3 ms·
> It doesn’t work like that. If an image processing algorithm takes 2 instructions per pixel, adding a check to every access could 3-4x the cost. Your understa
by Aurornis 2y ago
> It doesn’t work like that. If an image processing algorithm takes 2 instructions per pixel, adding a check to every access could 3-4x the cost.
Your understanding of how bounds checking works in modern languages and compilers is not up to date. You're not going to find a situation where bounds checking causes an algorithm to take 3-4X longer.
A lot of people are surprised when the bounds checking in Rust is basically negligible, maybe 5% at most. In many cases if you use iterators you might not see a hit at all.
Then again, if you have an image processing algorithm that is literally reading every single pixel one-by-one to perform a 2-instruction operation and calculating bounds check on every access in the year 2025, you're doing a lot of things very wrong.
> This is why if you dictate bounds checking then the language becomes uncompetitive for certain tasks.
Do you have any examples at all? Or is this just speculation?
- monkeyelite 2y ago> Your understanding of how bounds checking works in modern languages and compilers is not up to date. One I am familiar with is Swift - which does exactly this because it’s a library feature of Array. Which languages will always be able to determine through function calls, indirect addressing, etc whether it needs to bounds check or not? And how will I know if it succeeded or whether something silently failed? > if you have an image processing algorithm that is literally reading every single pixel one-by-one to perform a 2-instruction operation and calculating bounds check on every access in the year 2025, you're doing a lot of things very wrong I agree. And note this is an example of a scenario you can encounter in other forms. > Do you have any examples at all? Or is this just speculation? Yes. Java and python are not competitive for graphics and audio processing.
- MonkeyClub 2y ago> Java and python Java and Python are not on the same body of water, let alone the same boat. You can see some comparisons here: https://benchmarksgame-team.pages.debian.net/benchmarksgame/fastest/python.html https://benchmarksgame-team.pages.debian.net/benchmarksgame/...
- monkeyelite 2y agoYes. And why is that?
- Aurornis 2y agoInterpreted languages and garbage collection. Bounds checking barely registers as a factor.
- monkeyelite 2y agoYou can write Java that uses arrays that are allocated once - will it have loops as fast as C? Why not? And furthermore I don’t suspect you’re proposing we should be using C + bounds checking (that’s already a Gcc flag?). But rather bounds checking is one of many safety features. The whole pitch of Java is exactly what OP said - let’s just pay 10-30% cost to get these nice portability and memory benefits over C++, and it didn’t work that way especially as memory speeds have diverged from CPU.
- Aurornis 2y ago> Yes. Java and python are not competitive for graphics and audio processing. Because Python is an interpreted language and Java has garbage collection, not because they have bounds checking. Using a language like Rust the bounds checking overhead compared to C code is usually minimal, like 0-5%. The compiler can even optimize away many bounds checks (safely) if you structure your code properly, such as with iterators. Here's an example article : https://shnatsel.medium.com/how-to-avoid-bounds-checks-in-rust-without-unsafe-f65e618b4c1e https://shnatsel.medium.com/how-to-avoid-bounds-checks-in-ru... Make sure you read the note about how the author was mistaken about the 15% speedup from bounds checking. The actual bounds check change was much less, like a couple percent. You can even structure your code so that the bounds check disappears altogether when the compiler can confirm safety at compile time.