Ask HN: Single-person creations that have stood the test of time?
And I am not talking about only tech project creations. Anything extraordinary that comes to your mind?
And I am not talking about only tech project creations. Anything extraordinary that comes to your mind?
He was burned out from the state of commercial programming around the time and funded a sabbatical with his pension savings to work on Clojure. He had at least 3 attempts to bridge Common Lisp and the JVM or CLR runtimes and he had formed strong opinions on the need for Clojure to be hosted.
He kept up doing 90% of the work with the next few versions and even today it he calls all the shots on its development, it not being a "bazaar" style open source project. Of course it being open source anyone is free to write their own patches and make forks, but generally contributions are more likely to be accepted by the community as libraries, not language changes.
The whole story has been submitted here a few times and is quite interesting: https://download.clojure.org/papers/clojure-hopl-iv-final.pd...
Like the `while` loop is not "part of the language", it is a macro which rewrites it as a `loop`. Also `for` is just a macro, feel free to write your own version without needing to fork the language itself.
This might be a difficult concept to understand to people who have only used Python, JS, C# etc.
It occurred to me a few years later that the helium atoms he snapped together will likely survive until they are swallowed up by a black hole or maybe have a front row seat to a supernova or gamma ray burst. So likely hundreds of billions if not trillions of years.
1. Throw more mass into it, ideally 1+ stars under their own power via Shkavov thruster.
2. Compress it physically with mirrors and particle/photon beams (see inertial confinement fusion)
3. Compress it magnetically with incredulously powerful magnetic fields (see tokamak fusion).
4. Compress it gravitationally by putting a black hole in its center.
5. Make it more explodey by converting half of its mass into antimatter.
When enough "ash" is accumulated (and the star is still big enough) the previous ash begins fusing into some new heavier "ash" rinse and repeat until you get to iron "ash" but no further.
the reason is; every fusion lighter than iron is exothermic
two atoms fuse into something bigger than either, but smaller than both, plus releasing some energy (yay).
when you get to fusing iron you need a ball about 20 miles in diameter and fusing does not give up energy, fusing iron needs to take energy from its environment.
good thing it is not in the middle of a star big enough to grow a twenty mile ball of iron.
and that kids is where supernova come from once iron fusion starts it all goes in moments collapsing the core and the star above where the core was "falls" into that space gravitationally propelled with enough speed that when it hits the opposite side (falling just as fast in the other direction) everything fuses into "whatever"
that is about as exciting as it gets in this universe
The thing that I was interested in is whether or not a helium atom in the Earth‘s atmosphere has a way to be sucked into the fusing layer at some point. It just seems from everything that I’ve learned over the years that there’s a net outflow of that would likely keep the little atom some distance away even if it left Earths gravity well.
I don't know, but think it would be tough for a lone atom.
It seems to me there are three possible fates for that helium, and two of them involve being used up by stellar fusion. The helium atoms get somehow pulled into the sun and so eventually get used up in a fusion reaction, the sun expands during its giant phase and envelops the Earth and the helium eventually gets used up in a fusion reaction, or the helium nuclei gets propelled out of the solar system with the solar wind.
It'd actually be fun to estimate the odds of each option but I honestly have no idea. #3 feels like a long shot.
Would scaling up production of Farnsworth fusors help?
I say supposedly because I know Jonathan Blow paid someone to do all the art for Braid, and licensed all the music (I even saw him demo Braid before release with his programmer art at GDC back in 2008, it did not look anywhere near as pretty, so he made the right choice). But the artwork was so extensive I'm not quite sure that really counts for your criteria.
That may be true for some of the others in the list as well, if that's the case. Stardew Valley I know was all one person, though. Also I really enjoyed Gorogoa, even though it's a smaller game, and that was done by a single guy who quit his cushy tech job and spent 7 years teaching himself how to draw and make games before he released it. It's a super clever interactive comic panel puzzle game that was well-received, though. Think it's on Xbox Game Pass right now.
(But yes it's a bit weird not to say that the feat was, uh, making that whole game).
... because they did not write the compiler for their game.
Similarly, building an entire game from nand gates is a different kind of impressive than using an existing game engine + buying commodity art assets. However, using an existing game engine / art assets will probably result in something that looks better and can be more ambitious.
That's like saying a painting isn't a single person creation because the artist didn't make their own brushes from scratch. Maybe true technically, but splitting hairs to the point that it makes the conversation about something else entirely.
I, pencil is an interesting essay, but it's not what OP is asking about, I don't think.
In Braid's case, it was one developer and he contracted out the artwork, just as much artwork as in those two-person teams. So is it really a single person creation? It either could or it couldn't depending on your criteria (if he made all the critical decisions, some might say he is, while others care more about people who can both code and do art, two different fields that it's rare to find one person that can do both of them well). That's all I'm saying.
Also a compiler is a tool, much like a hammer is to a carpenter. If you want to get that precise then there's no point asking the question, because short of "I found some pretty rocks on the ground and stacked them", there wouldn't be hardly anything that would count as a single person creation.
Exactly, by the same logic Donkey Kong was made by a single person. The art and design was by Shigeru Miyamoto and the development was outsourced to a team of six.
Chuck E Cheeses - Founded by founder of Atari, Solely.
Really stretching the definition of solely when you have to ignore the people who did all the actual work in creating chuck e cheeses - i can “create” a restaurant, but without cooks, servers, hosts, graphic designers, construction/reno people, etc. it’s not going to be fun for the whole family
in other words ; 'cooks, servers' are generic components of a restaurant; the 'fun for the whole family' bit comes in as part of a custom set of things on top of the generic restaurant experience that were thought of and attributed to the people responsible for the concept -- in this case the Founder.
Lords Of Midnight by Mike Singleton springs to mind, as does Sandy White's astonishing Ant Attack.
Also if you count Flash/mobile games, there's a bunch there too. I would count in that instance. I released several Flash games back in the day where I did the code, sometimes art (sometimes I collaborated with an artist), sometimes I did the music too, and at least some of the sound effects (any that couldn't be generated with a synthesizer I purchased). And I released one game on iOS/Xbox 360 where I did everything as well.
But they were tiny games compared to behemoths like Stardew Valley.
Tiny utilities (typically under 100 kb), no ads, available for free, and they do exactly what they say very efficiently. The site has been around since 2001.
Charles Childers' work on the tiny Retro Forth language [2] is also worth pointing out. It was started by Tom Novelli, but according to interwebs, Childers took over in 2001. Really cool language if you like tiny VM-based systems and forth.
Another tiny one-man language is PicoLisp [3]. Created by Alexander Burger in around 1988, and he's still the maintainer. He has been using it in commercial application development ever since.
Today, I'm wishing a long and healthy life to Virgil Dupras' Collapse OS! [4]
EDIT:
Another project I really like is Willus.com's k2pdfopt [5], a small cross-platform tool that optimizes pdf files for mobile readers. Started in around 2011; judging by the website, very probably a single-person thing.
One more is, edbrowse, a "command line editor browser" with ed-like command language [6]. Originally written in 2002 by Karl Dahlke for blind users, but might be interesting to many others as well because of its scripting abilites.
Ah, and also mhwaveedit [7], Magnus Hjorth's wave editor, developed by him since around 2002. And Mark Tyler's mtpaint [8], which was apparently very much inspired by mhwaveedit.
A fascinating single-person made digital audio workstation is Non DAW by Jonathan Moore Liles [9], started in around 2006. I've used it a lot, great modular design, ran really well on an old Thinkpad T42.
Serenity OS [10] has also been in development for 3 years. Andreas Kling's amazing effort.
Possibly Dwarf Fortress [11] would also almost count as a two-man project, by Tarn and Zach Adams, going on since 2002.
Nils M. Holm's creations also deserve a mention for sure. Scheme 9 From Empty Space, Klong array language, several books, etc [12]. Really inspiring guy.
Not wishing to turn HN into Wikipedia, so I'll stop here. The dedication behind this kind of projects is amazing, really.
5: https://www.willus.com/k2pdfopt/
7: https://github.com/magnush/mhwaveedit
8: https://en.wikipedia.org/wiki/MtPaint
[1] Richard Hollins Murray contributed to the idea
You can't quite prove from there that humans were down there as well, but let's say it pushes the balance of evidence in the right direction.
Programmers used to be pretty damn productive.
The core of git is brutally simple and efficient, and maps incredibly well to "a linked list (and hash) of files serialized to disk".
Someone wrote somewhere "I used to wonder if Linus was a genius but Git proves it to be true. Anybody could have (re-)written a *NIX implementation, but Linus invented Git from whole cloth."
Linus copied the design from Monotone, by Grayson Hoare (who also made the original pre-pre-1.0 Rust).
"PS. Don't bother telling me about subversion. If you must, start reading up on "monotone". That seems to be the most viable alternative, but don't pester the developers so much that they don't get any work done. They are already aware of my problems ;)"
time for monotone to be rewritten in Rust
Git has corrupted its own store too many times for me ever to trust it anymore. Almost ready to switch to Fossil.
Check for "memtester", "smartctl", and "badblocks" ... understand what they do and see if they turn up anything.
I guess what I am saying is that there seems to be more than enough blame to tar everybody.
It's worth studying so aspirants understand no achievement exists in a vacuum, but not a means to divide something between solo and collaboration.
It's terribly unintuitive.
But it's also so powerful that nearly every use case you might have can be done with it.
Daemontools, a whole bunch of ciphers, tons of otero software. The man is a living legend.
* https://en.wikipedia.org/wiki/History_of_logarithms * https://en.wikipedia.org/wiki/John_Napier
> Another act which Napier is reported to have done, which may have seemed mystical to the locals, was when Napier removed the pigeons from his estate, since they were eating his grain. Napier caught the pigeons by strewing grain laced with alcohol throughout the field, and then capturing the pigeons once they were too drunk to fly away.
Exiled Kingdoms: https://www.exiledkingdoms.com/forum/index.php
Stardew Valley: https://www.stardewvalley.net/
Instapaper: https://www.instapaper.com/
MeetWell: https://www.meetwell.app/
Also worth noting, Chris Sawyer wrote the original Roller Coaster Tycoon code directly in Assembly. It's probably one of the reasons the simulation game worked so well on earlier systems.
[1] https://www.vice.com/en/article/8xypb5/the-dwarf-fortress-cr...
I love it and use it religiously, I just tracked down an article I remembered reading in 2015 the other day just by searching a keyword.
A Tale of Two Cities, Macbeth, Romeo & Juliet, Beowulf, Pride and Prejudice were the first ones that come to mind, though there are of course many more. Wish I knew more of the non-western classics
[1] https://en.wikipedia.org/wiki/The_Expanse_(novel_series)
Could be down to marketing - we like the idea of an "auteur" with a unique voice when we read a novel. Or it could be that having a consistent "voice" (or writing style I suppose) is a feature most people enjoy having in a novel.
All those book for sale at your grocery store? They sell massive copies and are not written by the author. Douglas Preston, John Grisham, Tom Clancy, and all those romance novels... A lot of those are written by someone AND someone else. In practice they are written by someone else and they slap a big name on it to sell copies. Those books keep the literature world running even if they are not great art.
Literature already is team based.
If book writing operated under the same rules as screenwriting, this would all be a lot more visible.
Don't forget editors. Maybe in some cases their impact is limited, but in other cases whole story lines are excised thanks to them. That's basically their job: to smoothen the narrative.
Most written works probably would look very different if not for the editor.
And they do already split some of the work for the text with people like editors.
There is another example of this effect: mathematical proofs typically have just a few contributors, and very often one. But mathematics has an overall consistency that allows different people's proofs to be simply glued together at the "edges", unlike literature where that doesn't work at all.
Japan still uses fax machines. No logical reasoning behind why.
Your typical TV show has more plot holes than cheesecloth, though. That doesn't seem like a fair comparison.
Brandon Sanderson, a modern fantasy author, is famous for having an entire corporation of editors, storyboarders, and researchers. But only his name goes on the book.
I've studied Shakespeare a fair bit, his plays were certainly not written truly solo. But as the lead creative, alongside with humanity's love of hero worship, only his name stood the test of time.
I think in most cases the best software is made by one or two individuals rather than large teams.
And don't even get me started on how much faster it is. Scrum is the mind-killer.
That person is Daniel Stenberg. He's been the main driver[0] (esp since its been called "curl"), and is still the majority committer, but the project has ~1000[1][2] commit authors (disclosure, I'm one). I'm in no way trying to diminish Daniel's work. It is excellent. Curl has however taken on a life of its own to an extent, with a few other heavy lifters and lots of occasional contributers. I actually think this speaks well to Daniels leadership; he seems to have a loose grip that guides the project and keeps things on the rails with minimal perturbation. He's still a heavy lifter, but not entirely Atlas[3].
[0] https://curl.se/docs/governance.html
[1] https://curl.se/docs/thanks.html
[2] https://daniel.haxx.se/blog/2022/01/30/1000-commit-authors/
Non tech stuff? The standard design nail clipper is something probably most of take for granted. Invented by David Gestetner and used all over the world. I feel like this is the kind of thing that people might end up missing in a real dystopian mad max style world (assuming they become hard to come by). If I have to use a toe knife like the great Frank Reynolds I will not be pleased.
To this day, Gosling is a little perturbed, from what I know. Stallman hasn't made major contributions to the project in well over a decade, and basically anything high level makes him confused. [0]
GCC has a long and storied history, and most of the code written by Stallman originally has been replaced with something better. Similarly, I don't believe he hasn't made major technical contributions in at least 20 years or so.
Chuck Moore devising his own programming language, which runs bare metal on x86 chips, in which he wrote a VLSI design program, to make chips, which run his programming language: it's safe to say this is the most control over a computing stack which a single human has at this moment in history.
Which is not to say GreenArrays is a one-man show, or detract from the hard work everyone else involved has done. I've not spoken to every one of them, but I can't imagine those I have disagreeing with me.
Joseph Shivers - Lycra/Spandex
Albert Hoffman - LSD
Anton Köllisch - MDMA (was killed in WW1)
John Pemberton - Coca Cola (significantly different to the modern drink)
John Baird Glen - Propofol
Percy Spencer - Microwave Oven
It's vanishingly uncommon for a first synthesis to have been undertaken by a true team, although this can get murky recently with drug discovery software narrowing search space. Nevertheless there is always a moment when (barring alien civilization) the compound never existed, then it does: it is nearly always one person doing the chemistry, with (at most) assistants who are fully supervised extensions of his or her will.
Aspirin, for only one example, is the creation of Charles Frédéric Gerhardt.
The guy was trying to make ivory-free pool balls and discovered something that changed the future.
For the better or for worse?
It's a humbling experience to see people like that.
Been around for 27 years now and maintained by Christian Ghisler pretty much on his own.
https://www.pcgamesn.com/rollercoaster-tycoon/code-chris-saw...
(I cannot recommend this service enough.)
Another way to look at it: all the tools you use now are team efforts after a lone developer perfected it to a usable release, where people were largely impressed to pick it up further.
[1][https://public.nrao.edu/news/grote-reber-radio-astronomy-pio...] [2][https://web.archive.org/web/20060927053750/https://www.bigea...] (by John Kraus who ran Ohio's "Big Ear" operation)
Around 1900 the Hubble family lived in Wheaton, Illinois, and as his 7th and 8th grade teacher young Edwin Hubble had a Miss Harriet Grote who later married Schuyler Reber and bore him Grote Reber as their first son. She often commented to Grote that young Edwin Hubble stood out from other students in his class and that she felt he would go far. In later years, when Hubble's fame was spreading she took special pride in his accomplishments and that she had been his teacher."
WOW
https://en.wikipedia.org/wiki/OMAC_(Buddy_Blank)
which has some delicious irony since it is about a "One Man Army Corps".
Kirby worked with Stan Lee and some others at Marvel comics to develop the characters and settings that set the foundation for Marvel being... Marvel. Kirby was kinda resentful that Stan Lee hogged the credits for the work. Particularly Stan Lee had a writing credit for all of the books he was involved in, but "writing" includes both planning the scenario and choosing individual words and the illustrators had a lot of input into the scenario.
Looking at OMAC you see Kirby do a really heroic job. The scenario planning is excellent and the writing at the sentence level is fine, but you can see that Lee had a special touch for that which Kirby didn't have.
Script:Jack Kirby
Pencils:Jack Kirby
Inks:Mike Royer
Colors:Jerry Serpe
Letters:Mike Royer
Serpe colored the whole comic’s run. In issue 2, D. Bruce Berry took over the inks; in 3 he took over the letters as well. This continued until the final issue (8), which was inked and lettered by Royer. And the final panel was “rewritten and redrawn by someone other than Kirby” due to the series being cancelled.
I have seen a lot of Kirby’s pencils (even inked over one of his pages as an exercise) and most of the comic is there in them, but there is a huge difference between them and an inked/colored/lettered page.
SID (the sound chip of the Commodore 64) by Bob Yannes.
John Chowning - FM Synthesis
And then Samuel Johnson's A Dictionary of the English Language, a dictionary researched and written by a single man. He thought it would take three years, ended up taking nine years. All the more impressive for the reckless confidence paid off by diligence.
The famous anecdote Boswell reported about Dr. Johnson's confidence in his own abilities as a scholar:
>ADAMS: But the French Academy, which consists of forty members, took forty years to compile their Dictionary.
>JOHNSON: Sir, thus it is. This is the proportion. Let me see; forty times forty is sixteen hundred. As three to sixteen hundred, so is the proportion of an Englishman to a Frenchman.
https://www.neh.gov/humanities/2009/septemberoctober/feature...
> And Johnson did not write his dictionary alone: He had half a dozen assistants, and the history of lexicography tells us that assistants influence dictionary-making more than either eighteenth-century social hierarchies or the Great Author theory behind Johnson’s reputation admits.
My understanding was that the work of the assistants was to copy down quotations Johnson had researched. I imagine they did other things too, but did they write the dictionary, or did they write down the dictionary? If there's a serious claim that they had input on the content, then that's one thing, otherwise I'll continue hewing to my unreconstructed Great Author narrative :)
I had also thought of Shakespeare when this question was asked, but then thought "well, I'm fairly certain he wrote the plays mostly by himself, but I'm also fairly sure other people, unattributed, committed a line here or there, and his actors probably made contributions that found their way into the folio editions, and so on, so maybe that's not a good answer to this question." I think Johnson, being as much an arrogant misanthrope as he was a titanic scholar, was probably less collaborative than Shakespeare in many ways, including his work on the Dictionary.
https://shiningrocksoftware.com/
One person made an outstanding village-building game that has not only stood the test of time (released 8 years ago) but also spawned dozens of imitators and a robust modding community. It's sold at least 2 million copies (based on outdated info I could find) and is frequently referenced in gaming media with the reverence of a classic.
"Aside from voice acting, soundtrack, and parts of the story, Dust was designed and programmed entirely by Dodrill. A self-taught illustrator and animator, he had previously done artwork and cinematics on Epic Games' Jazz Jackrabbit 2, and was in the process of creating an independent animated film, Elysian Tail.[14] He assumed it would take three months to complete the game; it actually took over three-and-a-half years. He originally envisioned the game as an 8-bit-style platformer, similar to earlier entries in the Castlevania series. Inspirations for the final game came from such titles as Metroid, Golden Axe, and Ys I & II, which Dodrill cites as his favorite games."
Maybe this helps?: https://twitter.com/aerique/status/1498684767696523271
Apologies for the initial claim.
Locate, sure.
newLISP [2] is still maintained and Lutz is frequent on forum.
Kawa Scheme [3], very good optimizing scheme on JVM, started 24 years ago. Per Bothner, author, is frequent on forum, although he'd like someone to continue developing it.
PicoLisp [4], started 34 years ago. Alexander Burger is still maintaining it, latest version was 2 months ago.
LFE [5], started by Robert Virding (one of Erlang authors), 14 years ago. Still developed.
Calibre [6], started 15 years ago by Kovid Goyal. Still in heavy development by Kovid.
[1] https://en.wikipedia.org/wiki/Gambit_(Scheme_implementation)
[2] https://en.wikipedia.org/wiki/NewLISP
[3] https://en.wikipedia.org/wiki/Kawa_(Scheme_implementation)
[4] https://en.wikipedia.org/wiki/PicoLisp
[5] https://lfe.io/
(For those interested in the technical aspects of how to compare resampling algorithms, you take a sound, graph it, speed it up then slow it down, graph it again. Comparison of the two graphs would ideally be identical but in practice that's nearly impossible. Mainly you're looking for where frequencies got created that were not originally there.)
https://en.wikipedia.org/wiki/On-Line_Encyclopedia_of_Intege...
https://www.theguardian.com/science/alexs-adventures-in-numb...
See for example https://en.m.wikipedia.org/wiki/Nordstr%C3%B6m%27s_theory_of...
It is a vacation home of sorts - but its a very large nice house ~7,000+ SF very well kept grounds - but rarely seen there.
https://en.wikipedia.org/wiki/Postfix_(software)
Seems initially the work of a single person, though I have no doubt others contributed greatly also.
Also, the architecture of Corbosier. Lots of other examples exist in architecture.
Great question, by the way.
Phil Zimmerman
is a list of projects where any one item - even the small ones, much less ffmpeg and qemu - would be at the top of anyone else's resume....quite impressive!
which does not describe ffmpeg. FB was pretty active for first 4-5 years, but he wasn't the sole dev even then.
Is this applicable to Minecraft if there is Michelangelos David, or Beethovens 5th Symphony?
Or some random tech stuff like Git if all the math is run on the shoulders of the likes of Euklid?
But to add something technical:
Nicolas Appert the guy who did canning before Pasteur could explain how it worked.
Maybe Minecraft will be gone in 5 years because Microsoft releases a Minecraft 2.
Then it will not have stood the test of time. while the Mona Lisa and canned food is still there.
In retrospect that seems like the more obvious Latinization. I wonder why Euclid is more common.
If the Elements was discovered now, we'd probably favour a more direct transliteration of the Greek: Eukleidos.
The Germans keep a lot of the Greek K in names where in general the latinized version with an C is used by others.
The series has been published continuously since the 1970s.
https://en.wikipedia.org/wiki/John_Sinclair_(German_fiction)
It may be on its way out, now.
[0] https://www.thoughtco.com/liquid-paper-bette-nesmith-graham-...
PCem (Sarah Walker)
[1] https://en.wikipedia.org/wiki/Venus_of_Doln%C3%AD_V%C4%9Bsto...
Also as a single-person creation, Linux is basically the implementation of an API which existed before -- not only did this bypass a huge amount of work that would go into designing a fundamentally new operating system, but it meant Linus's work would be useful because lots of existing software would run on it.
(I find it depressing that real innovation in operating systems seems impossible because everybody wants to run old software... Thus all the "bloat" has to get put back into the system.)
They all started as one man projects. Then others started to contribute little, then more.
There is another guy who is building a castle using mostly meieval methods.
I have heard that the RT-11 OS was a one man effort at DEC, which for many years was more popular than Unix on the PDP-11, but I don't know the details. Do any DEC experts know more about this?
The guy is a legend.
More seriously, 1) some mathematical fields, e.g. Galois groups. 2) some specific game strategies: - Bridge has the Stayman Convention (and others) - Chess has Alekhine's Defense (and others)
Talking about when it was in beta and the only way to play it was in the browser.
Then between 2011 and 2014 when it was sold to MS there were a lot more employees, development and outsourcing which is arguably when it exploded.
Wasn’t the JS release done specifically for the tenth anniversary?
His episode on how I built this is REALLY interesting.
The web has done alright.
There are hundreds of them.
Others like Gutenberg's printing press, Franklin's lightning rod, Whitney's cotton gin, Jenner's smallpox vaccine, Nobel's dynamite, Edison's phonograph, etc. etc. etc. are available in lots of lists.
Would Shakespeare have written Romeo and Juliet if there hadn't previously been Pyramus and Thisbe, or many other sources?
But a lot of essential tech tools, apps and libraries, are mostly one man creation (fitting in https://xkcd.com/2347/).
I think that calibre, keepassx (no longer maintained, but living as keepassxc) and curl are some examples.
Newtonian mechanics
The pythagorean theorem
(Am I cheating?)
There's even an XKCD[1] about it.
Sort of: this is the computer in its most raw and pure form, not for work but for play!
7-Zip, LZ4, Zstd.