churnedgodotdev

il y a 4 ansdiscuss

You're right this person can't be taken too seriously, but the linked website for hiring cheap Philippines workers appears to be legit.

churnedgodotdev

il y a 4 ansdiscuss

I read Donald Knuth does this. He uses a USB flash drive to transfer data to and from his main book-writing Linux workstation. Air gapping has the added bonus that the constant distraction and breakage risk of never ending app updates and security patches is mitigated. As Knuth is someone from the computing era before even floppies and network cards were a thing, when fast data transfer between sites meant mailing a deck of punch cards or reels of tape, it probably doesn't feel that slow to him. He also essentially quit email on January 1, 1990, after using it since 1975, because it was such a distraction from getting stuff done.

Also, George R. R. Martin uses WordStar 4.0 on a DOS machine:

  https://www.cnet.com/culture/george-r-r-martin-writes-with-a-dos-word-processor/

churnedgodotdev

il y a 4 ansdiscuss

Elon 1st Amendment comment:

  Twitter acting by itself to suppress free speech is not
  a 1st amendment violation, but acting under orders from
  the government to suppress free speech, with no judicial
  review, is

https://twitter.com/elonmusk/status/1598853708443357185

churnedgodotdev

il y a 4 ansdiscuss

> Basically stripped-down versions of Blender itself

Okay, that makes sense. The video on the linked Blender Apps page shows this really simple app where one drags models around and then drags materials onto the models to give the model that material. So something simple like that would work better in Godot. But Blender's 3D modeler is way more powerful than Godot's but has a very steep learning curve so distributing a simpler version of Blender's modeler could definitely make sense.

churnedgodotdev

il y a 4 ansdiscuss

Seems like Godot already does what Blender Apps does much better:

- Godot is great at importing Blender models/scenes.

- Blender is GPL, Godot is MIT.

- Godot produces much lighter .exe files. Blender Apps are at least 200MB.

- Godot has web export, too.

Godot still has stability issues that can affect larger projects (presently over 7000 open issues on Github[1]) but for smaller demos and art things Godot usually is pretty solid.

[1] https://github.com/godotengine/godot/issues

churnedgodotdev

il y a 4 ansdiscuss

> "gen z" students expect the "system" to adapt to address their struggles

I hope Gen Z soldiers don't have this attitude.

[Mortar round blows off leg of Gen Z platoon grunt.]

"OMG! Mortar rounds are SO UNFAIR!" Let's sign a change.org petition demanding mortar rounds only emit harmless pink smoke. If we end up with pink smoke in our foxhole we can demand more time to dig a deeper, better concealed foxhole.

churnedgodotdev

il y a 4 ansdiscuss

Of course, that's what I thought I said in my previous comment but I guess I wasn't crystal clear. The formal verification stack basically tells you, "There's no guarantee 0<N<100 holds if you get N from an I/O stream. N can be anything." So it's your job to insert a runtime check for N<=0 or N>=100 and decide how best to handle that, e.g. panic and quit, or clamp N to a valid range if it makes sense, or avoid accessing the array if it makes sense. But the system won't let you use a possibly out of range N as an array subscript so you are forced to code a solution to this special case "at compile time", typically by inserting a runtime check. Though if you are reading data from a trusted ROM you can also use a pragma or special assert to say, "This data can be trusted to have this invariant" and avoid the runtime check.

churnedgodotdev

il y a 4 ansdiscuss

ASML is the closest thing to the planet Dune in our world. Chips made with their machines extend life. Chips made with their machines expand consciousness. And he who controls the chips, controls the universe. The information universe, at least. I just wish for a messiah to arrive who leads the jihad against FAANG. The Kwisatz-RISC-V-Haderach. The Voice from the Open Source World. Fortunately, he need not fell Mark Zuckerberg in single combat with crysknife in order to lead our people to true freedom.

churnedgodotdev

il y a 4 ansdiscuss

Of course. The checks are performed where the values arrive. If a file or network socket provides a number N that is going to be used as an array index, then N must be proved to be within array bounds at compile time before being used as an array index, which means you are forced to catch the bug at compile time and decide how to signal bad values of N.

churnedgodotdev

il y a 4 ansdiscuss

Ada solved these problems in 1983.

More recently, you can use Frama-C to constrain allowable sequences of 0's and 1's for C types and formally verify correctness.

In Ada since 1983 you can, e.g, declare your own 8 bit signed symmetric type without the wart -128 like so:

  type Sym_8 is new Integer range -127 .. 127;

Then this fails at compile time:

  My_Signed_Byte : Sym_8 := -128;

SPARK can prove all execution paths through your program are free of such constraint violations. This means safe SPARK code can disable runtime checks and run faster than the safest Rust/Zig dev settings, which insert runtime checks for over/under flow.

In Frama-C, say you want a function that returns an absolute value. This function will fail to verify:

  /*@ ensures (x >= 0 ==> \result == x) &&
              (x < 0 ==> \result == -x);
      assigns \nothing;  */
  int abs (int x) {
      if (x >=0)
          return x;
      return -x;
  }

It fails to verify because you might have x==INT_MIN. So this will verify:

  #include <limits.h>
  /*@ requires x > INT_MIN;
      ensures (x >= 0 ==> \result == x) &&
              (x < 0 ==> \result == -x);
      assigns \nothing;  */
  int abs (int x) {
      if (x >=0)
          return x;
      return -x;
  }

churnedgodotdev

il y a 4 ansdiscuss

> can we just make it render on time?

Oh, time to go down a low-latency rabbit hole. Start here:

https://raphlinus.github.io/ui/graphics/2020/09/13/composito...

Some key mentions: Windows has a hardware overlay API. A hardware overlay is why the OS mouse cursor renders with low latency even with 'vsync on'.

For fullscreen, the ultimate in tear-free low latency is beam racing. Search for 'blur busters beam racing' to go down that rabbit hole.

churnedgodotdev

il y a 4 ansdiscuss

The cursor lag is really bad, particularly on Windows, but there are at least 2 ways the lag could be much lower:

1. Use 'desynchronized', e.g. canvas.getContext("2d", { desynchronized: true });

2. Instead of using requestAnimationFrame(), draw as soon as a move event is received.

Here's a paint app that uses techniques 1 and 2 to achieve much lower input lag drawing the pen strokes that chase the cursor:

https://paint.js.org/

Two other tricks to be aware of:

3. [Chromium only.] Use 'pointerrawmove' to get move events as soon as they happen instead of waiting for a big bunch of coalesced moves to all arrive after a delay.

4. Drawing where the cursor is predicted to be in order to compensate for lag works well during smooth cursor/pen motion. For example, the Windows Snip & Sketch tool actually renders the line you are drawing in front of the direction your pen tip is detected traveling because when using, say, a Microsoft Surface Pen on a physical Surface Pro screen the line being drawn would otherwise noticeably lag behind the physical pen tip (especially on a 60Hz screen) unless this predictive trick is used.

Finally, be aware of open Chromium issue #992954. When the devtools are open [F12], coalesced events aren't used so you get lower input lag but more events to pump (like 'pointerrawmove'). This bug is maddening because it depends on whether or not you have the debugger open!

https://bugs.chromium.org/p/chromium/issues/detail?id=992954

churnedgodotdev

il y a 4 ansdiscuss

Actually, a reason (of many) I switched from Apple to Windows was Apple's well-known audio drop-out fiasco caused by Apple's T2 security chip essentially hijacking data lanes when low-latency audio processing needed them. Here is an image of an audio capture of the T2 causing a hiccup:

https://tidbits.com/uploads/2019/04/T2-hiccup.jpg

This image was linked from this article:

https://tidbits.com/2019/04/05/what-does-the-t2-chip-mean-fo...

churnedgodotdev

il y a 4 ansdiscuss

Since many HN'ers live in the greater Seattle area, it's important to be aware of the Tacoma Smelter Plume:

https://ecology.wa.gov/Spills-Cleanup/Contamination-cleanup/...

For almost 100 years, the Asarco Company operated a copper smelter in Tacoma. Air pollution from the smelter settled on the surface soil of more than 1,000 square miles of the Puget Sound basin. Arsenic, lead, and other heavy metals are still in the soil as a result of this pollution.

churnedgodotdev

il y a 4 ansdiscuss

I'd fund it if I could use it to actually ship something and make money. I've donated to other software projects. If, say, for several thousand bucks they'd actually give me a tiny bit of support by fixing bugs in the "stable" branch that affect me then I'd pay up. At this point though if I want to actually ship using open source I find I can actually get things done faster using SDL2. Because SDL2 is a minimal layer over native if there is an issue I can actually figure out what native calls are happening in what order. If controller inputs are wonky or textures are the wrong color I can figure out why. Godot does try to use an understandability friendly subset of C++ (no template metaprogramming and such) but it is still too sophisticated for me to quickly dive in and understand how it all works; and it is certainly too complicated for its core maintainers to be able to quickly dive in and fix longstanding open issues in "stable".

churnedgodotdev

il y a 4 ansdiscuss

After quickly falling in love with Godot and several months of serious use I'm having to give up on it due to Godot's lack of funding to fix its more than 6000 open issues on Github:

  https://github.com/godotengine/godot/issues

Every single time a Godot bug bites me I find that a bunch of other devs reported the issue more than a year ago but no one has had time/money to fix the issue. For example, I was just bitten by #45628 opened Feb 1, 2021:

  https://github.com/godotengine/godot/issues/45628

When implementing controller support for a top-down Zelda-like game, it's normal to allow either the d-pad or left analog stick to control direction. But if you enable both at the same time in Godot then d-pad inputs cause your character to "freeze".

Eventually Godot will stabilize and be to game engines what Blender currently is to 3D modelers. If I had time to learn the Godot C++ code I would volunteer to help but I want to actually make video games instead of working on an engine.

churnedgodotdev

Submissions

We'll weather recession because Philippines workers only $10k/year

comments