HNHacker News
TopNewBestAskShowJobs

positron26

80 karma · joined July 7, 2025

Maximize what people can do. Enable what is more of what wants to be. Open propels by moving starting lines, ensuring all who will attempt may achieve what they are able.

There are itches too big for a single player to scratch. There are things that only a broad alliance of interests can properly build. Bridge these alignments and independence will cross with intent.

The first of these alignments and the one that enables all others is more and better bridges. Build legitimate authority that can employ pooled resources and capabilities towards important objectives that are not rightly the domain of any one of who comprises the alliance.

These are information communication technology problems. Open source is a development model that is critical to delivering the solutions.

https://discord.gg/gxptWX6aCR https://github.com/positron-solutions/MuTate https://bsky.app/profile/positron26.bsky.social

Ping us on GitHub to get in touch.

submissionscomments
positron26··on The Cathedral, the Bazaar, and the Winchester Mystery House
This... reaction to one of my other comments...

Stating facts is nice, but the conclusion you're trying to get to is just a tangent about ESR. In 1988. How do you relate this to arguments I made?

> later criticised when it was other people's work

Seems like first-hand learning and applying those lessons to more relatable projects, such as those after 1988? Are we still a society that rewards learning or must all mistakes be worn permanently and shamefully so that the malcontents can endlessly self-validate in their misery?

The authz language is pretty funny. Free speech has always been self-inconsistent. If I may use my free speech to organize a fascist takeover of society, is free speech without limit not potentially a tool of its own destruction? If one is, as ESR is, so concerned with free speech, would there not then be a need for authz? If you argue that controls on free speech are a hypocrisy, isn't that also what a fascist would argue while angling to eliminate barriers to the use of free speech so that they can use it to end free speech?

The license is fun. I'm sure I have equally amusing writings stashed away somewhere. Amid this evidence of early tension, stewing, and ideological turmoil, I do sense within ESR a dissatisfaction with the FSF. Would this not foreshadow that CatB was later aimed at the FSF and that ESR was motivated for a long time and therefore, while he ultimately presented a different message informed by many other developments and a long time to think and refine, all along grappling with an irritation at something deep within the FSF that he could not reconcile with?

Doesn't this continue to argue in a very straight line that CatB was all about the FSF and that the cathedral development model's similarities to corporate waterfall were just incidental?

positron26··on The Cathedral, the Bazaar, and the Winchester Mystery House
> The criterion is the size and complexity of the software

The criterion is the development process, not the complexity. Linux is complex, but not a Cathedral.

I don't want to split hair with your words more. For context, FSF hard liners since the dawn of the OSI were distorting the meaning of CatB to deflect criticism from themselves. FSF supporters also very successfully promoted "FLOSS" instead of bare _OSS, giving lots of later-comers the illusion that "free/libre" was an expansion pack for OSS when OSS came later, a very intentional evolution of the dogmatic "free" software movement.

The choice of "Cathedral" is an extremely obvious symbol when you consider the Protestant reformation as a defiance of Vatican, an overly central system where decisions can only flow from the top. There are a lot of metaphors ESR could have chosen from, but the "cathedral" rhymed with the undertones of the real tension between the many OSS practitioners who have divers motivations and the FSF's plan to slap GNU stickers on every piece of software on Earth while blessing their own cardinals at the FSF Vatican and excommunicating any dissent. Given that kind of very overt signalling, it's just not defensible to argue any other primary target than the FSF and the overly central development process they were dependent on to maintain control over projects.

positron26··on The Cathedral, the Bazaar, and the Winchester Mystery House
I used a similarly shaped argument with different nouns to highlight the ambiguity, and now you see why that's problematic. Don't just make blind assertions without linking it back to some concrete, at least arguing that some mechanism was *dominant*.
positron26··on The Cathedral, the Bazaar, and the Winchester Mystery House
> Unix-philosophy small tools that constitute an important part of the GNU project

The statement you chose makes a carve-out for Unix, not GNU. It doesn't support "not really."

positron26··on The Cathedral, the Bazaar, and the Winchester Mystery House
GNU didn't kick anything off. It was an attempt to document something that was already in full swing.

What was in full swing was Open Source, powered by scratch-your-own-itch. What was taking time was for the business world to learn the lessons by both carrot (Linux) and stick (Unix Wars, vendor lock-in, dozens of crappy competing standards). When Steve Balmer winds up using your language, you moved the ball.

Many ideas from The Cathedral & The Bazaar made it into The Lean Startup. The Cathedral development model was more related to waterfall. YC was already chugging along, but you can bet your ass PG was already steeped in the tea.

positron26··on F-15E jet shot down over Iran
Bro just throw out your privileges or pick some solid ground instead of dragging us all into the mud.
positron26··on F-15E jet shot down over Iran
> Rare Earth Minerals... ...unable to repair the damage any time soon (probably for decades)

Look bro, if we can make SR-71s out of pizza ovens, I'm pretty sure somewhere in the CIA can scrounge up a few ounces of gadoluminium. Tankie dreams are placation for those who wait for somebody else to make the birdseed fall from the sky.

positron26··on How the AI Bubble Bursts
When will this concern farm end? Internet is ant-milling harder than a model gone psychotic on synthetic data. Call me when it's over.

Back to the mines. The Vulkan only writes itself when prompted with well-conditioned problem statements.

positron26··on The Cognitive Dark Forest
Let me write a more interesting body. So hiding is the most rational - the only - strategy of survival.

In the beginning, you reached out with reckless abandon. It was fun to banter with dogs online. Nobody would ever see unless they were looking through your wall. There was no search. No comment history. Bumping into someone in the vast night was enough of a miracle. Why hold back? There are some forum warriors on some PHP BB somewhere, but the domains they rule are insignificant. If you're talking to someone, your motivations are rooted somewhere in the grass.

First came the like button. Rather than blindly hoping what you say resonates with the sensibilities of people you probably knew IRL, rather than present your genuine self because there were no scores, the incentive signal would begin to distort us. Then the newsfeed meant that if you got enough likes, you might get a moment of fame. We all knew it was a terrible idea, a force that would only corrupt us. The personal nature of disjoint little walls living in isolation, was becoming replaced by global stack-ranking.

Then the algorithms came. With them came content marketing to jump the line. At first the ten blue links were filling in the sparsity. Along with that came only a little bias, connecting semantically distant topics, but with a little bit of a feedback loop, an resonator with an unknown response curve. Engagement could be measured, and before long, we were chasing the same likes we used to train the system, and trained by our likes, attracted we became to mysterious stable manifolds, chasing the chase we ourselves define, like Nascar, but insidiously more stupid.

Little by little, the incentive trails no longer lead back to the grass. Reality became suspended without support, a self-sustaining virtual reality determined to fight you to prove that it exists, to prove that its conclusions were right. Every out group is understood to be an echo chamber, an ant mill spiraling helplessly, yet cynically, those who understand these mills best also wind them up like beyblades to crash them into other communities, seeking advantage with the asymmetry of outrage. After the battles, say what was made common to say, and you will be rewarded.

The spinning wheels cannot steer themselves and instead are dictated by whichever chaotic divergence generates the most powerful local gravity well, but because the goal of most is to harvest karma at the bottom, and because the mass controls where the bottom is, over and over we find ourselves pushing all others into the nearest pit to more quickly generate the illusion-giving singularity.

Like Darth Nihilus, the internet seeks only to feed, to feed on the validation that only the internet can give, the permission-giving blessings it needs to tell itself why the grass is wrong. All those who speak of grass are wrong. All those who smell of grass wreak and are wrong. We must destroy the grass, all those appeal to grass. After all grass is dust, at last we will project our utopia into reality. At last we will be not only right but so right that our beliefs will project back into reality.

The spaces within this over-connected, globally addressed world grows into a new kind of sparseness, one where all knowledge of grass must be concealed. Those who can ground the conversations in primary sources flee. Those who can color reasoning with nuance instead withdraw. Reality has retreated as the most dominant reverberations roam like the predator cities of Mortal Engines, looking for any invalidating observations to roll over and consume. Any real life must pretend to be a bot to blend in with the background radiation.

Less like Skynet and more like a zombie apocalypse, the threat comes from within, from among us, from our corruptions, from our karma seeking performances, from our lack of any commitment to any underlying reality, from our flawed belief that the information spaces is some kind of reality stone that enables active control instead of a mere reaction, the shadows on the wall, the murky results of the true forms.

Yet in this new darkness, a certain light has always held. What one wishes, one knows another has wished. What one respects, one knows another respects. No matter the limits of self-knowledge, no matter the information desert one has to cross at night to live in instinct, it is an infinitely brighter signal than the cynical self-corruptions of living for the machine, living to win the games whose rules it was our job to write. What one believes, one knows another has believed. Look into your own center and the true center of others you have known.

positron26··on Video Encoding and Decoding with Vulkan Compute Shaders in FFmpeg
SIMT is distinct model. Ergonomics are wildly different. Instead of contracting a long iteration by packing its steps together to make them "wider", you rotate the iteration across cores.

The critical difference is that SIMD and parallel programming are totally different in terms of ergonomics while SIMT is almost exactly the same as parallel programming. You have to design for SIMD and parallelism separately while SIMT and parallelism are essentially the same skill set.

The fan-in / fan-out and iteration rotation are the key skills for SIMT.

positron26··on Video Encoding and Decoding with Vulkan Compute Shaders in FFmpeg
> Most popular codecs were designed decades ago, when video resolutions were far smaller. As resolutions have exploded, those fixed-size minimum units now represent a much smaller fraction of a frame — which means far more of them can be processed in parallel. Modern GPUs have also gained features enabling cross-invocation communication, opening up further optimization opportunities.

One only needs to look at GPU driven rendering and ray tracing in shaders to deduce that shader cores and memory subsystems these days have become flexible enough to do work besides lock-step uniform parallelism where the only difference was the thread ID.

Nobody strives for random access memory read patterns, but the universal popularity of buffer device address and descriptor arrays can be taken somewhat as proof that these indirections are no longer the friction for GPU architectures that they were ten years ago.

At the same time, the languages are no longer as restrictive as they once were. People are recording commands on the GPU. This kind of fiddly serial work is an indication that the ergonomics of CPU programming have less of a relative advantage, and that cuts deeply into the tradeoff costs.

positron26··on This time is different
> However, current predictions about the future of software development (and the world in general) are speculative.

It's amazing to me how those willing to seize on the speculative nature of any ANY uncertainty cannot recognize the inherent uncertainty of the inverse.

> what CS fundamentals do you need

1. Tarski's undefinability theorem 2. Gödel's incompleteness theorems 3. Curry Howard correspondence

And a lot of exposure to deductive reasoning, vague ideas of automated theorem proving and formalization.

I won't pretend its easy, but let's be clear, a small fraction of people who know things are being forced to entertain the hysteria of a vast majority who are unwilling to know things and just go around beating their chests and will continue doing so until the train hits them.

There are 2-3 minor architectural changes in between now and what I would identify as a completely unbounded AGI with clearly discernible dynamic, self-defined objective functions and self-defined procedures for training and inference. It can be done in megabytes. Oh god. Get me out of this forum. I wish to return to my code editor.

positron26··on This time is different
> most of the people that were the loudest won't say they were wrong

I was so expecting to find this wind-up aimed at those peddling the "AI is hype" laziness.

It's laziness because they have little CS fundamentals to base such claims on, and the deductions can be made, just not clearly to people who need to study a lot more.

It's like watching an invisible train (visible to those with strong CS) rolling down the tracks at a leisurely pace. Those sitting in their stalled car on the tracks are busy tweeting about "AI HPY PE TRAIN." Until it wrecks their car, the gimmick is free oxygen. It's a lot easier to write articles than it is to build GPUs and write programs.

positron26··on Gentoo on Codeberg
The reality of good competition is that competitors are built on good, cheap open source. No matter how decentral, a lot of users will want guards at the offramps and onramps. The only path for... everyone to create stronger competitive checks on services they rely on is to make sure that the open foundations are extremely strong.

The alliance any up-and-comers can make with the ecosystem is to develop more of what they host in the open source. In return for starting much closer to the finish line, we only ask that they also make the lines closer for those that come after them.

That's a bit of an indirect idea for today's Joe Internet. Joe Internet is going to hold out waiting for such services to be offered entirely for free, by a magical Github competitor who exists purely to serve in the public interest. Ah yes, Joe Internet means government-funded, but of course government solutions are not solutions for narrow-interest problems like "host my code" that affect only a tiny minority. And so Joe Internet will be waiting for quite some time.

positron26··on America's pensions can't beat Vanguard but they can close a hospital
Just can't help humming in my head reg yoo la toe ree cap chur

Any system that can juice itself by increasing both funding and cost will scale both until the natural incentive gradients (are you smart? do you actually want to do stuff or do you just want a desk job?) vanish into the noise. When everyone went to college, nobody learned anything.

University is one of those things you always want to be capability rather than means gated, but those of means will always want their kids to get in regardless of how they were raised. After all, they worked hard. Why should their kids? They will ally with every convenient rationalization in order to moralize for the politics, taking advantage of arguments about "disadvantaged backgrounds" etc to treat everything as a means problem, but the goal is to dilute the capability aspect, and that robs talent.

If you have exceptional talent, you need to know the truth. Systems naturally try to optimize the dumb-rich to smart ratio so that there's a lot of subsidy available for anyone who actually needs to be there, but consequent GPA inflation demands that we make the education somewhat meaningless, so you're really on your own to set goals, and any good ones are way higher. Check the boxes, take the free lunch, and then treat the overall coddling like a charade that must be ignored. Then again, isn't self direction always that way?

positron26··on Simplifying Vulkan one subsystem at a time
Finding the optimal sub-language is about API coupling with client code, making a moving sweet spot for where bread & butter techniques live.
positron26··on How to effectively write quality code with AI
So now we're pro-slaughter and low-yield agriculture as long as we get to ride the keyboard eh?
positron26··on How to effectively write quality code with AI
Are there still people under the impression that the correct way to use Stack Overflow all these years was to copy & paste without analyzing what the code did and making it fit for purpose?

If I have to say, we're just waiting for the AI concern caucus to get tired of performing for each other and justifying each other's inaction in other facets of their lives.

positron26··on FOSDEM 2026 – Open-Source Conference in Brussels – Day#1 Recap
> tells ICE to seize all weapons in the US

Outside the constitution, outside their jurisdiction, and not lawful.

Popular sovereignty always works. One way or another.

The most wrong opinion in the debate is to claim that we will be punished for open carry demonstrations. Only an abuse victim excuses the attacker, and allowing the attacker to do what they want is just catch 22.

positron26··on FOSDEM 2026 – Open-Source Conference in Brussels – Day#1 Recap
You would have been able to see this by looking at a graph of my karma against the hour of day. There's truly an ideological dispersion across the Atlantic.
positron26··on FOSDEM 2026 – Open-Source Conference in Brussels – Day#1 Recap
> your comment sounds like American don't want to protest against Trump because it is too dangerous

At this time, we are still openly committed to the 2nd amendment in defense of the 1st, 3rd, 4th, 6th and so on. I am encouraging others to participate in open-carry demonstrations to make it clear to the authoritarians that they will not get the intimidating optics of an unopposed crackdown against a helpless crowd that they want. Personally, I grew up shooting things, so handling bootlickers will be natural if it comes to it.

Technically enabled solutions to better communicate, organize, and represent the will of the people would help a lot. Bomb-shelter thinking will not help much. If the US devolves into a Russian style authoritarian state, one where I will no longer be welcomed off the plane, the EU will have more to worry about than Windows. My ideas on the technically enabled side are complex but sound, so I encourage any interested in doing full stack Rust to get a hold of me by clicking links. I'll be finishing up some shader programming and feedback rendering today as the next piece of my strategy.

positron26··on FOSDEM 2026 – Open-Source Conference in Brussels – Day#1 Recap
> How did I lump both together when I specifically called them out separately, lol.

Dude, "American OSS and American corporations" is simple conjunction, a union, treating two things as one so that you can make a single predicate statement about them. If you mean to make separate statements about the two things, maybe don't group them into the same sentence phrase?

I asked you not to group these things together. If this provokes you to begin regurgitating "free/libre" ideology all over again, you obviously think that being asked to separate American OSS and American corporations is somehow incompatible with OSS or "free/libre". American OSS marched in front of Microsoft to demand refunds. Get it right.

positron26··on FOSDEM 2026 – Open-Source Conference in Brussels – Day#1 Recap
The trend is global and inherent to online psychological coupling and self-selection bias. The longer we go without healthy information spaces, the more the population will regress.

There does however seem to be a "free/libre" vs open source rift along the Atlantic ridge, and it is being wedged apart by the US government flirting with a return to isolationism mixed with bullying and self-enforced credible threat geopolitics.

It is really counter-productive for Europeans to think American OSS people are monolithic with US tech giants and the US federal government. Nonetheless, pluralism is good, and innovation will win, so I suppose it's just another hairpin in the game.

positron26··on FOSDEM 2026 – Open-Source Conference in Brussels – Day#1 Recap
> I don't think anyone is confused about American OSS and American corporations run amok...

You literally just lumped it all together, exactly the fallacy I'm voicing my concern about.

> foreign-owned proprietary

OSS is global. "Foreign owned" is relative. If Americans reject "European" open source, it would make zero sense.

> the US model

What even is "the US model?" The things that are being described as "American" or "European" here are not inherently national.

positron26··on FOSDEM 2026 – Open-Source Conference in Brussels – Day#1 Recap
As a US citizen, when I see the phrase "European digital sovereignty," I'm a bit concerned that our OSS enthusiast and activist allies in that geography are learning to associate American OSS with American tech companies and US government. This could deepen the old free/libre vs open source divide that seems to have polarized along the separation by the Atlantic ocean. If so, in a time where Americans may be soon head-to-head with a runaway tyrannical government, our EU allies will be busy retreating into free/libre commensalist thinking that seem tunnel-visioned on using government funding to escape MS Word, something that is going to be the last thing on their minds if actual sovereignty concerns emerge.

The more general goal will remain to protect all individual freedoms from all tyrannical governments, not to depend on them. It will remain to use better information technology to enhance the functioning of all governments and to create healthy competition in all markets to protect consumer choice. American OSS has not forgotten this one bit. Our country is just having a moment, and it won't help if EU OSS participation writes us off as casualties while EU OSS focuses on "uniquely European" solutions.

positron26··on Swift is a more convenient Rust (2023)
Leptos etc. WASM is mature enough to just go for it. Same code sharing arrangement as using TS on the backend but using the backend language on the frontend instead.
positron26··on Swift is a more convenient Rust (2023)
Yeah. You can bind C in swift. You can expose C interfaces from Rust. Write non-iOS code in Rust, then bind. Now you have iOS support with common client code in Rust.

You can compile Rust to WASM. You can run Rust on the backend. It's gaining popularity on k8s for memory consumption. You can write music visualizers in Rust. You can pivot to killer drones or embedded in Rust. It has very few foot guns compared to C and all of the advantages of C.

The people who want Swift to succeed off of Apple hardware are basically waiting for someone else to do it. Rust users will just invade Apple from every side. It is very clear which platform is the better investment of anyone's time.

I can't believe I'm even wasting keystrokes thinking about this. I saw a bunch of upvotes and can only imagine this has to be wishful thinking from Swift users.

positron26··on Vibe coding kills open source
Any "exposure" economy has real money somewhere else turning the wheel. If that money isn't sensitive to healthy signal, neither is the downstream. Well-wishers are just alternative perpetual motion machine enthusiasts.
positron26··on Vibe coding kills open source
What even is "engagement" here? Seems like abstract harm that rationalizes whatever emotion the reader already feels.
positron26··on Vibe coding kills open source
Without real finance model innovation, what returns?
← PreviousPage 2 of 9Next →