porgamrer

in 100 years all smartphones will look exactly the same as now, except they’ll all have the hapsburg jaw

In the many years of commuting to an open plan office I think 4 hours was probably the max, and usually it was less.

Since working from home, I’m not sure exactly but it’s a lot more, and honestly it’s more than is healthy. I’m still less tired overall though.

Honestly this is usually bad advice nowadays, for a bunch of reasons:

1. Modern allocators do the same thing as object pooling internally, usually faster. They rarely interact with the OS.

2. A GC will do things like zero old memory on another thread, unlike a custom clearing function in a scripting language.

3. Object pooling breaks the generational hypothesis that most modern garbage collectors are designed around; it can actually make performance much worse. Most GCs love short-lived objects.

4. Object pools increase code complexity and are very error prone; when you add a new field you have to remember to clear it in the pool.

5. If you are in a non-GC language you probably want something “data-oriented” like a slotmap, not a pool of object allocations.

Having said all that, it still all depends on the language/VM you’re targeting. The guy in the video clearly benchmarked his use case.

If I went back to the early days of C, it would be because John Connor sent me there in a time machine to destroy the first compiler before it could become self-hosting

lmao. as long as the people selling the stolen art pinky promise that the art theft machine didn’t steal any art, everything is above board

Game engines had a lot of growth speculation for the past decade. There were a lot of harebrained ideas about how game engine tech could disrupt loads of existing industries and provide the foundations for various new ones. e.g.

1. VFX studio offline rendering going to be replaced with modern game engine rendering any day now!
2. AR is about to take off and revolutionise every industry at any moment, if only someone can render the overlays!
3. The VR metaverse is here, and millennials love renting so much they are going to rent virtual flats and use unity to look at them!
4. The military will be desperate to spend their infinite budget on using unity to simulate warzones or something!
5. Wow Roblox found an amazing loophole for monetising child labour using a game engine. Let’s steal their idea and scale it up!

And so on.

For every idiot idea there is some large R&D team full of poorly-managed developers desperately trying to apply unity’s completely unsuitable technology to a problem it can’t solve, on the off chance that one of them turns into a money printer. There’s also probably a bunch of marketing people, sales people and suits trying to get past regulatory barriers, etc.

Whenever reality hits on one of these hype bubbles, a lot of people get fired. It just happened to VFX, for example.

I don’t think TIOBE rankings mean much, but C# is genuinely doing quite well.

In the games industry it has become the only language that is commonly used both for entry-level gameplay scripting and for implementing game engines.

It still has a lot of overhead compared to languages like C++ and Rust, but has proven to be a workable alternative for games that aren’t straining against the limits of modern hardware.

Aside from the anti-labour implications of this stuff, even as a consumer it is horrifying to imagine a world where we are constantly questioning whether or not some media was generated by AI. There’s something violating about it.

I don’t care if a machine learning system is being used in some very self-contained way, like to do basic voice-to-speech in a game with a lot of procedurally-generated text. That can even be an important accessibility feature.

However, as soon as it’s being passed off as a real human performance with emotional depth and authorial intent, it’s gross and I don’t want it in my brain.

This actually could have been much worse. I heard they are trialing a system where they lay off one employee every time unity is installed, but luckily there have only been 1800 installs since September.

I use a Tokonami KX450, which is not the newest but it’s the most widely available military-grade model that the average silicon shop is able to customise.

With that in mind you’ll want a uranium microreactor to really get that turbo button cranking out the keycodes (the french stuff is cheapest but ukrainian kit is worth the extra), as well as a mercury cooling solution and ideally a set of maglev keys for all the most common letters (NOT backspace; frankly you should remove that key entirely to avoid habits that damage your WPM).

Assuming you’ve got a solid pair of high-torque power gloves that should get you up to at least 20000 WPM, which admittedly won’t cut it if you’re trying to keep all the NPM dependencies up to date in a modern bank’s transaction processing software, but it’s probably enough if you’re just doing a bit of data analysis in python.