Summary

  • A Wikipedia page reports a 382 km (237 mi) Wi-Fi connection in June 2007, held by Ermanno Pietrosemoli and EsLaRed of Venezuela, transferring about 3 MB/s of data.
  • A separate institutional record counts some 50 ICTP Wireless Lab training activities and more than 2,400 participants since 1995, lists EsLaRed among its links, and does not name Pietrosemoli or assign him any role.
  • Between those two facts sits the question this article can only answer negatively: what documented method, curriculum, instructor roster or later independent deployment carried the practice from the record to the people who were taught.

The number that gets repeated is 382 kilometres. It is a good number, and it is genuinely in the record. The Wikipedia entry reports that among long-distance links using non-standard devices, one in June 2007 covered 382 km (237 mi), held by Ermanno Pietrosemoli and EsLaRed of Venezuela, transferring about 3 MB/s of data. The Internet Hall of Fame inductee profile states the same achievement in its opening line: he "holds the world's record for the longest WiFi connection at 382 km."

Read that carefully and the credit is not a solo performance. It is a person and an organisation, named together, in the same clause. That distinction is the whole story, and it is the first thing a celebratory version of this article would sand off.

What the excerpt actually contains

The passage held for that Wikipedia page breaks off after the word "between." Everything that would make the demonstration reproducible sits on the far side of that break: which two sites were joined, over what terrain, at what frequency, with which radios and antennas, under whose regulatory authority, measured how, in which direction, over what interval. None of it is in the record I can cite. Anyone who tells you the equipment list or the link budget is telling you something this evidence does not contain.

The number itself deserves the same care. "About 3 MB/s" is a rate without a stated method. Payload or raw link rate, one direction or two, averaged over what window, against what packet loss — the excerpt is silent. A figure that travels without its measurement protocol is a headline, not a specification. It proves that something impressive happened at a stated distance on a stated date. It does not prove that anyone else could rebuild it.

Two different kinds of credit

The inductee profile is the more generous document. It says he has been "promoting low-cost data communications technologies for Internet access by teaching in many countries," and it lists wireless links built in twenty named countries: Argentina, Comoros, Ecuador, the United States, Spain, Italy, Kenya, Lesotho, Malawi, Mexico, Micronesia, Morocco, Mozambique, Nicaragua, Peru, Trinidad, Senegal, Seychelles, Venezuela and Zambia.

That is a substantial claim about reach. It is also, in the form held, a profile's summary of a career rather than a record of instruction. It names no course, no syllabus, no cohort, no date for any of the teaching, and no institution that hosted it. Twenty countries is a map, not a curriculum.

The institution that kept the tally

The most concrete training number in this record belongs to an institution that does not name the man. The ICTP T/ICT4D Wireless Lab page reports that "since 1995, the Wireless Lab has held some 50 training activities attracting more than 2400 participants, from Africa, Asia, Oceania, Europe and Americas as well as from international organizations." Among the page's links: ITU/BDT and EsLaRed.

Put plainly: the lab has a documented training history, and EsLaRed appears in its link list. Pietrosemoli does not appear in the passage at all. No role is stated. No activity is attributed to him, no participant was taught by him in that sentence, and the roughly 50 activities cannot be billed to his account without evidence that does not exist in this record. The appearance of EsLaRed in a links list also tells us nothing about what kind of relationship that was — informational, collaborative, contractual or merely referential.

The chronology matters just as much. The training record begins in 1995. The reported link is dated June 2007. Twelve years separate them, in that order. The cumulative participant total cannot be an effect of a demonstration that came later, and the demonstration cannot be the origin of a programme already running.

Four stages, and only two of them are populated

Capability transfer has stages, and conflating them is how a distance record gets mistaken for a legacy.

The first is proof of possibility: somebody demonstrated that the thing can be done at this distance with this class of equipment. The June 2007 link occupies this stage, under the limits noted above.

The second is codification: the method written down well enough that someone else can follow it without the original team in the room. This record contains no such document. The 382 km figure is not a method.

The third is dissemination: named instruction, in a named setting, to named participants, with material that can be examined. The ICTP aggregate is the closest thing held, and it is an institution's tally of activities and attendance with no curriculum attached and no instructor named. Attendance is a count of people in a room. It is not a count of people who can now build the thing.

The fourth is outcome: independent teams deploying, extending or improving on what they were taught. Nothing in this record reaches that stage.

So the honest scorecard reads: possibility, documented at a distance and a date; codification, absent; dissemination, asserted at institutional scale but not tied to the person; outcomes, absent. Two populated stages, two empty, and no joined evidence running between the ends.

What would close the gap

Evidence of transfer is not exotic. It is boring, and that is why it is rare. A dated workshop roster with his name on the instructor line. A course syllabus listing the exercises. A slide deck preserved from a session. A participant who later built a link and said where the method came from. A contemporaneous technical report giving the endpoint sites and the measurement method, so the demonstration itself becomes teachable rather than merely impressive.

None of these were obtained. Stating that is not a complaint about the record; it is a description of what the record holds. The distinction between "this happened and was not written down" and "this happened and the writing is not in front of me" is one this article refuses to blur. What can be said is narrower and firmer: the documents reviewed establish a spectacular distance and an institutional training programme of real scale, and they do not connect the two.

The uncomfortable version

There is a version of this story in which the record-holder and the teaching institution are the same achievement at two moments in time, and the reader is invited to feel that the second grew out of the first.

Nothing held supports it.

The teaching record predates the demonstration by twelve years, the institution that kept the tally does not mention him, and the profile that praises his teaching attaches no document to it.

What survives is more interesting than the myth, and less flattering to everyone involved. A very long link proved something about physics and antennas and stubbornness. A training programme proved that an institution could convene people at scale for three decades.

Neither one, on the paper available, proves that the capability moved.

Pietrosemoli's name sits at the junction, and that is precisely why the junction deserves a receipt rather than a tribute. The EsLaRed site in the passage held presents the organisation as an online reference platform for technology information in Venezuela; it displays no founding year, no founder list, no office holder and no institutional chronology. The 2017 inductee row places his name under the country entry Venezuela in a list, which is a fact about a list and not a statement about what he controlled.

Credit should follow the receipt. On this record the receipt names a team for the distance, an institution for the tally, and leaves the transfer unwritten.

That is the finding, and it is worth publishing precisely because the flattering version is so easy to write and so poorly supported.