The Docket — July 21, 2026

I · Patent

TrackTime, LLC v. Amazon.com Services LLC

U.S. Court of Appeals, Federal Circuit · July 2, 2026

A software patent can live or die on whether its own words name a working thing or only a result someone wants. The Federal Circuit sent a claim back for the trial court to decide whether the phrase "executable program code configured to" perform a task describes real structure or an empty function, and the answer shapes how software patents can be drafted at all.[1]

The setup

TrackTime holds two patents on synchronizing recorded audio with a written index, and it sued Amazon and its Audible unit for infringement.[1] The trial court held two claims of one patent indefinite, meaning too unclear to be valid, because it read the claim phrase "executable program code configured to" as a functional term with no disclosed machinery behind it.[1] A jury separately found a claim of the second patent invalid as anticipated, meaning a single earlier product already disclosed everything the claim covered.[1] TrackTime appealed both rulings.[6]

The dispute runs through Section 112(f) of the Patent Act, which lets an inventor claim a "means" for performing a function, but only if the patent's written description actually discloses the structure that carries the function out.[2] A claim that invokes that provision without disclosing the structure is indefinite and void.[2] Whether the provision applies is a question of claim construction that the appeals court reviews de novo, meaning fresh and without deference to the trial judge, while any underlying factual findings stand unless clearly erroneous.[1]

The question

Software has no gears or levers, and code is defined partly by what it does, so the usual line between a structure and a function blurs.[5] The precise issue was whether "executable program code configured to facilitate annotation," and a sister phrase about synchronous playback, are means-plus-function terms that must point to disclosed structure, or instead recite enough of how the code operates to count as structure themselves.[1] The answer was not obvious, because the claims never use the word "means," which raises a presumption that the provision does not apply, yet that presumption can be rebutted, and "code" is a term that can name either a real thing or a bare result.[3]

The reasoning, walked

The panel started from the governing rule that Section 112(f) turns on a distinction between function and structure.[1] A term avoids the provision if a skilled engineer would read it to denote structure, either because the term is itself a known name for structure, or because the surrounding claim language describes how the code operates rather than only the result it reaches.[1] Because the claims omit "means," the court began with the presumption against applying the provision and placed the burden on Amazon to rebut it by a preponderance of the evidence.[3]

That framework comes from Williamson v. Citrix, an en banc decision, meaning one heard by the full court rather than a three-judge panel, which tightened the test for functional claiming.[3] A later panel decision, Dyfan v. Target, then softened the edge for software by holding that claimed code, coupled with language describing its operation, can connote structure, and that off-the-shelf code can serve as structural evidence.[4] The load-bearing move was the court's insistence that Dyfan states no rule that code is always structural, so context decides each case.[5]

Applying that, the panel faulted the trial court for a test done before Dyfan and too crude to survive it.[1] The trial court had called "executable program code" merely generic, but it never asked whether engineers knew the term as a name for structure at the time of the invention, never weighed whether passages describing the code's operation limited the phrase, and never developed a record on how the art actually used the words.[1] The panel added a second point that did real work, observing that the claims require the same software to perform annotation, synchronous playback, and other functions together, which itself constrains what the code must be.[1] On that reasoning it vacated the indefiniteness ruling and remanded for a fresh determination, inviting new evidence and argument.[1]

The second patent followed a separate track and did not survive.[1] TrackTime argued that the earlier program, a courtroom transcript tool called LiveNote, did not disclose a mobile device or touch input, but the court held that the LiveNote user guide's reference to a tablet computer, together with expert testimony, was substantial evidence, meaning enough that a reasonable factfinder could rely on it, from which a jury could find anticipation.[1] One argument was forfeited because TrackTime had not raised it below.[1] The panel was unanimous, with no dissent, yet the opinion reads as a careful correction of doctrine rather than a routine affirmance.[6]

What turns on it

For anyone who drafts or defends software patents, the ruling sharpens a familiar risk, that a claim reciting "code configured to" do something, with no account of how the code operates, can collapse into a means-plus-function term and then fail for want of disclosed structure.[5] The practical lesson is to describe operation and disclose a corresponding algorithm in the specification, not merely name a desired outcome.[5] The decision also warns trial courts away from the shortcut of branding code "generic," and it settles the contextual inquiry, rooted in Williamson and refined by Dyfan, as the governing frame for the next wave of software validity fights.[1] This is a reading of how one court reasoned about claim language, offered to explain the doctrine and not as guidance for any real patent dispute.

Sources

II · Property

Boerschig v. Rio Grande Electric Cooperative, Inc.

Supreme Court of Texas · May 22, 2026

A power line has crossed this Texas ranch since 1947, standing on a right-of-way the cooperative never filed in the public records. The state's high court held that the rancher who bought the land decades later is bound by the modest line he could see, but not by the tripled and taller version the utility built to serve new customers.[1]

The setup

Boerschig bought a 6,397-acre ranch in 2002, and a Rio Grande Electric Cooperative distribution line already ran across it, roughly seventeen to twenty wooden poles carrying four wires over about a mile and a half.[1] The cooperative had built that line in 1947 under an unrecorded 1945 writing signed by the executor of a prior owner's estate, a paper it never entered in the county property records.[1] Boerschig could see the line and it appeared on his survey, but he had no notice of the writing itself.[5] In 2012 the cooperative rebuilt the line without his consent to serve a new gas compressor station and substation, tripling the poles to about sixty, raising their height, and nearly doubling the wires, and Boerschig sued for trespass.[1]

The claim rests on an easement by estoppel, a right of way a court will recognize even without a valid written grant when a landowner represented that an easement existed, the user believed it, and the user relied to its detriment.[3] That doctrine is an equitable exception to the ordinary rule that an easement needs a signed writing and that recording is what binds later buyers.[3] A jury split ten to two found that such an easement existed and that the upgrade stayed within its scope, the trial court entered judgment for the cooperative, and the court of appeals affirmed.[1] Boerschig took the case up on the argument that the evidence was legally insufficient.[2]

The question

Two questions sat at the center.[4] The first was whether an unrecorded writing that fails as a formal easement can still supply evidence of the representation that estoppel requires.[1] The second was harder, what fixes the scope of an easement by estoppel, and whether tripling the line exceeded that scope as a matter of law.[1] Neither answer was obvious, because estoppel is an equitable workaround of the recording system, and the court had to decide how far it can stretch against a buyer who never saw the paper.[4]

The reasoning, walked

Writing for a five-justice majority, Justice Busby set out three elements for an easement by estoppel, a representation that an easement would be conveyed, belief in it, and detrimental reliance.[1] The first holding was that a writing which fails as an express easement can still be some evidence of the representation, so the 1945 paper, plus sixty-five years of construction, use, and maintenance, supported the jury's finding that an easement arose.[1] Because Boerschig could see the line when he bought, he took the ranch subject to that visible burden.[3]

The scope analysis was the heart of the case.[1] The majority held that three things cabin the scope, the landowner's representations about permitted use, the holder's actual uses in reasonable reliance, and the later purchaser's notice of those representations or uses when buying.[1] Its pivotal line was that knowing about a current use does not show that the scope reaches a greater future use that is more than trivial.[1] The court grounded that narrow reading in policy, reasoning that easements by estoppel must be strictly limited because they undercut the writing and recording requirements, and that binding later owners sits in tension with the statute protecting good-faith purchasers.[3]

Applied to the record, the cooperative's reliance investment was the 1947 line as it stood, and no evidence showed the far larger rebuild was needed to keep serving existing customers rather than to reach new ones.[1] Because Boerschig had neither actual nor inquiry notice of the unrecorded writing, meaning no knowledge of it and no visible clue that should have prompted him to search for it, whatever the original parties may have intended about future scope could not bind him.[1] The court therefore held the upgrade exceeded the easement as a matter of law, reversed, rendered judgment for Boerschig on trespass, and remanded for relief, while noting the cooperative could still condemn a larger easement through eminent domain.[2]

Justice Bland, joined by three colleagues, dissented and would have kept the verdict.[1] Her path put the burden on Boerschig to prove the upgrade exceeded scope, treated the jury's answer as a failure to carry that burden rather than an affirmative finding for him, and warned that the majority's rule could stop utilities from ordinary additions or line replacement even for necessary maintenance.[4]

What turns on it

Landowners and utilities across Texas now face an old unrecorded easement frozen at the footprint that was actually built and relied upon, so growth requires a fresh grant or a condemnation.[3] The ruling reinforces the recording system and the protection of buyers who purchase without notice, and it aligns the scope of an estoppel easement with the older principle that use fixes the scope of a prescriptive one.[5] The four-justice dissent marks the real fault line, a contest between shielding the settled expectations of a purchaser and preserving the operating flexibility of infrastructure that has sat on the land for generations.[4] What appears here traces the court's reasoning as a matter of doctrine, and none of it should be taken as advice about an actual property dispute.

Sources