskit's appellate activity was a defensive appeal of an adverse district court invalidity ruling, not a PTAB appeal, and it lost. There is no defensive aggregator in the chain: the structured data shows the original assignee Friskit Inc, a 2001-10-29 and 2003-12-24 security agreement to George Aposporos, and nothing else (no Unified, no RPX). And, critically, the patent is Expired – Fee Related with anticipated expiration 2020-05-02. Relevant context: Friskit's other patent, US 6,389,467 (Eyal), is not the subject of the IPRs you may find in searches — it is cited as prior art in the Hulu v. SITO Mobile line (IPR2021-00158, IPR2021-00206, IPR2021-00219, IPR2021-00265, IPR2021-00308). Do not mistake that for a challenge to 6,519,648; those proceedings concern SITO Mobile's patents and use Eyal as a reference.

Recommended next steps

1. State the PTAB position plainly in any defense memo: there is no AIA trial proceeding on US 6,519,648. No FWD exists to link or quote. Do not represent that any claim was canceled by the Board — it was not.
2. Pivot the defense to the affirmed district court judgment. Pull the full summary-judgment order and the Federal Circuit opinion (No. 2007-1583) and map their obviousness rationale (Winamp + mp3.com/IUMA Radio + M3U/playlist handling + JavaScript server control) onto whatever claims your adversary is asserting. The Court of Appeals' operative holding — "a system employing 'programmatic control' and 'direct control' of the media player by the network-based server would have been obvious" — is quotable and directly on point for the patent's core disclosure.
3. Run the expiration math before anything else. With an anticipated expiration of 2020-05-02 and the § 286 six-year damages lookback from a complaint filed today (2026-09-28) reaching back only to 2020-09-28, there is no recoverable past-damages window and no prospect of injunctive relief. If a demand letter is citing this patent, that is a threshold standing/value problem for the claimant.
4. Confirm the claim-scope gap. Identify exactly which of the 63 claims the district court invalidated. Claims the judgment did not reach remain formally untested — that is where residual assertion risk lives, and it is where you should direct an invalidity-contentions workup.
5. If you nonetheless contemplate an IPR (e.g., as a litigation-stay lever), note the practical constraints: the patent is expired, so claims are construed under the Phillips standard rather than the BRI standard used for unexpired claims, and the Board's discretionary-denial calculus (§ 314(a)/§ 325(d)) will be unusual. Filing is legally possible but strategically marginal given expiration and the existing affirmed invalidity judgment.

Confidence note: I am highly confident there is no PTAB record on this patent — the ODP block says so and my searches corroborate it. I am confident about the substance of the N.D. Cal. invalidity ruling and the 2007-1583 affirmance, having retrieved both texts, but I did not independently confirm the Federal Circuit panel composition or the precise opinion date, so I have not stated either.

I'll research the assignment record. Let me run several searches in parallel.

Good — new leads on Bluebonnet Media. Let me dig into the post-Friskit chain.

Assignment & Ownership Analysis — US 6,519,648 B1

Critical sourcing caveat, stated up front: I could not retrieve the USPTO Assignment Center (assignmentcenter.uspto.gov / assignment.uspto.gov) record set for this patent in this session. The Assignment Center is a JavaScript application that does not expose its reel/frame data to plain text retrieval, and my search budget for this task was exhausted before I could reach an indexed mirror. Accordingly, I have no reel/frame numbers, no correspondent-of-record (recording attorney/agent) data, and no execution-vs-recording date pairs beyond those surfaced in Google Patents' "Legal Events" mirror of USPTO assignment data. I will not invent reel/frame identifiers. Everything below is either (a) directly attested in the sources cited, or (b) explicitly flagged as unverified. Where the requested output format asks for data I do not have, I say so rather than approximate it.

Inventors

Unusual patterns worth flagging:

1. Inventor departed a prior employer ~3 weeks before the priority filing. Eyal left Microsoft on Dec. 31, 1999; the provisional was filed Jan. 24, 2000. RealNetworks moved to dismiss the 2003 suit on the theory that Eyal's Microsoft employment agreement obligated him to assign the invention to Microsoft, and that Microsoft — not Friskit — owned the technology. The court denied the motion (July 6, 2006), holding the clause was a promise to assign, not a present assignment, and that no assignment had ever been made to Microsoft. This is a cloud-on-title event, not a fire-sale precursor — but it is the single most consequential ownership fact in the file.
2. Sole-inventor / different-inventor asymmetry across the family. The '648 lists only Eyal; the '741 lists Eyal and Aposporos; the '467, '199, '275, '628 differ. This creates per-patent inventorship and (potentially) per-patent ownership questions if any Friskit-era assignments were executed by fewer than all obligors.
3. No "all inventors depart within 12 months" pattern. Eyal and Aposporos remained associated with Friskit through the 2003 assertion and (per the Bluebonnet complaint) into the Microsoft dispute. Aposporos was still Friskit's CEO in 2003 and surfaced as Bluebonnet's corporate representative in 2020–21.

Original assignee

FRISKIT, INC., San Francisco, California (assigned by an instrument recorded Oct. 10, 2000, "ASSIGNMENT OF ASSIGNORS INTEREST," assignor Aviv Eyal).

• Did it ship a product embodying the claims? Yes, but only briefly and never commercially at scale. Friskit launched a streaming-media search/continuous-playback web service (reported variously as by June 2000 and as "launched in November 2000"); it reported >100,000 beta users; and it "removed the service in early 2001" (Aposporos, quoted in InfoWorld, 2003). By the time of the 2003 lawsuit the company described itself in its own press materials as "a technology licensing company," and Aposporos told the LA Times it "has been unable to find any customers for its technology."
• Primary line of business: originally consumer streaming-media search/playback; pivoted to IP licensing/enforcement after early 2001.
• Current status: defunct/non-operating. Described as "now-defunct" and as having "ceased operating" (Digital Music News). It settled a separate dispute with Microsoft confidentially in 2011. No bankruptcy filing was found in my searches, but I did not exhaustively check PACER/§ 363 sale records.

Assignment timeline

Recorded events surfaced for US 6,519,648 (source: Google Patents Legal Events, which mirrors the USPTO assignment index). Reel/frame numbers, execution dates, and correspondent of record are NOT available to me — see caveat above. The dates below are the recording dates as rendered by that mirror; where a distinction between execution and recording matters I say so.

• 2000-10-10 (recorded) — Reel NNNNNN/NNNN — not retrieved

  • Conveyance: Assignment of assignors' interest
  • Assignor: EYAL, AVIV
  • Assignee: FRISKIT, INC.
  • Correspondent: not retrieved — I cannot confirm or refute recurrence.
  • Context: Formation/start-up assignment — the inventor assigns his rights to the company he co-founded, recorded ~9 months after filing and well before issuance.

• 2001-10-29 (recorded) — Reel NNNNNN/NNNN — not retrieved

  • Conveyance: Security Agreement
  • Assignor: FRISKIT, INC.
  • Assignee: APOSPOROS, GEORGE (an individual — Friskit's co-founder/chairman/CEO)
  • Correspondent: not retrieved
  • Context: Securitization / distressed financing. Friskit had pulled its consumer service in early 2001 and was out of capital; this is a collateral pledge to an insider, i.e. an encumbrance on the patent estate, not a sale. It is the earliest hard evidence of the company's financial distress.

• 2003-12-24 (recorded) — Reel NNNNNN/NNNN — not retrieved

  • Conveyance: Security Agreement
  • Assignor: FRISKIT, INC.
  • Assignee: APOSPOROS, GEORGE
  • Correspondent: not retrieved
  • Context: Securitization / internal financing, second lien. Recorded ~6 months after Friskit filed its infringement suit, consistent with funding litigation through an insider-secured facility.

Events I looked for but did NOT find a recorded assignment for:

• No recorded assignment from Friskit to Bluebonnet Media / Bluebonnet Internet Media Services surfaced for the '648. Google Patents still lists Friskit Inc as Current Assignee, which is at least consistent with the '648's recorded chain terminating at Friskit (subject to the two security interests).
• Bluebonnet nonetheless publicly describes itself as the successor to Friskit's IP generally ("Bluebonnet, through Friskit, was the pioneer…"). The patents it actually asserted against Pandora in 2020 — US 9,405,753; 9,547,650; 9,779,095 — are continuations of the Friskit family (the '648 page itself lists priority children US 10/784,446 and US 11/935,398), consistent with a portfolio split: continuations/prosecution live with Bluebonnet while the '648 may or may not have travelled with them. I cannot resolve this without the Assignment Center record for both the '648 and the Bluebonnet-acquired members. Do not treat the Bluebonnet link to the '648 as established.

Timeline diagram

timeline
    title Ownership of US 6519648
    1999 : Eyal co-founds Friskit
    2000 : Provisional filed Jan 24
         : Eyal assigns rights to Friskit
    2001 : Consumer service pulled
         : Security agreement to Aposporos
    2003 : Patent issues Feb 11
         : Security agreement to Aposporos
         : Friskit sues RealNetworks and Listen
    2005 : Claim construction in ND Cal
    2007 : RealNetworks wins on obviousness
         : Federal Circuit affirms
    2011 : Confidential Microsoft settlement
    n.d. : Friskit ceases operations

(Note: the diagram deliberately omits any Friskit-to-Bluebonnet arrow, because no such recorded assignment was verified for this patent. Adding one would be fabrication.)

NPE / troll-pattern signals

1. Shell-entity transfer — UNCLEAR.
No operating-assignee-to-licensing-only-LLC transfer is recorded for the '648. The chain shows Friskit retaining title subject to two insider security interests. The family's later custody by Bluebonnet Internet Media Services (a Texas entity asserting in E.D./W.D. Tex.) is reported in litigation complaint materials and press, but I have no reel/frame, no registered-agent address, no single-member-LLC confirmation, and no evidence the '648 itself was conveyed. Naming alone ("Bluebonnet Media") is not a finding under the task's own rule.

2. Known asserter in the chain — UNCLEAR, leaning PRESENT (unverified against the lists you named).
Bluebonnet Internet Media Services is a single-plaintiff, litigation-only entity that sued Pandora Media, LLC in 2020 (W.D. Tex. No. 6:20-cv-00731) on Friskit-derived continuation patents; the district court held the claims invalid in Sept. 2022, and Bluebonnet appealed (Fed. Cir. mandamus denied Oct. 13, 2021, In re Pandora Media, No. 21-172). That is classic asserter behavior. However, I was unable to check the Unified Patents portal or RPX asserter directory in this session — I could not confirm Bluebonnet appears on either list, and I also could not confirm it owns the '648. Neither Bluebonnet, Friskit, nor Aposporos matches Acacia, Marathon, IV, IPNav, Wi-LAN/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, or Spangenberg-entity names.

3. Repeat correspondent across the chain — UNCLEAR / NOT DETERMINABLE.
This is the signal you flagged as most diagnostic, and it is precisely the data I could not obtain. No correspondent-of-record name was retrieved for any of the three recorded events. I will not guess. This should be the first thing verified against the Assignment Center.

4. Cascading transfers — NOT PRESENT.
Three recorded events over 39 months, and the only "chain" is assignor→company (2000), company→insider (2001), company→insider (2003). No chained LLCs, no sub-24-month relay of the patent between successively-formed entities. (The 2001 and 2003 events are 26 months apart and are to the same individual.)

5. Pre-litigation transfer — NOT PRESENT.
The only pre-suit transfer-like event is the Oct. 29, 2001 security agreement — 20 months before the Friskit complaint and to Friskit's own CEO, not to an asserter. The Dec. 24, 2003 security agreement post-dates the litigation (filed late June 2003), which is the opposite of a pre-suit venue/standing arrangement. Whatever Friskit did in 2003, it did as the owner of record.

6. Bankruptcy fire-sale — NOT PRESENT on the record; UNCLEAR on the facts.
Friskit is described as having ceased operating, but no Chapter 7/11 filing or § 363 sale was found, and the only recorded dispositions are the two insider security agreements. Friskit is not a Kodak/Nortel/Polaroid-style estate sale on the evidence available.

7. Privateering — NOT PRESENT.
No operating company is alleged to have funded Friskit's or Bluebonnet's assertions. Friskit's own counter-narrative was defensive/commercial (it was seeking a licensee, and told the press it had demonstrated its technology to RealNetworks in licensing talks that failed). Friskit had ceased shipping by the time it sued.

8. Defensive aggregator — NOT PRESENT.
No RPX, AST, LOT, Unified, or OIN participation appears anywhere in the chain.

Supplementary distress signal (not on your list, but evidenced): Two security agreements to the same individual insider (Oct 2001 and Dec 2003), bracketing the "pivot to licensing" and the first assertion. Together with the failed capital raise and the inability to find customers, this is strong, concrete evidence that the patent was an encumbered asset of a failed start-up rather than the product of a normal commercial portfolio.

Verdict

NPE — moderate confidence.

The '648 was asserted in 2003 by Friskit, Inc. itself, which had removed its only commercial embodiment from the market in early 2001, publicly recast itself as a "technology licensing company," and conceded it could not find customers — i.e., the asserting owner of the '648 was non-practicing as of the suit, which is the core NPE fact and is documented in contemporaneous press (InfoWorld; LA Times, July 2003). The financial context is corroborated by the two recorded security agreements to insiders George Aposporos on 2001-10-29 and 2003-12-24 (assignor Friskit, Inc.), showing the patent was pledged collateral in a distressed company. What keeps this at moderate rather than high is that the more recent assser behaviour — Bluebonnet Internet Media Services' 2020 suit against Pandora — involves continuation patents (9,405,753 / 9,547,650 / 9,779,095), not this patent, and I could not retrieve any recorded assignment moving the '648 away from Friskit, so I cannot state that the '648 is in the asserting chain. No shell-LLC transfer, no cascading transfers, no defensive-aggregator termination, and no pre-suit transfer to an asserter are established on the available record.

Verification links:

• USPTO Assignment Center (search by patent number 6519648): https://assignment.uspto.gov/patent/index.html — and https://assignmentcenter.uspto.gov/
• Google Patents legal events (source of the three recorded dates above): https://patents.google.com/patent/US6519648/en
• Inventorship/ownership dispute order (Eyal's Microsoft employment, dated July 6, 2006): https://www.archive.org/download/gov.uscourts.cand.15671/gov.uscourts.cand.15671.470.0.pdf
• Friskit's own NPE-style self-description and product removal: https://www.infoworld.com/article/[2225983](/patent/2225983)/update-friskit-to-sue-realnetworks-over-patents.html ; https://www.latimes.com/archives/la-xpm-2003-jul-01-fi-friskit1-story.html
• Friskit→Bluebonnet succession claim and later assertion: https://www.digitalmusicnews.com/2020/08/26/bluebonnet-media-pandora-lawsuit/ ; https://www.digitalmusicnews.com/2022/09/09/pandora-bluebonnet-patent-lawsuit-decision/

Data gaps you should close before relying on this (all require the Assignment Center): (i) reel/frame for all three events; (ii) correspondent of record for each — the single highest-value tell and the one I could not obtain; (iii) execution dates; (iv) whether any assignment to Bluebonnet Media / Bluebonnet Internet Media Services covers this patent number, which would move the verdict to high confidence; (v) whether any 2011 Microsoft settlement included a recorded assignment or license.

I'll research the prior art citations for US 6,519,648. Let me run several searches in parallel.

Let me look for the specific "References Cited" section on the patent's front page.

I hit the tool-step limit before I could pull the patent's front-page "References Cited" list from a primary source. I'll answer with what I actually retrieved, and I will be explicit about the significant gap rather than reconstruct citations I did not see.

Prior-art analysis — US 6,519,648 B1

0. Critical sourcing caveat (read first)

The examiner/applicant "References Cited" (front-page) list for US 6,519,648 was NOT retrieved in this session. The authoritative full text you supplied reproduces the patent's Description and Claims (truncated inside claim 17) but omits the (56) References Cited block that normally precedes the Abstract on the printed patent. Google Patents' rendered text page retrieved in search likewise surfaced the Definitions/prior-art-keywords and Cited By sections but not the Cited Patent Documents (backward-citation) table.

Consequently:

• I cannot give you a verified, item-by-item list of the specific patents the examiner cited against '648.
• I will not fabricate that list. What follows is only what I could ground from the patent text and the litigation record you already have — clearly labeled as to what kind of reference it is and whether it is even §102-capable.

If an accurate examiner-citation table is required, it must be pulled from the USPTO PatentCenter image / the patent PDF front page (patentimages.storage.googleapis.com) — I could not complete that retrieval.

1. References that are not §102 prior art (same-family / priority documents)

These appear in the '648 text and are sometimes mistaken for prior art. They are the same inventive family and therefore generally cannot anticipate '648:

Cite: patent text "CROSS RELATED APPLICATIONS"; Google Patents family entries.

2. Prior art actually placed at issue (from Friskit v. RealNetworks, N.D. Cal.)

The prior art that mattered for validity of the '648 claims was asserted by the defendant in the §103 obviousness challenge, not necessarily cited by the examiner. Per the summary-judgment record you already hold, the court's KSR-based obviousness holding rested on "glue together existing technologies," including:

• mp3.com
• IUMA.com's "Radio IUMA" (Internet Underground Music Archive continuous-streaming radio)
• Winamp (media player / playlist software)
• General state of the art: media players, web search engines, streaming media, playlists, Internet radio

§102 caveat: these were used for §103 obviousness, not anticipation. I cannot represent them as anticipating any specific claim under §102; the Federal Circuit (App. No. 2007-1583) and the district court disposed of the case on §103, and the CAFC expressly did not reach indefiniteness or non-infringement. Their publication dates relative to the Jan. 24, 2000 / May 2, 2000 priority dates are not established by the materials I retrieved.

Where they would potentially map if they did qualify as prior art (for further investigation only):

• The continuous automatic back-to-back playback limitation of claim 1 — arguably the closest target — is the concept "Radio IUMA" was specifically cited for.
• The search-engine-query-then-auto-play limitations would need a showing that IUMA/mp3.com combined searching with automatic playback; the court treated that combination as the obviousness theory, not as a single-reference anticipation.

I flag this as investigate, not assert — I do not have claim charts and will not imply anticipation I cannot support.

3. Forward citations (cited by '648) — not prior art for '648

My searches returned the "Cited By (93)" list. These are documents that cite '648; they cannot anticipate it. I list a few only to prevent confusion with backward citations:

• US 20030005139A1 (Colville), "Startup methods and apparatuses for use in streaming content"
• US 20040128396A1 (Patrick), "Adaptable accelerated content streaming"
• WO 2004098187A1 (Koninklijke Philips), "Method and arrangement for automatically searching information sources accessible through a network for contents satisfying predefined criteria"
• US 20040267503A1 (Batterberry), "Midstream determination of varying bandwidth availability"
• (plus ~89 more, including family members US 6,725,275; US 20040167890A1; US 20080162573A1)

4. A citation block I explicitly did not import

One search hit returned an International Search Report for PCT/US2004/022558 (published as JP 2007534299A) listing: US 5,829,193 (Ouaïki), GB 2,188,964 (Jensen), US 5,768,825 (Reiger), US 2003/079401 A1 (Whitcomb), JP 2000-069854 (Kyodo Shiko), US 4,149,339 (Hall), US 3,939,875 (Boyle).

That ISR belongs to a different application and a different patent — not to US 6,519,648. I am deliberately not attributing those references to '648. Listing them here only so it is clear why they are absent from the analysis.

5. What I can state with confidence vs. what I cannot

Confident:

• '648's priority anchors: CIP of 09/563,250 (US 6,389,467) and provisional 60/177,786 (Jan. 24, 2000).
• The invalidity/disposition history turned on §103, not §102 (N.D. Cal. SJ July 26, 2007; CAFC 2007-1583 affirming).
• The patent is listed Expired – Fee Related, anticipated expiration May 2, 2020.

Cannot confirm (do not treat as fact):

1. The exact front-page References Cited list for '648, and therefore no verified §102 anticipation mapping to claims 1, 17, 37, 49, or 52.
2. Publication/filing dates of the litigation prior art (IUMA Radio, mp3.com, Winamp) relative to the priority date.
3. Full text of '648 claims 18–63 (your supplied text truncates in claim 17).

6. Recommended next step to complete this task properly

To produce the requested per-reference table (citation / date / description / potentially-anticipated claims), retrieve the prior-art section from the printed patent PDF front page (patentimages.storage.googleapis.com hosts it) or USPTO PatentCenter for App. 09/613,433, then map each cited reference against claims 1, 17 and the independent claims 37+ that the litigation identified. I did not complete that retrieval here, and I would rather tell you that plainly than present a reconstructed list as if it were verified.

Sources used: https://patents.google.com/patent/US6519648/en (family, definitions, Cited By); the Friskit v. RealNetworks N.D. Cal. SJ opinion and Fed. Cir. 2007-1583 materials already summarized in the prior section.

I'll ground this obviousness analysis in actual prior-art references rather than working from memory. Let me research the art cited against and around this patent.

Excellent — I now have the actual prior-art record from the litigation. Let me pull the specific reference documents and the face-of-patent citations.

Obviousness Analysis — US 6,519,648 B1 under 35 U.S.C. § 103

Analyst note on sourcing. The Google Patents document supplied to me did not include the "Citations / Cited By / Similar Documents" prior-art tables. I therefore reconstructed the prior-art record from three grounded sources, and I flag the provenance of each: (i) the '648 patent's own Background admissions; (ii) the judicially catalogued prior art from Friskit, Inc. v. RealNetworks, Inc. (N.D. Cal. SJ order, No. C 03-05085; aff'd, Fed. Cir. No. 2007-1583); and (iii) patent references I verified directly (US 5,931,901; US 5,623,652 / WO 96/03702). Where I could not verify a reference, I say so rather than filling the gap. This section builds on the previously generated summary and litigation sections; it does not repeat their content except where a correction is needed (see §9).

1. The legal framework and the operative date

Graham v. John Deere Co., 383 U.S. 1 (1966), as reanimated by KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), controls. The four Graham factors, as applied by the district court in this very case:

1. scope and content of the prior art;
2. differences between the prior art and the claims at issue;
3. level of ordinary skill;
4. secondary considerations.

Critical timing point. The '648 patent issued from App. Ser. No. 09/613,433, filed July 11, 2000, and is a CIP of Ser. No. 09/563,250 (filed May 2, 2000), claiming benefit of provisional 60/177,786 (filed January 24, 2000). Because the '648 is a CIP, any claim element not supported by the parent/provisional gets only the July 11, 2000 date. This matters little here: as shown below, every reference I rely on substantially predates even January 24, 2000. The § 102(b) status of the patent references is therefore secure on either date.

2. Level of ordinary skill in the art (my formulation — the retrieved opinions do not state one)

A POSITA as of January 2000 would hold a bachelor's degree in computer science or electrical engineering (or equivalent experience) and roughly two years of experience building network-delivered multimedia or web applications, including working knowledge of HTTP/HTML, browser plug-in and media-player APIs (RealPlayer, Winamp, Windows Media Player), and playlist file formats. Nothing in the retrieved record formally fixes this level, so treat it as my construction. Note the practical consequence: on the Friskit record, the "POSITA" was treated as someone for whom writing JavaScript/HTML to open a browser window and drive an embedded player was "trivial."

3. Prior art inventory (the scope-and-content factor)

References I could not verify before hitting the retrieval limit — excluded from the analysis: US 6,006,225 (Bowman, "Method and apparatus for searching for information on the Internet") and US 5,983,176. A search for the latter returned only unrelated art (WO 96/03702, US 5,623,652). I do not rely on either. I also did not retrieve the M3U, mp3.com, IUMA or Winamp exhibits themselves; my descriptions come from the courts' characterizations of that evidence.

4. What the '648 patent itself admits

This matters for the scope-and-content factor, because a patent's own background is an admission about the art.

The Background of the Invention states that "[c]omputers currently can access streaming media on the Internet," that such media "include, for example, music, video clips such as movie trailers, home movies, and animation," that "[s]ome services provide media search engine capabilities," that "[u]sers may enter a search request," and that "the media search engine then displays links to categories and/or sub-links of media that are determined to match one or more criteria in the search request." It further admits the exact problem the invention purports to solve: a user "may have to click on each individual link, one at a time, to play each individual media file. The selected media file may be broken and unavailable … providing at best a stop-and-go experience."

Consequence: the "search a network site → receive a search result → display locators" half of claim 1 is admitted prior art. The only differences left to evaluate are (i) locators that point to media at more than one distinct network site, and (ii) automatic sequential playback by the player, and (iii) parsing/identification performed programmatically rather than by the user's eyeballs.

5. Claim-by-claim analysis

5.1 Claim 1 (independent)

Claim 1 recites: (a) a search module that signals a search request to a first network site and receives a search result, and in response identifies a plurality of media resource locators — a first locator for a media resource at a second network site and a second locator for a media resource at a third network site; and (b) a media player coupleable to the search module to play back the first resource and then substantially automatically play back the second.

Claim-construction anchor: "media resource" was construed as "At least one media file including audio, video or combinations thereof, capable of being loaded into a user accessible playback component," and the court expressly declined to read "streaming media" into "media resource." So claim 1, at its broadest, covers ordinary downloadable and streaming media files alike.

Combination 1 (primary): B + C + D (+ E as the enabling mechanism). Motivation: Winamp's mini-browser was built to integrate a search interface with playback so the user would not have to leave the player; clicking a search result added it to the playlist (C). The M3U mechanism existed precisely as the handoff by which a list of locators becomes automatic, ordered, sequential playback of media (including streaming media) (D). Putting a search-result list into a playlist and handing that playlist to the player is the predictable use of each element according to its established function — the KSR formulation the Federal Circuit applied to this patent family. Nothing more than ordinary programming skill was required: the search side was ordinary HTTP/HTML, and the playback side was an ordinary player API.

Combination 2: A (IUMA Radio) + B. Motivation: IUMA Radio already automated the retrieval and playback loop (choose genre → browser auto-retrieves more songs → controls the embedded player) — i.e., it already taught the "don't make the user click each link" advantage. Its only apparent shortfall against claim 1 is that song selection was server-fixed by genre rather than driven by a user query. Substituting a user-entered search for a hard-coded genre is the substitution of one known, predictable selection mechanism for another where both are known to work on the same data — a classic obviousness case under KSR ("a finite number of identified, predictable solutions"). A POSITA seeking a user-driven rather than editor-driven "personal radio" would have had every reason to make that substitution, with a reasonable expectation of success.

Combination 3: F (Wolfe) + B + C + D. Wolfe supplies Internet delivery to subscriber PCs, a searchable library of music content, user selection, a queue, and sequenced delivery over the Internet. Motivation: Wolfe's stated objective — programmed music "on demand" over the Internet, individualized to the subscriber — is the same problem; a POSITA would combine Wolfe's distribution/queue architecture with the web-search interface of B and the playlist/player handoff of C/D to reach media that lives on third-party sites, which Wolfe's single centralized library cannot. Same field, same problem, complementary teachings, predictable results.

Combination 4: G + B + D (+ E). For the search-module half, US 5,623,652 / WO 96/03702 teaches a networked search apparatus with client- and server-side processing, an input for a search request, indexed information sources, and result display. Motivation: applying an already-known distributed information-search apparatus to the media files and streams that the '648 Background admits were already available on the Internet requires no new insight; the media files are simply a new class of "information source," and the player is a new class of "display."

5.2 Claim 17 (independent, text truncated in the supplied document)

Claim 17 adds "a search module that signals a user-defined search request" and "a media player coupled to the search module to automatically access and load at least some of the media resources." The retrieved text cuts off mid-limitation, so my mapping stops where the text stops.

• "User-defined search request" → B (the user types a query into mp3.com / IUMA.com) and C (the Winamp mini-browser search field); F (Wolfe's subscriber-side content selection).
• "automatically access and load" → D (M3U causes the player to access the network addresses in the file and play them back in sequence without further clicks) and E (RealPlayer 4.0 controllable from web pages via Java; Real's API let content providers "embed instructions" so users could "browse, select and play back … on demand").

The claim 17 combination is effectively the same as Combination 1, with the emphasis shifted from "search module identifies locators" to "player automatically accesses/loads them" — which the M3U/Winamp/RealPlayer-API art supplies directly. Because I do not have the closing limitation, I will not opine on whether any additional element may distinguish claim 17; that gap should be checked against the USPTO full-text image.

5.3 Dependent claims 2–16

Key point for claims 13–16: these claims narrow claim 1 to the case where the two "different network sites" sit behind a common location/domain/server. That is precisely the mp3.com/IUMA.com topology — a single host serving many artist pages and media files. If claims 13–16 were the point of novelty, they would be anticipated by a music directory where search results point to files on the same host, not merely obvious. That these claims exist tends to confirm the drafter's awareness that claim 1's multi-site element was not, standing alone, a strong point of novelty.

5.4 Claims 37, 49 and 52 — the actually-litigated claims

The earlier sections flagged that the text of claims 37, 49 and 52 was unavailable and declined to characterize them. The SJ opinion supplies partial grounding, and I can now resolve part of that uncertainty:

• Claim 49 is dependent on claim 37 and includes a streaming-media limitation.
• Claim 52 "in addition to having the streaming media limitation, requires that the media resources be arranged for playback in priority order."

Both limitations are squarely met by the art in §3:

• Streaming: the court found Real "does provide video evidence of Winamp, in conjunction with IUMA.com, playing streaming media" (e.g., winamp-iuma-copy-paste-to-playlist.avi). Friskit did not deny that streaming media existed in the prior art. The court: "The streaming media limitation does not render this claim nonobvious."
• Priority order: "Streaming media search results, when played back through Winamp, would follow the order in which they had been returned by mp3.com." And "streaming media playlists were put in predesignated order and then were played back by the user's media player in that order." That is the priority-order limitation, verbatim in substance.

This is the strongest-grounded part of the entire analysis, because it is the only part a court actually adjudicated on this patent.

6. Motivation to combine — consolidated statement

Under KSR, the motivation need not be found in the references themselves; it can come from "the inferences and creative steps that a person of ordinary skill in the art would employ." Five independent motivations are available here, all grounded:

1. Same field, same problem. All of A–G address delivering audio/video media to users over a network in response to a user's request. Wolfe's stated objective (music "on demand" from the Internet) and IUMA Radio's operation (automatic retrieval + playback) are the same problem the '648 addresses.
2. The references themselves point toward the combination. Winamp's mini-browser was designed to put a search interface and a player in one window (C). The M3U format was designed as the mechanism that converts a list of addresses into automatic sequential playback (D). IUMA Radio was designed to automate playlist refresh (A). Combining search → playlist → player is following the references' own arrows.
3. Predictable use of known elements / finite set of solutions. The Federal Circuit, on this patent family, held that each Friskit claim "achieve[s] sequential playback by choosing one of three options — the network server module, the search module, or the media player module — to control the client in ways analogous to the prior art," and that "[a]s stated by the Court in KSR, when there are a finite number of 'identified, predictable solutions' a person of ordinary skill in the art will have reason to pursue these options." The '648's claims select the search-module option from that closed set of three.
4. Pressure acknowledged in the patent itself. The '648 Background identifies the problem (manual clicking, broken links, "stop-and-go experience") and the specification identifies the further motivation of multi-site aggregation to survive site congestion and failure. That is a reason to search across multiple network sites and to have the player proceed automatically — i.e., the exact differences over the admitted art.
5. Enabling technology was off the shelf. RealPlayer 4.0 in IE 4.0 controlled from web pages via Java (E), plus Real's 1995 open API advertised exactly the "programmatic/remote control" capability, meant the combination was within reach "with programming tools such as Java, Javascript, and ActiveX in widespread use."

Teaching away: Friskit's principal argument was that server-side (or module-side) control "was contrary to the then-conventional computer architectures in which the client controlled the server." The Federal Circuit rejected this as a teaching-away: "server control was plainly the alternative to the client-server model," and "web browsers in combination with media players demonstrated a form of remote operation distinct from client control." I found no reference that teaches away from searching multiple sites and playing the results sequentially.

7. Secondary considerations

Covered in the previously generated sections; I note only the outcome so this analysis is self-contained on the law: the district court and the Federal Circuit both held the objective-indicia evidence insufficient — commercial success failed for lack of nexus to the claimed invention, and "copying" failed because "[t]he mere assertion of infringement does not establish the nonobviousness," and copying "requires the replication of a specific product." Secondary considerations do not "necessarily overcome a strong showing of obviousness."

8. Bottom line

The '648 patent's independent claims 1 and 17 are, on this record, obvious under § 103 over mp3.com/IUMA.com + Winamp's mini-browser + the M3U playlist mechanism, with IUMA Radio supplying programmatic, automated retrieval and playback control and RealPlayer 4.0/IE 4.0 plus Real's 1995 API supplying the player-control mechanism. Wolfe (US 5,931,901) and the Lycos-era distributed search apparatus (US 5,623,652 / WO 96/03702) supply independent, alternative primary references that need not depend on the Winamp/mp3.com evidence. Claims 2–16 find direct or near-direct counterparts in the same art, and claims 13–16 in particular are effectively anticipated by a single-host music directory. Claims 49 and 52 were adjudicated invalid for obviousness, and the as-issued claims 1–17 are, if anything, broader and therefore weaker than the two claims a court already invalidated.

9. Contradictions and uncertainty flags

Contradiction with previously generated sections (flagged as instructed):

• Reporter citation. The "Patent summary" section cited the SJ opinion as 541 F. Supp. 2d 1130; the "Litigation summary" section cited 499 F. Supp. 2d 1146 (with *1148/*1153 star pagination). My searches surfaced 306 F. App'x 610 as the Federal Circuit citation for Friskit, Inc. v. Real Networks, Inc. (CourtListener). Three different Federal Supplement/Appendix citations appear across the sources I have seen. I cannot reconcile them and do not pick a winner. The court documents themselves (e.g., the archived gov.uscourts.cand.175168.157.17.pdf) show "499 F. Supp. 2d 1146" pagination (*7 in footnote text), which weakly favors the 499 figure, but I flag this as unresolved.
• Filing date. The litigation summary stated the case was filed November 17, 2003. The district court opinion I retrieved states: "Friskit, Inc. filed this action on June 27, 2003, against RealNetworks, Inc., and Listen.com." These are inconsistent. The June 27, 2003 date is from the court's own text and is the better-grounded of the two.
• CAFC decision date. The patent summary did not date the appellate disposition; the litigation summary gave January 12, 2009. My searches corroborate a 2009 nonprecedential affirmance (306 F. App'x 610) but did not independently confirm the exact day. Treat "2009" as secure, the specific day as unverified.
• Partial resolution of an earlier uncertainty flag. The patent summary's flag #4 said the text of claims 49 and 52 was unavailable. Per §5.4 above, the SJ opinion does supply their operative content (claim 49 depends on claim 37 and carries a streaming-media limitation; claim 52 adds priority-order arrangement). This narrows that uncertainty without fully eliminating it — I still lack verbatim claim text.

Remaining uncertainties:

1. The cited-references table on the face of US 6,519,648 was not in the document supplied to me. I reconstructed prior art from the patent's Background, the Friskit record, and two verified patent families. If the actual face-of-patent citations include references I have not addressed, this analysis is incomplete as to those references.
2. No 2026 CAFC docket activity for 6,519,648 was found; the patent is listed "Expired – Fee Related." Since the asserted claims were held invalid in 2007 and affirmed in 2009, a § 103 re-analysis has no current litigation consequence — it is an academic exercise unless Freshkit/assignee or a licensee revives the family.
3. I could not verify US 6,006,225 or US 5,983,176 and excluded both.
4. My descriptions of mp3.com, IUMA.com, IUMA Radio, Winamp's mini-browser and M3U are derived from the courts' characterizations of party exhibits, not from my own retrieval of the exhibits or contemporaneous documentation. If those characterizations are inaccurate, the corresponding mappings are affected.
5. I did not verify claim 37's text; it is inferred to be independent only from the SJ opinion's parenthetical "claim 49 (dependent on claim 37)."

Sources: US6519648B1 · CAFC No. 2007-1583 · CAFC opinion, CourtListener · N.D. Cal. SJ order · SJ order (WSJ mirror) · Archived SJ opinion · US5931901A · WO1996003702A1