News for hackers. Stuff that matters.
Re:Vint Cerf on why TCP/IP was never patented by emot on Thursday November 19, 2026 @08:06PM (#50568862)
full transcript here: ourfuturestories.eu/vint-cerf
Re:Vint Cerf on why TCP/IP was never patented by commandersaki on Friday November 20, 2026 @01:55AM (#50571059)
I didn't find the answer that interesting:

Vint Cerf: In the early 1970s, while we were developing this, we asked ourselves, should there be patents associated with it? Should there be other constraints? Should it be a licensable thing? The government does license technology that it has paid to develop, but it can also release it openly. And we went through a kind of analysis. Remember that we were thinking this would be a substrate for the American command and control system that would use computers to support command and control. And that we knew that it would have to be interconnected to our allies. So then in 1973, '74, we're saying, well, who are our allies now in 1974? And there's this list. And then we said, well, who were our allies 25 years before that? And it was a different list. And we said, who will our allies be 25 years from now? And we didn't know. So we thought the only way we can be assured that the allies will have access to this is to simply open it up. So we published this openly. All the protocols were made available openly. We encouraged implementations and interoperability. Interoperability. Could I just say one thing about that word, interoperability? Literally, everything in my career has revolved around the ability of disparate and distributed systems to interwork with each other. And interoperability is its own reward. And so I think if there was any theme in my history and in my career, it's been the seeking interoperability among disparate systems

Re:Vint Cerf on why TCP/IP was never patented by lioeters on Friday November 20, 2026 @02:26AM (#50571303)
> interoperability is its own reward

That was one of my favorite points of the interview. He found value in building relationships, improving the space between systems and how they work together. That explains why TCP/IP was not patented - the reasoning was that the protocols needed to adapt to an unknown future when your allies could become enemies, or vice versa, so the best solution was to publish the invention and open it up to everybody.

"Adversarial interoperability" is a phrase I saw somewhere, and may apply here too.

Re:Vint Cerf on why TCP/IP was never patented by harrouet on Friday November 20, 2026 @03:11AM (#50571618)
That was when the US had allies.
Re:Vint Cerf on why TCP/IP was never patented by meialabscom on Thursday November 19, 2026 @08:12PM (#50568911)
I feel like this sort of long term thinking is not likley to happen now. Are we too focused on monetization now? I feel this is mostly the influence of VC's in "innovation".
Re:Vint Cerf on why TCP/IP was never patented by robmusial on Thursday November 19, 2026 @09:24PM (#50569396)
Signal Protocol, RISC-V, WireGuard just off the top of my head are pretty major recent innovations that aren't patent encumbered. There are plenty of people doing interesting work not influenced by VC money.
Re:Vint Cerf on why TCP/IP was never patented by lioeters on Thursday November 19, 2026 @10:57PM (#50569960)
I read that if we didn't have public libraries and someone suggested the concept these days, it would never be built because it sounds like a socialist scheme. "Free access to books and self-education! Who does that profit?" Everyone, of course, society benefits.

But there are good examples of technology for public benefit, and the more I pay attention, I see people are making efforts in various corners, not just in technology but a variety of social activism on collective and personal level.

The Internet, the web, and personal computers were miraculous for that historical period, it liberated the creative energy of following generations. The accelerating enshittification of our present time has managed to capture a lot of that value and cultural relevance, but the cutting edge has moved on elsewhere to find fresh expression.

That is to say, I believe there is hope as long as people continue to engage with meaningful projects creating value for the commons, and even the toxic atmosphere of global exploitation may turn out to be the wake-up call and challenge to humanity to manifest social changes, including socially beneficial technological innovations, for a better future.

Re:Vint Cerf on why TCP/IP was never patented by miki123211 on Friday November 20, 2026 @01:14AM (#50570805)
If libraries were invented in the 21st century, it would have been by a Silicon Valley entrepreneur "disrupting the copyright monopollies" with the innovative practice of "book sharing". They would quickly be criticized by authors and journalists alike and banned by some local governments.

Libraries that required their members to pay fees did in fact exist historically. It's how many libraries in the US got started. They were private enterprises meant to make money (or at least be self-sustaining non-profits relying on a mix of donations and fees), not government-run projects.

Libraries are an obvious workaround for the western copyright system[1]. Because this system lets us share existing book copies without the permission of the author, but doesn't allow us to print new ones for free even when the capability exists, the obvious way to make books cheaper is to serially share them across many readers and have each pay the fraction of the price. This is essentially what these libraries did, with fees as a stand-in for the price sharing that simplified their operation.

[1] Yes, I'm aware that copyright isn't exactly necessary for libraries to exist; high printing costs would do it too. I do think that they would be entirely unnecessary and almost extinct (in well-functioning 21st-century democracies) if not for copyright, given our printing costs.

Re:Vint Cerf on why TCP/IP was never patented by lioeters on Friday November 20, 2026 @03:07AM (#50571591)
"Uber for books"
Re:Vint Cerf on why TCP/IP was never patented by frankling_ on Friday November 20, 2026 @01:21AM (#50570849)
[...] but the cutting edge has moved on elsewhere to find fresh expression.

Any pointers? Been looking for these elsewheres.

Re:Vint Cerf on why TCP/IP was never patented by cowlevel on Friday November 20, 2026 @05:21AM (#50572674)
The pointers are necessarily not available in public, or they'd become enshittified the same way.
Re:Vint Cerf on why TCP/IP was never patented by nine_k on Friday November 20, 2026 @03:14AM (#50571644)
> sounds like a socialist scheme

Public libraries were invented in 1830s, and a few existed (notably Boston's), but the proliferation of tax-supported public libraries is entirely a capitalist scheme, implemented privately by Andrew Carnegie, a steelmaking magnate.

Re:Vint Cerf on why TCP/IP was never patented by ninjalanternshk on Friday November 20, 2026 @06:16AM (#50573116)
The most capitalist thing about capitalism is that capitalists will do anti-capitalist things if it benefits them.
Re:Vint Cerf on why TCP/IP was never patented by brookst on Friday November 20, 2026 @10:19AM (#50575913)
That’s not really how it works… something isn’t a capitalist policy because a capitalist person supported it.

Fire departments are 100% socialist. That does not change depending on who supports or opposes them.

Carnegie was an interesting character. Certainly a capitalist as a person, but his support of tax-funded libraries does not transform public libraries into capitalist institutions.

Re:Vint Cerf on why TCP/IP was never patented by xhkkffbf on Friday November 20, 2026 @11:25AM (#50577184)
Free access? Libraries aren't free. You pay taxes for them. Also, libraries don't pirate. They pay for their books.
Re:Vint Cerf on why TCP/IP was never patented by kstenerud on Friday November 20, 2026 @12:58AM (#50570701)
JSON, HTTP, Markdown, CSV, HTML, almost every programming language, SSH, AV1, ODF, Open protocol, Linux, POSIX, OpenGL, Wayland, Vulkan, Opus, SeaBIOS...
Re:Vint Cerf on why TCP/IP was never patented by hn_submit on Friday November 20, 2026 @02:35AM (#50571366)
My favorite non-patented CS invention of all time is...(drum roll) UTF-8!!

Many people don't appreciate the amount of DRAM and disk memory saved by this brilliant invention. In the 90's Unicode (16-bit) was all the rage. But it required double the amount of storage for text which was mostly ASCII.

The inventors should receive a Turing award (if they haven't already).

Re:Vint Cerf on why TCP/IP was never patented by nine_k on Friday November 20, 2026 @03:01AM (#50571543)
The design of UTF-8 is by Ken Thompson and Rob Pike; the former actually has received the Turing Award (for other work).
Re:Vint Cerf on why TCP/IP was never patented by hn_submit on Friday November 20, 2026 @03:09AM (#50571601)
For his work on C and Unix no doubt.
Re:Vint Cerf on why TCP/IP was never patented by hobo123 on Friday November 20, 2026 @05:16AM (#50572631)
I love that they just invented UTF8 while working on Plan 9 because they didn't like UTF16.
Re:Vint Cerf on why TCP/IP was never patented by trashb on Friday November 20, 2026 @08:25AM (#50574332)
legend has it was designed on a placemat in a New Jersey diner and implemented in a day.

https://www.cl.cam.ac.uk/~mgk25/ucs/utf-8-history.txt

Re:Vint Cerf on why TCP/IP was never patented by coldpie on Friday November 20, 2026 @10:15AM (#50575830)
When I first learned how UTF-8 encoding works (to fix a text rendering bug in Wine) I was blown away. It's such a beautiful and simple encoding and it manages to be backwards-compatible with the most important encoding that existed before it. What an accomplishment.
Re:Vint Cerf on why TCP/IP was never patented by hn_submit on Friday November 20, 2026 @03:07AM (#50571585)
Protocols are a dime a dozen. Every competent developer could make one up just by mixing and matching things.

And RISC-V was originally an educational ISA which got out of hand and found traction in the commercial sector. Primarily because China feared getting locked-out of American and European ISAs.

Re:Vint Cerf on why TCP/IP was never patented by throawayonthe on Friday November 20, 2026 @04:06AM (#50572057)
at least for the signal protocol it doesn't really apply, not at all something 'every' competent developer could make by mixing and matching things.
Re:Vint Cerf on why TCP/IP was never patented by hn_submit on Friday November 20, 2026 @08:50AM (#50574623)
And why's that? I could easily come up with a comparable encryption protocol just by using different ciphers, hashes etc.
Re:Vint Cerf on why TCP/IP was never patented by keeda on Friday November 20, 2026 @11:50AM (#50577717)
I would say the difference is in the high-level structure of the protocol and the guarantees it provides. You’re right that once the structure is established the components could be fungible, but depending on what guarantees are being promised, coming up with the structure can be very non-trivial.

Like, in security there are a zillion ways your protocol could be compromised even if there are no bugs in the code. I’m not familiar with the Signal protocol at all, but I do know the combination of guarantees it provides were unmatched until it came along.

On the flip side “protocol” is a broad term, encompassing even a basic file format or a message between a client and server (hence “protobufs”!) where the only guarantee is “you will find this data at this location.” Those indeed are a dime a dozen, but they’re on the simpler end of the spectrum.

Re:Vint Cerf on why TCP/IP was never patented by hn_submit on Friday November 20, 2026 @01:23PM (#50579411)
There's nothing special about the Signal protocol, except (maybe) the Double Ratchet mechanism which is merely a way to generate ephemeral message keys using hash functions.
Re:Vint Cerf on why TCP/IP was never patented by euroderf on Friday November 20, 2026 @04:12AM (#50572110)
Is USB-C patent-encumbered ? It finally rid the world of serial and parallel ports. A local maximum of nirvana.
Re:Vint Cerf on why TCP/IP was never patented by cowlevel on Friday November 20, 2026 @05:20AM (#50572664)
I'm not sure but don't forget it's illegal because trademarks to say your product has a USB port, unless you're a paying member of the USB-IF and it's been compatibility tested.
Re:Vint Cerf on why TCP/IP was never patented by jbmchuck on Thursday November 19, 2026 @10:34PM (#50569826)
Yup. Now it would be called a moat.
Re:Vint Cerf on why TCP/IP was never patented by wmf on Thursday November 19, 2026 @11:06PM (#50570030)
I think people appreciate the power of free and open standards much more now than in previous decades. Companies have shifted to more subtle forms of monetization than just making everything proprietary.
Re:Vint Cerf on why TCP/IP was never patented by miki123211 on Friday November 20, 2026 @01:17AM (#50570824)
Open standards have much better network effects, and so they're generally favored by the market.

Companies would prefer for their standards to be proprietary, but open standards products often do better than proprietary ones, because of said network effects.

This is not always true; sometimes the open standard and the required interoperability hampers innovation, hence why most enterprised moved off of standards-based Cisco products to Slack and Zoom.

Re:Vint Cerf on why TCP/IP was never patented by abofh on Thursday November 19, 2026 @11:29PM (#50570178)
Well, exactly. His boss wasn't measuring him by patents
Re:Vint Cerf on why TCP/IP was never patented by ares623 on Friday November 20, 2026 @01:16AM (#50570817)
Yes. The "open web" would be considered an abject failure of maximizing revenue and shareholder value.
Re:Vint Cerf on why TCP/IP was never patented by Ey7NFZ3P0nzAe on Thursday November 19, 2026 @10:02PM (#50569636)
Reminds me of:

> - Who owns the patent on this vaccine?

> - Jonas Salk: Well, the people, I would say. There is no patent. Could you patent the sun?

- Jonas Edward Salk (October 28 1914 – June 23 1995) was a medical researcher and author, the inventor of the Salk vaccine against Polio

Re:Vint Cerf on why TCP/IP was never patented by saghm on Thursday November 19, 2026 @10:16PM (#50569729)
Also worth remembering that the Eisenhower administration tried to resist giving the vaccine out for free because it was "socialized medicine".
Re:Vint Cerf on why TCP/IP was never patented by rexpop on Friday November 20, 2026 @01:12AM (#50570792)
Yes, senator Barry Goldwater likewise characterized a government vaccination program as “socialized medicine by the back door, not the front door.”

Goldwater would later to on to commit such atrocities as voting against the Civil Rights Act of 1964 and floating the use of atomic weapons in Vietnam.

Re:Vint Cerf on why TCP/IP was never patented by nine_k on Friday November 20, 2026 @03:28AM (#50571767)
Giving valuable stuff out for free is often unwise. It won't be valued and hence wasted and squandered. It would be hoarded and resold if seen as valuable, e.g. if the supply is limited.

There should be a cost, but it does not have to be monetary. A bureaucratic cost often works well enough: you don't pay for the stuff but you give identity in exchange, and are tracked to limit repeat orders. Or you can buy any amount of the stuff at the market price, but without all that.

Re:Vint Cerf on why TCP/IP was never patented by euroderf on Friday November 20, 2026 @04:21AM (#50572186)
Email awaits some such solution.
Re:Vint Cerf on why TCP/IP was never patented by cowlevel on Friday November 20, 2026 @05:23AM (#50572687)
It's called - and Hotmail had this in the 90s - exclusive spam filter. Any email from anyone not on your contact list is deleted.
Re:Vint Cerf on why TCP/IP was never patented by saghm on Friday November 20, 2026 @08:35AM (#50574456)
> A bureaucratic cost often works well enough: you don't pay for the stuff but you give identity in exchange, and are tracked to limit repeat orders.

You're describing literally exactly how vaccines are distributed, surrounded by a lot of apologism for the arguments against the exact thing you described.

Re:Vint Cerf on why TCP/IP was never patented by nobody9999 on Friday November 20, 2026 @03:48AM (#50571925)
[delayed]
Re:Vint Cerf on why TCP/IP was never patented by saghm on Friday November 20, 2026 @08:42AM (#50574530)
From your first source:

> President Dwight D. Eisenhower expressed his belief that every child should receive the polio vaccine, without indicating how that would happen. Eisenhower charged Health, Education and Welfare Secretary Ovetta Culp Hobby to work out the details in coordination with Surgeon General Leonard Scheele.

> Congressional Democrats advocated for a plan that would make the polio vaccine free to everyone, which Hobby rejected as a “back door to socialized medicine.”

Yes, there are a lot of other details in the history of the polio vaccine, but I fail to see for this isn't exactly what I described. The person in Eisenhower's cabinet who he put in charge of figuring this out used those words in the context I described. The fact that it was preceded by "back door to" is only as "complicated and nuanced" as the understanding that a back door still leads to the same house as the front door.

Re:Vint Cerf on why TCP/IP was never patented by nobody9999 on Friday November 20, 2026 @09:49AM (#50575412)
[delayed]
Re:Vint Cerf on why TCP/IP was never patented by saghm on Friday November 20, 2026 @12:33PM (#50578596)
I didn't said "insisted on not providing the vaccine for free". I said "tried to resist", which implies that they acquiesced. Why do you insist that this is not an accurate picture?
Re:Vint Cerf on why TCP/IP was never patented by nobody9999 on Friday November 20, 2026 @01:16PM (#50579315)
[delayed]
Re:Vint Cerf on why TCP/IP was never patented by crossroadsguy on Friday November 20, 2026 @01:05AM (#50570761)
[delayed]
Re:Vint Cerf on why TCP/IP was never patented by nine_k on Friday November 20, 2026 @03:19AM (#50571697)
Novelty would be hard to demonstrate. The amount of prior art has a literally galactic scale.
Re:Vint Cerf on why TCP/IP was never patented by skinfaxi on Friday November 20, 2026 @09:21AM (#50575025)
If the prior art is nature it's still patentable no? Like if you designed a helicopter that flew like a dragonfly?
Re:Vint Cerf on why TCP/IP was never patented by kps on Friday November 20, 2026 @09:57AM (#50575544)
They did. US 2780765 Solar energy converting apparatus (1954).
Re:Vint Cerf on why TCP/IP was never patented by brookst on Friday November 20, 2026 @10:23AM (#50575994)
There are many patents on solar panel tech. That is not patenting the sun, any more than some EV motor patent is “patenting roads”.
Re:Vint Cerf on why TCP/IP was never patented by acheron on Thursday November 19, 2026 @11:29PM (#50570183)
And then he went to work for Google.
Re:Vint Cerf on why TCP/IP was never patented by paul7986 on Friday November 20, 2026 @12:07AM (#50570405)
I heard him speak at Google in DC in 2014. It was interesting...
Re:Vint Cerf on why TCP/IP was never patented by wetwater on Friday November 20, 2026 @12:52AM (#50570671)
anything in particular ?
Re:Vint Cerf on why TCP/IP was never patented by siren2026 on Friday November 20, 2026 @01:55AM (#50571060)
this is extremely sad.

That a spyware adtech company is now even able to hire one of the greatest principled brains of our generation. For what it is worth he got hired when most of us still had faith in Google.

Re:Vint Cerf on why TCP/IP was never patented by pak9rabid on Friday November 20, 2026 @12:29PM (#50578536)
Could be worse, at least he didn't accidentally end up at Facebook like a certain hero of mine did.
Re:Vint Cerf on why TCP/IP was never patented by therealmarv on Friday November 20, 2026 @09:52AM (#50575473)
It's worth it to research what he worked on while being at Google.
Re:Vint Cerf on why TCP/IP was never patented by 7e on Friday November 20, 2026 @12:04AM (#50570390)
In the early 70s the government owned most DARPA-funded research. So there would have been little point to pursuing a patent.
Re:Vint Cerf on why TCP/IP was never patented by JSR_FDED on Friday November 20, 2026 @12:45AM (#50570621)
I was in a meeting with Vint Cerf 20+ years ago. He’s an engineer’s engineer and you immediately pick up on the combination of intelligence, focus, discipline and an ability to zoom out and see the bigger picture. Very impressive. He’s also humble and very approachable - and dresses better than you :-)
Re:Vint Cerf on why TCP/IP was never patented by jibalt on Friday November 20, 2026 @01:22AM (#50570854)
When I worked for the UCLA CompSci dept in 1969 I dressed like a hippie (as did Jon Postel and many others), but Vint always wore a 3-piece suit and carried a briefcase. (His bud Steve Crocker was more a pocket protector engineer type.)
Re:Vint Cerf on why TCP/IP was never patented by jibalt on Friday November 20, 2026 @01:18AM (#50570832)
I knew Vint when I worked in the UCLA CompSci dept under Steve Crocker. Both were super nice guys. (Jon Postel, the third in the Van Nuys High triumvirate, was far more inscrutable.)
Re:Vint Cerf on why TCP/IP was never patented by ehecatl42 on Friday November 20, 2026 @03:35AM (#50571823)
I found myself at a dinner table with him during an ACM conference. The fellow's manners are as impeccable as his clothing. The sort of man who irons his own socks... unlike RMS, with whom I also had a one-on-one dinner after bringing him onto campus.

The only two computing legends I ever personally met.

Re:Vint Cerf on why TCP/IP was never patented by hn_submit on Friday November 20, 2026 @02:02AM (#50571104)
What interests me even more is that TCP/IP was never "finished." Cerf and others still had a to-do list they were working on when the DoD halted development and froze TCP/IP.

What was this list of to-do items (I've seen it listed once) and what-if they had finished it? What would the world look like then?

Re:Vint Cerf on why TCP/IP was never patented by p_l on Friday November 20, 2026 @03:05AM (#50571568)
32bit addressing was supposed to be temporary stopgap with v4 being retired by 1990, for example
Re:Vint Cerf on why TCP/IP was never patented by nticompass on Friday November 20, 2026 @10:49AM (#50576504)
Nothing is more permanent than a temporary solution.
Re:Vint Cerf on why TCP/IP was never patented by LeFantome on Friday November 20, 2026 @11:35AM (#50577388)
I read the other day that the 8086 and its ISA was a “temporary solution”. 50 years and counting on that one too.
Re:Vint Cerf on why TCP/IP was never patented by 1718627440 on Friday November 20, 2026 @04:54AM (#50572442)
Is there software that is ever finished and doesn't have a todo list?
Re:Vint Cerf on why TCP/IP was never patented by prmoustache on Friday November 20, 2026 @05:54AM (#50572926)
There used to be many of them when internet wasn't widely available outside of universities and customers didn't expect to be beta testers.
Re:Vint Cerf on why TCP/IP was never patented by mhh__ on Friday November 20, 2026 @07:37AM (#50573846)
Quant finance is interesting because often your pricer for a given product will be useless until it's done and then never touched again whereas most software is useful from the beginning but never finished.
Re:Vint Cerf on why TCP/IP was never patented by amomchilov on Friday November 20, 2026 @08:25AM (#50574330)
Are there ever new “insights” that would lead to an update?
Re:Vint Cerf on why TCP/IP was never patented by eugenekay on Friday November 20, 2026 @11:00AM (#50576729)
That becomes a new Product, sold at a higher rate to Favored clients.
Re:Vint Cerf on why TCP/IP was never patented by mhh__ on Friday November 20, 2026 @12:49PM (#50578885)
Yes, typically a different class of models either from some new burst of creativity or computational tractability.
Re:Vint Cerf on why TCP/IP was never patented by knome on Friday November 20, 2026 @11:40AM (#50577511)
Software, like art, is never finished. Only abandoned.
Re:Vint Cerf on why TCP/IP was never patented by ilkkao on Friday November 20, 2026 @06:19AM (#50573142)
TCP and IP both have extension support (options). New options are still being standardized. Latest TCP option is RFC9768, added this year.
Re:Vint Cerf on why TCP/IP was never patented by sybercecurity on Friday November 20, 2026 @12:09PM (#50578124)
That's the joke, it's never finished. The Internet is basically an instance of building a plane in flight that succeeded.
Re:Vint Cerf on why TCP/IP was never patented by zoobab on Friday November 20, 2026 @03:20AM (#50571702)
Remember that computer programs are not patentable, in the US under SCOTUS jurisprudence Alice, and in the EU under EPC art52.2.

Nevertheless, the Unified Patent Court just ignored the exclusion for a smartwatch, saying having a button bypasses the exception.

Re:Vint Cerf on why TCP/IP was never patented by cowlevel on Friday November 20, 2026 @05:19AM (#50572652)
I believe the phrase "computer programs are not patentable" betrays a complete misunderstanding of patents, it is like saying that numbers are not copyrightable.
Re:Vint Cerf on why TCP/IP was never patented by dv_dt on Friday November 20, 2026 @05:20AM (#50572660)
Software hardware combinations have always been patentable even if software alone only has copyright protection
Re:Vint Cerf on why TCP/IP was never patented by freejazz on Friday November 20, 2026 @09:19AM (#50574995)
That's not what Alice stands for at all
Re:Vint Cerf on why TCP/IP was never patented by kube-system on Friday November 20, 2026 @10:30AM (#50576158)
True, you only have to worry about patents if you want to do things like... actually use software on a computer, or provide it to others to use on a computer.
Re:Vint Cerf on why TCP/IP was never patented by crazybonkersai on Friday November 20, 2026 @04:29AM (#50572248)
A cool thing about Vint Cerf is that he looks like The architect from The Matrix series. I wonder if the cast was intentional or it is just a coincidence
Re:Vint Cerf on why TCP/IP was never patented by gbxyz on Friday November 20, 2026 @07:03AM (#50573555)
It was intentional.
Re:Vint Cerf on why TCP/IP was never patented by justin66 on Friday November 20, 2026 @10:23AM (#50575988)
[delayed]
Re:Vint Cerf on why TCP/IP was never patented by pak9rabid on Friday November 20, 2026 @12:22PM (#50578375)
Holy jeez, you're right!
Re:Vint Cerf on why TCP/IP was never patented by chuckadams on Friday November 20, 2026 @10:04AM (#50575660)
[delayed]
Re:Vint Cerf on why TCP/IP was never patented by keeda on Friday November 20, 2026 @11:58AM (#50577893)
Hah, I never realized Cerf was a homonym for “surf!” Nominative determinism or an eponymic Easter Egg?