Video

Exactly, there is a disconnect between implementation and understanding the hw it runs on, everything is abstracted to a high level, causing a lot of performance loss. Lots of programmers have no clue of assembly, you need to look at the generated assembly to understand how effective your code is using the underlying hardware. Todays programmers press compile and call it a day, with ever increasing hw performance you can get away with that only for so long. The abundance on new programming languages also has not helped, there is a loss of focus by all this choice, fundamentally these new languages have brought more abstraction and tolerate more clueless programmers.

Many abstraction layers in code are justified and they exists because lack of standardization.

@@gruntaxeman3740 Abstraction is very usefull it is part of designing good architectures, within these architecture it does not have todo with standardisation, abstraction makes sense from a design perspective. Of course you need abstraction layers for all kind of things outside your design as well, to address HW, etc etc, and these layers do add to less efficient code, nonetheless i would not argue against these kind of abstractions. The kind of abstractions i am against are the ones used in "modern" programming languages, to assist clueless programmers, the ones that don't know how to handle pointers, cause memory leakage, don't know what memory allocation entails and freeing up memory, they write their programs in languages that treat them like little children preventing the children from making mistakes abstracting away basic concepts required when you write in C or C++. The kind if programms, programmed by these programmers with highly abstracted and protected languages, generate very inefficient code, of course i can write inefficient code in C too, but generally C and C++ programmers know their language and the relation with the HW it runs on, this is usualy not the case for high level languages.

@@gruntaxeman3740 Yes, it's not that some, or even many, of the abstraction layers aren't justified but it does remove the programmer from the underlying hardware to an extend where I see people today teaching low-level programming languages but not understanding the difference or ranges of a signed vs an unsigned integer, let alone memory allocation/boundaries/alignment/stack/cache/heap, intrinsics, etc. The OS is an abstraction layer in itself, which is why we don't use one when we need things to run predictably or fast. My point is not that they are inherently bad. My point is that even if you are working in an environment on top of a lot of layers (JS in the browser), you should still know how the hardware works and how your code get's from your environment to the hardware. If you do not know this, you end up not initializing your arrays with a size even in situations where you know the size at "compile time" or you write loops that iterate over date in a way that can not be optimized even by the best compiler. Thereby leaving many X performance on the table even though it would not have been any more difficult to write the code much more optimized.

Except for the part where he is actually proposing and working on solutions... Not just complaining.

Yeah, he's thinking about it as a game dev, someone that doesn't need to care about IO as much as other devs.

It's much simpler than that. Most software engineers do not have an intuition about what's slow and what's fast, you can make things fast in any domain, but people just do not have this knowledge. I've worked on everything from websites, web applications to low-level analysis software and even some embedded software, in most cases people do not care, and when they do their focus in the wrong place. Like trying to optimize some algorithm when in reality it's your memory that's messed up and causing a huge slowdown because CPU is busy doing nothing for most of the time. Education on performance is very lacking, where it stops for most people is at: switch to binary search if your element count exceeds XYZ.

@@filipg4 Focus should be code maintainability. When software runs slowly, then something stupid is usually done. Clean, maintainable code makes easier to spot if something stupid is done and makes modifications simple.

He is very aware of this and talks about it specifically in other lectures. This is just one clip

When will he finally release Jai to the public 😢

I would argue that there was never a need for it, but after years of being exposed to slow stuff people just stopped to care. "Just buy w faster machine" is all that you need.

I'm loving what Jai has done being one of the best curly brace programming languages. Just wondering the specific things about its design make it so engineers only spend a week doing the same reflectors in 2 months using other similar languages?

The problem is abstractions are getting higher (now we just tell AK ai what to program) but machines are not really getting faster. "But more cores" you say. All cores have to write to the same memory at the end of the day. If two cores write to the same memory address at the same time there goes your program

You have no idea what you are talking about. Games are one of THE most complex software systems out there. Sure, they don’t do that much io waiting (actually they do too but it’s not the main computation there), but neither does Pro Tools when i open it. What do you think a DAW is doing for 60 seconds when it’s opening? Surely nothing that couldn’t be avoided with a codebase that has a couple million lines less boilerplate

What counts as "real software"?

@@drxyd probably enterprise java codebases and react

the vast majority of the kinds of programs he mentions in the talks (compilers, editors, photoshop etc) make few to no network calls

When your program only waits on network io it is structured wrong. Maybe instead of sending and waiting for requests individually you should batch them up into larger requests or just a single giant one, maybe network io is not even necessary -- not everything has to be a web application -- or maybe you run all the code on the client side where it does only need to load once and then never talk with a server far away. What do you mean "virtually all seful programs", MS Word has a slow Math editor, Photoshop and GIMP take multiple seconds to start, and web applications and sites are now multiple megabytes making them slower (the average website is now larger than DOOM). I could say that my program spends most of the time in memcpy, then it is not the problem that memcpy is slow, the problem is that the program creates useless copies everywhere. When you spend time waiting on network io, you are probably making too many requests and you are not doing them in parallel.

I've seen a number of projects where they have to simply shrug their shoulders and start working around poorly understood behaviors that seem to be emergent properties of the specific mix of proprietary black boxes which nobody is willing to own or investigate. Whether that's "bad" or not is left as an exercise for the reader.

description says "Reboot Develop 2017"

Because it's planned to be in the future (when it's actually fully featured and ready) alongside a full-feature product (Sokoban game) to prove its capability? Why are you even here?

it's not closed source at this point yet, either. it's in very restricted beta still as far as I know.

Not yet. Tbf its like a raw food in a newly opened restaurant. Not sure if it should be served yet, but we'll see.

@@9hoot789 true, though I can't wait to see Jon handle the toxic side of the open source community. I don't think he'll be able to handle it without doing a Torvalds

​@@joseduarte9823Is he going to update the language after 1.0? Or is it like zig where they plan to never update after 1.0

Do it!

You can't really do reflection in Rust, so you'd have to also metaprogram it in; like making macros that automatically allows getting some debug output from a struct. It's definitely possible but pretty complex (like it was to implement for Jai as well). The simplest solution would be just implementing the Debug trait which can be derived for a lot of types.

Rust is trash for gamedev. Please stop following hype trains. In game dev you HAVE to be flexible and fast prototyping, fast changing the code. Rust is too strict which lessens errors, but slows change speed to 0.

@RichardLofty 1) I do need fewer iterations. I think more before and after. 2) I can get faster CPU. 3) CPUs are getting much faster these days in the first place. 4) Rust front end will receive a major update soonish to reduce compile times further. 5) I fail to see why I can not use slower compile time to slow down myself for a break too. You may not burn out but I do tress out over time. 6) I am not going to dissmis an entire language because "oh my gosh the compiler is too slow". C# compiles faster. And yet there are engines written entirely in C# which is far worse at run time. Like Space Engineers. Rust does not fail at compile time at least. Whether it be safety or speed.

@@RichardLofty You've never done gamedev in Rust, have you?

@@techpriest4787 I don't think he was talking about compile times, but the time it takes the programmer to change the source code.

& MacOS has vendor lock-ins, its updates regularly break things & users only have little control what & how MacOS updates.

Not true and not even applicable either considering in that hypothetical, he has built a bridge (two bridges, in fact), and understands exactly how to build one (not just "two logs and some rope"), so he has every reason to be able to complain and point out the problems with the industry.

@@9hoot789 huh? The joke Ken Thompson (I think that’s KT..) is making is just a joke. Writing a barebones OS on a rudimentary machine is something that is a tractable problem for a single, highly skilled developer. Years later, Linus did something similar, but had to leverage GNU to get complete functionality. He has a semblance of a point in this, but it’s taken to such an extreme that it’s a functionally useless point. You can’t fix something if you don’t address the complicating issues.

​@@enginerdy The person you're responding to is pointing out that Jonathan Blow has built complex programs and is currently building his own programming language - he has a solid foundation to base his claim that software is worsening on because he knows what sorts of things you can do when you don't use the kind of abstractions people claim save time. You're right that the tasks people are completing these days are more impressive, but by what magnitude? For ten times the engineers, are we making software that is ten times as good? Blow argues that not only are we not making better software, software reliability is actual becoming worse, so we may be radically increasing the number of engineers required to make software of roughly equivalent quality. Is that worth it? You can watch Blow's full talk for a much more thorough explanation of all of this, if you haven't already.

@@MrLordFireDragon thanks for the further explanation. I do think his examples undermine his main points though, which I broadly agree with. The infrastructure to handle modern volume is large, and has many active components. We will eventually develop better tools to handle scale, but this comes in cycles: people expend vast effort to do a thing that’s very hard, people spend time studying the problem and making tools to make it easier, problem becomes more tractable. Even his laptop example is goofy, don’t know how old he is, but computers in the era he’s talking about crashed CONSTANTLY. “Five nines” was what you could expect from carrier telephone and internet data service, but that was about it outside of aerospace and medical (where it’s even higher).

Not true. China builds massive bridges and apartment buildings in weeks. Some even in days.

Some say 9 is so reliable it makes time but who knows how long ago that ended

It was also used by Blizzard for cinematics and games like WoW (most likely before that, especially around the Diablo 1 era), and Squaresoft for their PS1 era games, and beyond. It made for really cool advertising, especially off the heels of Toy Story a year before FF7 released, and likely contributed to the quality of the game being high, but it is a huge money sink and risk that paid off in pushing the medium forward to surpass movies and books...but at what cost? (FF7 was one of the most expensive games at the time, with a manpower of 100 people, some of whom are from the film and anime industry)

John is working on a new programming language and compiler for it, both specifically for games.

The point he is trying to make is not that we need it, but the whole concept of it just disapeared, we don't really strive for things like that anymore. It's always product first, I don't really buy big games on releases anymore, I don't install new software versions right away, because you know for sure - day0 is always bugged.

Windows updating without user concent a d crashing the entire OS is not a "hardware problem", its part of a rotten culture problem and your dismissal is also part of it

Linux can be upgraded without a reboot. I found that one out the hard way when I updated my system (as one does before installing new things), that update came with a kernel upgrade, then I tried starting my new thing (it was the docker daemon IIRC) and it failed. I did reboot it at that point, but I'm sure I could've reloaded the appropriate components of systemd at that point and gotten it to work, so my point stands. Sure, some hardware isn't up to the task of five 9s. But inevitably, that hardware was created in service to the same culture that produced modern software that is _also_ incapable of five 9s.

@@mage3690yes, you can upgrade anything on linux with no reboot, including the kernel. On personal computers, it’s better to not bother an just reboot, it’s easier. To have live kernel upgrades you must set it up so I opt for just rebooting

"That’s a hardware issue, not software". sure, for five 9s that might be true. but the vast majority of problems I've had with machines I've worked with have been software problems, not hardware problems. most reliability struggles most of us experience are software problems. even striving for three 9s and achieving two would be a big improvement.

The first two minutes make it clear what context 9 9's is made in.

Presumably user uptime, given the talk of productivity. A bug on one frame could take arbitrary time for the user to work around, so pretty much all software falls far from five 9s with that definition.

Refers to the percentage of running vs down, not the chance to break. You'll get more than one frame of downtime after the system breaks.

In other words it would have to reboot and restore adobe in that one frame for that to keep 99,999% runtime

@@sabamacx It's not. He talks about losing the progress on the programs. How does it correlate to the uptime? You might lose your progress in a one-frame bug, or if system restarts for a minute or two. The annoying part is that you lost your data.

but he is 100% right and i a new. I love using Rust because i can low and hight level stuff.

Gratuitous incompatibility for user lock-in. If it's something new, make it complex enough that any attempt to clone the API will have at least a few bugs that prevent compatibility (e.g., every Microsoft, Apple, or Google interface). If it's something that a smaller competitor already did, add something similar to your own stuff (so it won't fall behind), but make it different enough that something written for your ecosystem won't be portable to their systems (e.g., when Plan 9 introduced the "rfork" system call, all the BSDs copied it compatibly, then Linux adopted it but tweaked it a little and changed the name to "clone"). Standard libraries help a little, but the obvious countermeasure is to add new features faster than the competitors can keep up.

Also, the need for higher performances at the price of incompatibly. Everyone got an opinion on what is great performance depending on who you ask.