Nicolas Fodor was born in 1963 and grew up in Paris. At nineteen, he was hired to book top models and manage a modeling agency — but what he really wanted was to be like the attorneys he worked with. In 1988, he graduated from Université Paris II Panthéon-Assas, one of France's most prestigious law faculties.5 He never practiced. Instead, he taught himself to code — and did something that would define the next thirty years of his life.
In 1989, in Paris, he founded Liaison Micro Systems — one of the first companies to build client-server development tools for Oracle on Mac. By 1991, CERN — the laboratory that would invent the World Wide Web — had licensed his GraphTools software. In 1992, he patented an electronic commerce system48 — years before the term "e-commerce" entered common use.5
In 1992, Fodor moved to New York. And in 1993, in Miami, he published something the world had never seen before: CommSurfer — the first email client for Windows LAN and Internet, designed for mobility.1
He incorporated SetNet in 1994.6 And in 1995, on a trip to Porto Santos with a friend named Jim, struggling with phone lines too unreliable for data, Fodor had the idea that would consume the next two decades of his life: the first mobile app ever conceived — email delivered to any phone, anywhere, without a data connection.5
SetNet's concurrent-session technology handled 400,000 simultaneous users for ZAZ in Brazil in 1996 — years before anyone else had solved the same problem at scale. This architecture is the prior art that would later threaten to invalidate an entire patent empire.
By 2001, SetNet was not a startup. It was infrastructure.
SetNet had moved from Miami to San Mateo County, California. It had powered the launch of Vodafone Live with SFR in France, reaching 15 million subscribers.2 It had launched the first color-screen mobile email for SFR, the first PushMail for smartphones. In 2002, it entered a formal Cooperation Agreement with Hewlett-Packard France7 — HP would act as intermediary between SetNet and SFR, declaring and paying license fees based on actual usage.
The contract was clear: approximately $5,000 per simultaneous session.7 As Vodafone Live grew, as millions of French mobile subscribers used SetNet's email platform, the financial obligation to SetNet grew with it.
Then HP stopped counting.
SFR generated €1.653 billion in revenue from platforms powered by SetNet's technology.
SetNet received €525,000.
The €525,000 payment was not a full and final settlement. An internal HP email from October 1, 200450 explicitly states it was for pre-2003 professional licenses only: "La base éligible pour le montant des licences 2003 était de 1,635,500 Euros avant le comptage de Décembre 2003."
Every consumer who bought a mobile phone from SFR was automatically assigned a phonenumber@sfr.fr email address and could listen to their email over the phone using SetNet's VoxMail technology. This was not an opt-in feature — it was factory-default on every Vodafone Live handset. That means every one of the 15 million subscribers was a SetNet user.5
Under the original HP Cooperation Agreement, the pricing was approximately $5,000 per concurrent session.7 SetNet's filing at the Tribunal de Commerce in February 2007 claimed between €30 million and €6 billion.51 SFR's defense to the judge: they cannot owe €6 billion.
SFR's central defense was simple: we couldn't count the licenses, so we don't know what we owe.
This was a lie. Olivier Cadène, an SFR employee (olivier.cadene@cegetel.fr) in charge of all DRM matters for the SetNet platform,52 provided two sworn declarations (attestations sous serment):
If the 15 million consumer users were properly counted — as Cadène's own software was designed to do — the licensing obligation under the original concurrent-session pricing was orders of magnitude higher than what SFR declared.
The forensic evidence went further. SetNet proved to the judicial expert that SFR submitted two identical user counts from two different database tables — the USERS table and the MAILBOXES table — which by definition cannot produce the same result, since each user could have multiple mailboxes.53 The counts were fabricated. SFR also deleted user records before counting, deflating the numbers — but no trace of these deletions was ever produced.
The Cadène declarations — the highest-value evidence on the scale of proof — were among the three exhibits that disappeared at the Cour d'Appel.
By 2014, every other attempt to kill the case had failed. Inside SetNet's boardroom, Pierson had tried to build a majority with Kleidman to oust Fodor from his own company — a corporate coup that would have ended the litigation. But Fodor had transferred half his shares to Georges Daou, leaving himself a minority. The swing vote was Daou's. He didn't flip.
Kleidman's financing veto had already destroyed the California action and blocked the LAEP investment. But the French case was still alive. The Cadène declarations were devastating — sworn testimony from SFR's own engineer proving the counting fraud. If those exhibits reached the Cour d'Appel, the case could not be stopped.
So they didn't reach the court.
LMT Avocats' Christophe Héry documented what happened in an email to Fodor on March 13, 2014:54
The RPVA — the French courts' electronic filing system — "glitched." The three most critical exhibits vanished. Not random documents. Not procedural filings. The three exhibits that proved SFR fabricated its user counts. The boardroom coup failed. The financing blockade wasn't enough. The proof substitution in the lower court wasn't enough. This was the last move: make the evidence disappear before it reached the appeals court.
The Cour d'Appel refused to reopen pleadings. It then repeated the lower court's reasoning verbatim without re-examining the case, ratified the Znaty expert report "without any further comment," and found SetNet guilty of gross fault for having appealed — awarding HP additional damages for the "frivolous" act of exercising the right to appeal.54
While SetNet's case moved through the French courts, SFR's corporate ownership was undergoing a transformation of extraordinary political significance — one involving the highest levels of the French state.
In early 2014, Vivendi SA began seeking a buyer for SFR. Two bidders emerged: Bouygues Telecom (backed by Economy Minister Arnaud Montebourg, who publicly championed the bid)57 and Patrick Drahi's Numericable, controlled through his Luxembourg-registered, Amsterdam-listed holding company Altice.58 Drahi himself was a Swiss resident with holdings in Guernsey — a corporate structure Montebourg publicly criticized.57
At the Élysée Palace, Emmanuel Macron served as Deputy Secretary-General (secrétaire général adjoint) — not yet a minister, but in one of the most powerful advisory positions in the French executive.56 French media reported that the Élysée did not oppose the Drahi bid — contradicting Montebourg's public position.59
Now read the dates carefully:
Forty-eight hours. That is the gap between the Cour d'Appel's ruling and Vivendi's announcement.
A pending litigation claim of up to €6 billion against SFR — backed by sworn employee declarations, an independent expert assessment, and documented counting fraud — represented an existential threat to the acquisition's valuation. A €17 billion deal cannot close with a €6 billion liability on the books. The Cour d'Appel eliminated that threat on March 12, 2014. The three exhibits that would have proven SFR's fraud had already disappeared via the RPVA "glitch."
The political connections run deeper. Bernard Mourad, a close associate of both Macron and Drahi, helped facilitate the SFR acquisition while working for Drahi.59 In October 2016, Mourad left SFR Group to join Macron's En Marche! movement — the political party that would carry Macron to the presidency six months later.59
By the time the Cour de Cassation issued its final rejection on January 11, 2017, SFR had changed hands over two years earlier. The company that committed the fraud no longer existed in its original form. Drahi's Altice controlled everything. And the minister who had opposed the deal had been fired and replaced by his rival from the Élysée.
In November 2016 — the same month SetNet's case was pending its final appeal — the Autorité de la concurrence opened a "gun-jumping" investigation into Numericable for implementing the SFR merger before receiving regulatory clearance, potentially facing a €500 million fine.62 Even the regulator acknowledged the deal had been rushed.
To understand what happened to SetNet, you need to understand what was happening to the entire mobile email industry at the same time.
In the early 2000s, a company called NTP held patents on wireless email.8 In 2000, NTP sent notices to every major company offering wireless email, offering licenses. None took one. NTP sued Research In Motion — the maker of BlackBerry — in the Eastern District of Virginia, a court known as the "rocket docket" for its speed. RIM fought. RIM lost. In March 2006, RIM paid NTP $612.5 million4 — roughly a third of RIM's $1.8 billion in total cash and investments — to make it stop.
That settlement proved something: wireless email patents were worth fortunes.
Visto Corporation, run by CEO Brian Bogosian9 — formerly of Bell Atlantic, who built Visto from a Redwood Shores startup into a major mobile patent licensor, later describing it as "the industry's first mobile enterprise unicorn" at a self-reported $1.25B valuation10 — had a different strategy. Rather than sue first, Visto built a web of non-disclosure agreements with the world's largest mobile operators — including Vodafone Group, the parent of SFR. Under these NDAs, operators agreed not to disclose that they held Visto patent licenses to their incumbent mobile email vendors — vendors like SetNet, whose technology the patents allegedly covered.
Then Visto sued. Microsoft (December 2005).11 Good Technology (January 2006).12 RIM (May 2006).13 Seven Networks.14 The patents were asserted against everyone building mobile email.
NTP licensed its portfolio to Visto.8 Visto licensed its portfolio to Vodafone, AT&T, Sprint, T-Mobile. SFR — Vodafone's French subsidiary — appears on Visto's published operator customer list alongside these global carriers.
But Visto's ambition went far beyond patent licensing. Visto signed and announced an exclusive global deal with Vodafone to standardize all business email across every phone manufacturer on the Vodafone network worldwide. This was not a licensing agreement — it was a platform play. One company, one email system, every handset, every market. The goal was total control of mobile business email on the world's largest mobile operator.
This was also a plan to kill BlackBerry. At the time, RIM's BlackBerry was the only mobile messaging system that mattered. It was mission-critical infrastructure for governments — so critical that India at one point demanded all BlackBerry servers be hosted in-country to prevent foreign intelligence access. The politicization of mobile communications infrastructure had begun. Whoever controlled the email layer controlled the pipe.
There was no App Store. There was no iPhone. Independent software vendors like SetNet had no distribution channel except through the mobile operators themselves. You didn't build an app and publish it — you negotiated directly with the carrier, on their terms, in their infrastructure. They were the gatekeepers. And when they decided to replace you, there was nowhere else to go.
This is the world Fodor was operating in. No platform to protect him. No alternative distribution. His technology ran inside the operator's network. When SFR and Vodafone decided to replace SetNet with Visto's system — using technology built on patents that Fodor's 1993 prior art predated — there was no App Store to route around them, no direct-to-consumer channel, no cloud deployment. The operator was judge, jury, and executioner.
No mobile operator removes a working, mission-critical email infrastructure serving millions of users without a pre-arranged licensed replacement already under contract. SFR terminated SetNet in 2005 without explanation. Visto lists SFR as a customer. The 2004 MoU — which capped and settled SetNet's most powerful concurrent-session claims for €525,000 — came just before the termination.
Before the 2005 termination, Visto approached Nicolas Fodor with an acquisition offer. Fodor refused.
Nicolas Fodor was the only obstacle. Every other vendor settled, licensed, or was acquired. Good Technology was acquired by Motorola in 2007.15 RIM paid $880 million total across both NTP and Visto settlements.416 The iPhone 3G launched July 11, 2008 with native Exchange ActiveSync corporate email17 — walking directly into this combined patent portfolio.
SetNet, operating on millions of SFR subscribers with CommSurfer/Datawave (1993) — published prior art that predated every Visto patent by years — was the single most dangerous independent vendor alive. Fodor's technology, if properly presented to a court, could have invalidated the entire NTP/Visto portfolio and potentially unwound billions in settlements.
By refusing to sell, Fodor left a $612.5 million ecosystem with a hole in it. The SFR termination closed that hole. Or so they thought.
A French-born inventor published a mobile email system in 1993. Ten years later, the patent industry tried to claim they invented it. His technology was the proof they hadn't.
After the New York Times article, the defendants in the Visto patent litigation — RIM, Microsoft, Good Technology — needed Fodor's original CommSurfer source code. It was stored in a Miami storage unit. A federal court issued a subpoena compelling its production.
Good Technology's General Counsel — a man named Jim, who knew Fodor personally — called Fodor and recommended voluntary cooperation. RIM and Microsoft's lawyers nonetheless petitioned the court. The subpoena was issued.
Fodor needed a lawyer. He asked Alexander Brown — a former Orange Telecom executive — for a recommendation.5 Brown referred him to Morgan, Lewis & Bockius LLP, one of America's largest law firms. Morgan Lewis assigned Andrew Gray — Andrew J. Gray IV, Partner in the Intellectual Property Practice, Palo Alto.19 Physics background. Semiconductor specialist. Named "Innovator of the Year" finalist by The Recorder in 2019.20 Admitted before the U.S. Supreme Court and the USPTO.19 On paper, a strong choice. In practice, the single attorney through whom four concurrent adverse client relationships would flow.
Gray was a problem from day one. Not because he was incompetent. Because of who else he was representing.
HP was not just any Morgan Lewis client. It was one of the firm's most significant relationships:
When Fodor confronted Gray about the HP conflict, Gray told him a "Chinese Wall" had been erected inside the firm. This was false as a matter of law. Ethical screens cure only lateral-hire conflicts. They do not cure concurrent conflicts — representing directly adverse clients in active litigation at the same time. There is no screen that makes that permissible. Misrepresenting this to a client asking about a conflict is itself a professional responsibility violation.
In June 2008, Morgan Lewis drafted SetNet's Series A term sheet (internal document reference: 1-PA/3694111.5, Philadelphia office):
This structure gave Series A investor Peter Kleidman — Cambridge PhD mathematician (finite group theory, 1987),30 six years at Goldman Sachs, fifteen years in investment banking,31 Hollywood executive producer (Wonderland, 2003, starring Val Kilmer),32 and prolific civil litigator across California courts with over a dozen lawsuits including contract disputes, evictions, and civil rights claims against judges33 — a veto over all future financing. He exercised it:
A financing structure designed to be a trap, reviewed by the partner managing the adversary's relationship, drafted by the firm representing the adversary. The concurrent conflict was not incidental. It was structural. It was total.
In March 2009, Fodor delivered the CommSurfer/Datawave source code to Morgan Lewis on a CD-ROM, pursuant to the federal subpoena. RIM and Microsoft were supposed to collect it.
Microsoft's lawyers never came.
The CD-ROM sat uncollected. Microsoft had no interest in deploying prior art that would destabilize its own Visto settlement negotiations and potentially expose its prior NTP settlement to refund claims. The documents sat in Morgan Lewis's custody while the discovery period ran.
Then something unexpected happened: RIM's CEO personally learned that the prior art had never been picked up. Someone at RIM — not a lawyer, someone inside the company — had found out. The CEO alerted RIM's lawyers. RIM's lawyers called Gray immediately: we need to come Thursday to collect the CD-ROM.
Gray called Fodor. Fodor couldn't make Thursday. He said: let's do Monday.
Monday: nobody showed up. Nobody called.
Wednesday, Fodor called Gray.
The deadline had fallen between Thursday and Monday. RIM's lawyers had known this when they called Thursday — it was precisely why the CEO had gotten personally involved and they'd called same-day, ready to fly the next morning. They hadn't told Gray the deadline was Thursday-or-never. Gray had no independent way to know. A lawyer not financially tethered to RIM would have demanded the hard deadline date before calling his client. Gray didn't ask. And when the deadline passed, he didn't call.
Fodor found out because he made the call himself, two days later.
In France, SetNet's case was handled by Francis Teitgen — a partner at Weil, Gotshal & Manges in Paris35 — a name that carries extraordinary weight in French legal circles. Teitgen served as Bâtonnier of the Paris Bar (c. 2000), the highest elected position in one of the world's most prestigious bar associations.36 His father, Pierre-Henri Teitgen, was Minister of Justice under Charles de Gaulle's provisional government (1945–46) and a founding architect of the European Convention on Human Rights.37 Francis later became vice-president and general director of Ouest-France, France's largest-circulation daily newspaper, co-founded by his father in 1944.38
Teitgen abandoned the SetNet case and demanded payment at full hourly rates — repudiating the fee agreement that had been drafted as a fixed cash amount plus a share of litigation proceeds.39 The fixed cash component had already been paid in full, in advance. Notably, his own firm, Weil Gotshal, had declared a conflict of interest but refused to disclose its nature.40 By walking away mid-case and demanding full-rate payment on top of what had already been paid, Teitgen left SetNet without lead counsel in the French proceedings at a critical moment.
What followed was worse. The matter went before the Paris Bar's fee arbitration — the Bâtonnier's own institution — which validated Teitgen's claim in full. The Paris Bar then effectively forbade other lawyers from representing SetNet until the disputed fees were resolved. SetNet was blacklisted from legal representation in Paris — by the institution whose former leader had abandoned the case. Fodor and investor Eric Pierson went to see Basile Ader, the Bar member then in charge of disciplinary matters, seeking help.5 None came.
The Paris Bar then sued SetNet directly. SetNet lost in first instance — because no law firm could defend them. The Bar had forbidden it. Fodor filed the appeal himself, pro se. The Paris Bar never responded. The appeal disappeared into silence.5
The financial burden of changing counsel fell on Thales Martins, a common stock investor who had initially taken $500K alongside Kleidman's two Series A tranches of $500K each. Martins put in additional capital specifically to fund replacement lawyers — each time a new counsel was retained, the Paris Bar's machinery would force them to abandon, creating a cycle of legal paralysis that drained resources while the litigation clock ran. Despite the blacklisting, LMT Avocats (Christophe Héry, Jérôme Rousselle) ultimately took the case49 — the firm whose independent assessment had found 80–90% success probability.
One final detail: HP won the French litigation but never collected. The Tribunal awarded HP approximately €1.5 million in fees and costs. HP never enforced the judgment. A company that spent years fighting SetNet in court, represented by the same law firm that represented SetNet, won — and then walked away from its own victory. The judgment was never worth collecting. The purpose had already been served: SetNet was destroyed.
Meanwhile:
Fodor's documents were used in RIM's USPTO reexamination defense. Fodor has no accounting of what was filed.
The attorney representing Fodor, funded by RIM, telling Fodor he cannot find out what RIM did with Fodor's documents.
The decade-long French litigation ended in 2017 with total defeat.42 Nicolas Fodor filed for Chapter 7 bankruptcy in 2021.43
Morgan Lewis simultaneously represented SetNet/Fodor and HP France — parties in active adversarial litigation in the Tribunal de Commerce de Paris.44 Per se nonconsentable under Rule 1.7(b) and Restatement §122(2)(c). Firm-wide imputation under Rule 1.10.
Gray told Fodor a screen had been erected. Screens cure lateral-hire conflicts (Rule 1.10), not concurrent conflicts (Rule 1.7). Misrepresenting applicable law to a client asking about a conflict is dishonesty under Rule 8.4(c).
RIM paid Fodor's legal fees (~$45,000+) while simultaneously using Fodor's documents in USPTO proceedings and having its lawyers withhold the discovery deadline from Gray.
Rule 1.8(f) requirements: informed consent ✗, independence ✗, confidentiality ✗. None met.
Gray filed Cellcentric patents without freedom-to-operate analysis including HP's portfolio. USPTO duty of candor (Rule 56) created impossible dilemma: disclose HP confidential info (breach Rule 1.6) or withhold material info from USPTO (inequitable conduct). No ethical path existed.
Series A structure (pre-money $200K, 88.2% dilution, 5x redemption from HP France proceeds) reviewed by HP relationship partner Kellerman. Gave Kleidman veto over all financing. Destroyed California action. Blocked LAEP despite 80–90% independent success assessment.
Four concurrent adverse clients through one attorney. Structural breach risk. French criminal exposure for professional secrecy violations (RIN Art. 2; Penal Code Art. 226-13).
| Rule | Violation |
|---|---|
| ABA Rule 1.7 | Concurrent conflict — HP France v. SetNet |
| ABA Rule 1.7 | Concurrent conflict — RIM v. Fodor/SetNet |
| ABA Rule 1.8(b) | Using client information (HP France litigation) against client (SetNet financing) |
| ABA Rule 1.8(f) | Third-party payor — RIM paying fees, controlling representation |
| ABA Rule 1.1 | Patent prosecution malpractice — no FTO, no litigation record |
| ABA Rule 1.6 | Confidentiality breach risk — concurrent adverse clients |
| ABA Rule 1.10 | Firm-wide imputation of Gray's conflicts |
| ABA Rule 8.4(c) | Misrepresentation of Chinese Wall as applicable cure |
| 37 C.F.R. § 1.56 | USPTO duty of candor — impossible dilemma created by conflict |
| FL Rule 4-1.7 | Florida analog (SetNet Florida corporation) |
| French RIN Art. 4.1 | French professional conflict rules |
| French RIN Art. 2 | French professional secrecy |
| Penal Code Art. 226-13 | Criminal breach of professional secrecy |
| Document | Status | Key Facts |
|---|---|---|
| Series A Term Sheet | Available | ML ref 1-PA/3694111.5; Jun 2008; $200K pre-money; HP France named |
| Lipski Expert Report | Available | €16.35M missed revenues + €27M lost opportunities |
| LMT Avocats Memo | Available | Mar 1, 2011; 80–90% success; €15–50M; Kleidman blocked |
| Gray Email Mar 30, 2009 | Available | "Nick cannot get RIM to provide any information..." |
| California Complaint | Available | Filed Jan/Feb 2014; 7 causes of action |
| CIV-110 Dismissal | Uploaded | Jun 27, 2014; WITHOUT PREJUDICE |
| Cellcentric Forensic History | Uploaded | Names Gray; Chinese Wall; Kleidman obstruction |
| NYT Markoff Article | URL | Apr 16, 2007; CommSurfer as prior art |
| Aeon Timeline File | Uploaded | 341 events, 1963–2020 |
| Federal Subpoena | Pending | RIM + Microsoft co-request; March 2009 |
| USPTO Reexam Filings | Pending | What RIM filed with Fodor's documents |
| Patent Applications | Pending | Cellcentric patents filed by Gray |
All four elements met. Relationship: confirmed by billing, term sheet (ML ref 1-PA/3694111.5), emails. Breach: per se concurrent conflict (Rule 1.7); third-party payor (Rule 1.8(f)); patent malpractice (Rule 1.1); financing trap (Rule 1.8(b)); Chinese Wall misrepresentation (Rule 8.4(c)). Causation: but-for the financing trap → California action abandoned; but-for the CD-ROM failure → prior art not deployed; but-for HP conflict → litigation strategy compromised. Damages: €43M+ (Lipski) + U.S. action value + Chapter 7 bankruptcy losses.
State bar complaints available in: Pennsylvania (ML headquarters), California (Palo Alto/SF offices, Gray, Kellerman), Florida (SetNet incorporation). French bar: Conseil National des Barreaux (RIN Art. 4.1 violations).
Series A term sheet drafted under conflict conditions — potentially voidable. Patent applications filed without FTO analysis and without Rule 56 compliance — challengeable for inequitable conduct.
Interference with attorney-client relationship. Unauthorized use of confidential materials produced under federal subpoena. Per se Rule 1.8(f) violation with potential tort exposure.
For press inquiries, legal co-counsel, or document requests, contact Nicolas P. Fodor.
All documents available upon request to verified press and legal counsel.
This site is maintained as a public interest record. thesaga.fodor.app
In HBO's Silicon Valley, the final season follows Richard Hendricks as he pitches a vision that sounds absurd — a new, decentralized internet. No firewalls. No gatekeepers. No government backdoors. Infrastructure owned by the people who use it. The show played it for laughs. The audience laughed.
Nicolas Fodor wasn't laughing. He was building it.
They bankrupted him. They buried his patents. They erased his evidence. And then he went and built the infrastructure that makes the next internet possible.
After losing everything — his company, his patents, his decade-long litigation, his life savings — Fodor did what he has always done since that first line of code in 1989: he built something new.
SETIP.IO is a platform that puts your entire infrastructure — DNS, routing, certificates, authentication, WAF, DDoS protection — into one JSON file that can be versioned, diffed, and deployed. Infrastructure as actual code.65
It's what happens when the man who invented multi-threaded email routing in 1993 spends another thirty years thinking about how the internet should work.
150+ REST API endpoints. Self-hosted, hosted, or hybrid. LXC isolation. WireGuard tunnels. Zero vendor lock-in.
Deploy from a Raspberry Pi. Deploy from bare metal. Deploy from anywhere. Your data never touches their servers unless you want it to.
AI-native: MCP interface for Claude, agents, and CI/CD pipelines to manage infrastructure autonomously.
The same concurrent-session architecture that powered 15 million SFR subscribers in 2002 now powers a multi-level load balancing system with sub-millisecond routing decisions at enterprise scale. The tech was never the problem. The problem was the people who wanted to own it.
Driver.House runs on SETIP.IO. It's an open-source, AI-powered rideshare platform that lets passengers book directly with professional drivers — zero commission, zero surge pricing.66 Smart outbidding algorithms save riders 20–40% on every trip.
The same principle as SetNet Mail in 1995: cut out the middleman, connect people directly, make the technology invisible. Except now the middlemen are Uber and Lyft instead of AOL and CompuServe.
Every startup needs an internet address. Most burn money on cloud infrastructure before they've found a single customer. UrlyUp gives any developer a public HTTPS URL for their localhost in seconds — encrypted through WireGuard, zero configuration, free to start.67
It's the front door to SETIP.IO. And it solves the problem Fodor saw thirty years ago in Porto Santos: how do you get on the internet when the infrastructure doesn't want to let you in?
The idea: startups shouldn't need to spend money on infrastructure until they reach escape velocity. Build first. Get customers. Then scale — on your own terms, on your own hardware, with your own data.
In the Silicon Valley finale (S6E7, "Exit Event"), Russ Hanneman pulls up in his orange McLaren and confronts Richard Hendricks: "Are you full of shit again? You don't really have the tech." Richard's answer: "So what? We didn't have it last time either." He pitches a peer-to-peer internet — no firewalls, no surveillance, no gatekeepers. The show treated it as both dream and cautionary tale. The fictional version was too dangerous to release.
Silicon Valley S6E7 — Russ and Richard discuss the new internet
The real version is already running. It's not on phones — it's on any hardware you own. A Raspberry Pi in your closet. A NUC under your desk. A rack in your garage. SETIP.IO doesn't care. It routes. It encrypts. It deploys. And nobody can take it from you because you own the hardware.
They took his company. They took his patents. They took his evidence. They never took his ability to build.
In 1993, he published the world's first mobile email client. In 1995, he conceived the first mobile app. In 1996, his technology handled 400,000 concurrent users for ZAZ in Brazil. In 2002, it powered 15 million subscribers on SFR. A $17 billion acquisition needed his case to disappear before it could close.
In 2026, the man the New York Times identified as the prior art that could have invalidated an entire patent empire18 is still writing code. Still building infrastructure. Still refusing to be bought, broken, or silenced.
The underdog is still here. And he's still building.
Verification standard: Every factual claim about third parties has been independently verified against public records, court filings, corporate biographies, news archives, and/or primary source documents from a 1,025-document litigation archive. Claims sourced solely from first-person testimony are marked accordingly. Where a claim could not be independently verified, it is labeled as such. Corrections to initial drafts based on verification are documented below.
Source types used in this document:
The following claims were corrected during the verification process:
nytimes.com/2007/04/16/technology/16email.html with Markoff author attribution in URL parameters. Date restored to April 16, 2007. Source upgraded from "Under Review" to "Public Record."