cyber_kinetist

Nowadays C# has good AOT support, but it wasn't when Flutter was in its infancy.

Yeah good point. Although C# nowadays has good AOT support, that wasn't the case in the early days of Flutter (Google could have collaborated with Microsoft to develop this, but why would they?...)

There is Slint (https://slint.dev/), which was a company founded by ex-Qt devs and seem to be trying to make a better alternative to Qt. The core engine is built wiht Rust, but they also provide C++ and JS bindings and also have a QML-like scripting language you can use to design UI quickly. So far I think they've solidified their revenue model by catering to embedded devices, but I haven't seen a solid usecase for desktop apps (yet...)

> I very much reject a "desktop framwork". Qt has its own abstractions for everything from sockets to executing processes and loading images, and I don't want that.

Yes, that is the consensus of why Qt sucks - it's a massive framework that tries to do everything at the same time with a massive toolset of in-house libraries. This is inherently tied to the revenue model of the Qt Company - sell custom modules that work well with the Qt ecosystem at a high enterprise-level price. I also wish to just use the "good" parts of Qt but I can't, since it already has a massive QtCore as its dependency.

However, there is still no cross-platform framework except for Qt that can actually do the most important things that a desktop framework actually needs: an actual widget editor, styling and theming, internationalization, interop with native graphics APIs (though I have gripes with their RHI system), etc. That's why I'm rooting for PanGUI (https://www.pangui.io/) to succeed - it pretty much completes all the checkboxes you have, but it's still WIP and in closed alpha.

> I hoped someday Flutter might be mature enough for desktop development >> Anything that forces a specific language and/or runtime is dead in the water.

Yeah, but at that time I thought this was at least better than wrangling with Qt / QML. You can write the core application logic ("engine" code) in C++ and bind it with Dart. There are already some companies I've seen gone a similar route with C# / WPF.

They're making the initial version with C#, but they have plans to make the library language-agnostic. Rather than creating bindings, they'll write it in a subset of C# and then transpile it to C++/Jai/Zig/etc, so you can get the best language integration without the hassle of also wrangling with your build system.

Not bad for mobile apps, but still sucks a lot for desktop support.

Also, really wished they've opted for a more general language like C# rather than Dart - but that's inevitable since Google needed to make use of their Dart language after they've failed to standardize it on the Web (and I think they don't want to use a language developed by Microsoft of all companies)

Absolutely, they need Qt in order to design and theme a UI that actually doesn't look terrible (They already had good experience in porting Musescore from vanilla C++ Qt5 to QML widgets, so I think they'll use a similar system for Audacity)

The main consensus in the native space is that Qt is still miles ahead of any other cross-platform desktop framework (including WxWidgets). Doesn't mean that Qt is anywhere good - it's just the least worst option out of all.

I hoped someday Flutter might be mature enough for desktop development, but so far they've focused most of their efforts on mobile and I don't think this will change in the future.

I think the real conclusion is: someone has to make a native cross-platform desktop UI framework that doesn't suck. (Yeah Qt exists, but it really sucks...) Until then, everyone will default to just using the browser for a desktop app, and the beatings will continue.

Because of this, I'm really looking forward for PanGUI to step up (https://www.pangui.io/), their UI framework is very promising and I would start using it in a heartbeat when the beta actually releases!

This is probably the community he was talking about:

https://handmade.network/

Here's the manifesto: https://handmade.network/manifesto

Yeah, reminds me of the "Security" xkcd (https://xkcd.com/538/) - a threat from a good ol' 5-dollar wrench defeating state-of-the-art encryption.

Never estimate how state actors can use violence (or merely the threat of it) to force people to do things. The only way to respond to that is not through code or algorithms or protocols, but through political action (whether it be violent or non-violent)

No, 2026 is definitely not the year of home servers, because hardware has become too expensive.

Maybe viable if you have a bunch of spare parts laying around. But probably not when RAM and storage prices are off the charts!

That's true if you're using a CAD-like tool, but that's typically not used for art (more for engineering / mechanical design)

Game / VFX artists heavily use mesh-based tools such as Maya or Blender.

Hah, we had this since the beginning in CUDA! (Circa 2008)

It's baffling how little has been invested in shader language tooling - probably set back graphics programming in the game industry about a decade or so. I just wish there was a proper CUDA-like shading language with a simple API, like what Sebastian Aaltonen suggested in https://www.sebastianaaltonen.com/blog/no-graphics-api.

The problem is that Moore's law is dead, silicon isn't advancing as fast as what we've envisioned in the past, we're experiencing all sorts of quantum tunneling effects in order to cram as much microstructure as possible into silicon, and R&D for manufacturing these chips are climbing at a rapid rate. There's a limit to how we can fight against Physics, and unless we discover a totally new paradigm to alleviate this issues (ex. optical computing?) we're going to experience diminishing returns at the end of the sigmoid-like tech advancement cycle.

A nice tip: make sure to install the LTSC version of Windows 10, which is the most perfect setup you'll ever have (well, other than installing Linux)

No Cortana, no Copilot, no Windows Apps. Just pure unadulterated Windows, with extended support until 2032 (if you install the IOT version)

To add some context, gold was actually something backed by the US government during the Bretton Woods era (40s-70s), where 1 ounce of gold was pegged to 35 dollars. This was only possible because the US accumulated so much wealth relative to rest of the world after WWII, so they controlled the majority of the gold supply. After the golden age of Keynesian America ended with stagflation in the 70s, the US government had to stop all of their gold from fleeing the country, so this guarantee had to end. Which leads to the Nixon shock, where the dollar (and all other currencies as well) became free-floating, and we enter a brave new world where humanity hasn't lived before (neo-liberalism).

Given all that, it's easy to see why the value of gold has plummeted during the 70s - 00s. Though I could see two reasons as to why gold prices are rising during the last decade:

- Gold is actually just a part of the asset bubble (in the same group as housing, stocks, and crypto), and investment in it is aided by too much money printed by the US government not being used towards productive ends but towards rampant asset speculation.

- The current era of neo-liberalism is going to end pretty soon, and some goldbugs are rooting for the revival of late 19th-century classical capitalism, where gold was actually the international standard. I think this is very unlikely though, even if the US dollar loses its status with the end of the petrodollar system. My guess is we're going to deal with free-floating currencies for quite some time, especially when wars are going to happen and governments have to print more money to sustain their war efforts. (I think the best monetary system would be neither gold or crypto, but instead something like the Bancor (https://en.wikipedia.org/wiki/Bancor))

> Make things that pay off even after you’re done making

Which means you need to have some kind of ownership of the means of production (equity, in this world of capitalism), else you're just a wage slave.

Current datacenter GPUs are optimized for LLM compute, not for real-time rendering. The economics for running such beefy GPUs just for game streaming won't add up.

I think China will then try to sell their own PC parts instead, their semiconductor industry is catching up so who knows in a decade.

But perhaps then the US will probably reply with tariffs on the PC parts (or even ban them!) Which is slowly becoming the norm for US economic policy, and which won't reverse even after Trump.

Yup you've actually pointed out the most important difference: SDL3 is designed to be compatible with the APIs and devices of the past (2010s), whereas this proposal is designed to be compatible with the newer 2020s batch of consumer devices.

If you look at the details you can clearly see SDL3_GPU is wildly different from this proposal, such as:

- It's not exposing raw GPU addresses, SDL3_GPU has buffer objects instead. Also you're much more limited with how you use buffers in SDL3 (ex. no coherent buffers, you're forced to use a transfer buffer if you want to do a CPU -> GPU upload)

- in SDL3_GPU synchronization is done automatically, without the user specifying barriers (helped by a technique called cycling: https://moonside.games/posts/sdl-gpu-concepts-cycling/),

- More modern features such as mesh shading are not exposed in SDL3_GPU, and keeps the traditional rendering pipeline as the main way to draw stuff. Also, bindless is a first class citizen in Aaltonen's proposal (and the main reason for the simplification of the API), while SDL3_GPU doesn't support it at all and instead opts for a traditional descriptor binding system.

Over a couple of decades VCs have invested in vanity startups that cost billions of dollars like it's nothing, countless times.

I think half a billion isn't that expensive for a program that searches for a potential "theory of everything" that can profoundly change our understanding of the universe (even if it brings no results!)

We already have an abundance of natural resources (at least... for the developed countries) The problem is that they aren't evenly distributed enough, to the extent that a lower population cannot counter the increasing wealth inequality.

There's a lot of simple cloth sim examples on the internet, so I see why LLMs can code these kinds of demos easily.

Kinda interesting that they were able to resurrect the WebKit engine on Windows, might have been quite an engineering feat. In the past (2007-2012) Apple distributed a Windows version of Safari, but it's been more than a decade since it was deprecated and the engine only supported Apple's ecosystem.

If you really want to read how horrible and scary the idea of geo-engineering can get, read these two papers:

https://brill.com/view/journals/hima/30/4/article-p3_1.xml?l...

https://brill.com/view/journals/hima/31/1/article-p3_1.xml?l...

It's a wild read, that not only explains how this idea is technically horrible - he also analyzes the social and psychological effects this would have on humanity, and it's harrowing:

- We will be unable to see blue skies anymore, it will simply become white. This will be a traumatic event for humanity, and would have consequences beyond our understanding. Also, sunsets will be much more bloody-red, and we wouldn't be able to see the stars as clearly as before.

- The nature of aerosol deployments is that once we start this to combat global warming, corporations and governments will just rely on this instead of also reducing CO2 emissions. So as time passes, an exponentially larger amount of aerosols would be needed to block out the sun at every year, which makes the side-effects of geo-engineering grow immensely over time.

- Once aerosols hit a limit, then there lies the termination shock - once we "give" up on emitting aerosols due to it becoming exponentially expensive to maintain over time, the Earth's temperature will suddenly shoot up at unprecedented rates, quick rendering the planet inhabitable for humanity. His analogies to Freud's theory of repression is apt - we constantly repress our traumatic thoughts with stopgap measures, until the fantasy becomes too expensive to maintain and enter into the fully destructive phase.

Not all jobs are fun, but they can be bearable if meaningful enough (whether that being useful for other people, or even just provide a living wage to support your family)

If they really cared about performance (which I think is the case since they're using C++ for this) - they should use something else than nholmann-json. It's simple to use, but has one of the most atrocious performance for a C++ JSON library. (The safest choice would probably be RapidJSON)

As a side note, Intel's discrete GPUs are also famous for high-quality video transcoding - it was quite popular for streamers who needed a second helper PC only for OBS streaming.