I think this doesn't get talked about enough. If your input is a big run of data that is being checked/transformed in one shot, it works well. But, if you're likely to have to make a decision on several bytes of the input, SIMD will be the same or slower than the scalar method. It's not a magic "go fast" button.
I think the big miss is that the former is achievable far more often than people imagine, which leads to settling for the later, aka pessimization. Reducing your allocations from millions per run to handfuls per run during initialization is effectively a "go fast" button, and is generally more reliable. It is very common for teams to spend big effort getting a 2-3x speedup on allocation-heavy code by disabling branches when a 100-1000x speedup can be had if restructuring allocations is a strategy under consideration (even before the extra 4-8x you might see if you go all the way to hand-tuned SIMD).
That is not necessarily true. simdjson exists. But it's far from simple.
I have been told SIMD is good for data-parallel loops, like the GPU is, and that is true, but it can also be used piecemeal, unlike the GPU. Because it is just the CPU, you can read 32 unaligned bytes, scan for the index of the first space, and take a branch based on that.
I am speaking exactly of simdjson. If you look closely, you'll see that the high performance is really only achieved by skipping over large chunks of the input data. Which is great, if your use case is operating on a subset of data. But, if the application really needs to process every piece of the entire input, there's no performance win.
Think about it, JSON structure is almost entirely made up of single bytes ({ } [ ] , : "). If you're not skipping over stuff, you're not going to get away from having to examine those bytes and branch on them. For something like `{"a":1,"b":2}`, the SIMD scan has a bunch of overhead to find the interesting parts and then you go back and practically have to reprocess every single byte.
You are leaving out other trade-offs, like the program crashing when something memory-unsafe is done. It is not simply a performance trade-off. Many folks would rather have the language catch this at compile-time instead of hoping that the test suite exercises all the code in just the right way to tickle the bug.
> Another key pro to capabilities is that you can actually have a memory safe interface without having to leak lifetime and other info across boundaries.
One of the benefits of having lifetimes in the type system is thread-safety. Again, there is a lot of focus here on just memory-management: use-after-free, double-free, etc... But, being confident in the code running in a threaded environment is important as well.
Looks like Mr. Kelly agrees with you. From the zen of zig:
- Runtime crashes are better than bugs.
- Compile errors are better than runtime crashes.
- Incremental improvements.
- Avoid local maximums.
Considering every memory safe language that I’m aware of panics on index out of bounds, there’s spectrum here.
Personally, I’d like to see refinement types that allow people to elide the runtime overhead/panics. Though that may be too much complexity to stomach.
> Looks like Mr. Kelly agrees with you. From the zen of zig:
... what? No one is going to argue against those vague statements. He obviously does not agree with people that put more weight on making things compile errors instead of runtime errors. Which is fine, but saying he is in agreement is nonsense.
> Considering every memory safe language that I’m aware of panics on index out of bounds, there’s spectrum here.
Many of those languages go to great lengths to avoid raw indexing: rich iterators, arenas with branded indexes, checked gets, and so on... But, sure, if you ignore all that and do `get(i).unwrap()`. Then, yes, things are exactly the "same".
I have no problem with people choosing the set of tradeoffs that works for their context. I have a problem with people acting like those tradeoffs don't exist.
> No one is going to argue against those vague statements. He obviously does not agree with people that put more weight on making things compile errors instead of runtime errors
> With the first and only commit 2 hours ago, the author of this project didn't let it "bake". They didn't exercise it locally to see what issues it might have, and I have a hard time believing the very first iteration of this software is perfect.
You have no idea whatsoever how many iterations were done before the initial commit. The VCS log is not representative of anything that happened before the first public release. Even before LLMs, people would grind away on stuff until they were happy and then put it in a VCS for public consumption.
Rust is one of the safer languages, but saying that it is "the safest language" is just a baseless exaggeration.
Decades before Rust and long before the simplified language that was C, there were safe programming languages, where all invalid operations, numeric overflows or out-of-bounds accesses generated exceptions and where use-after-free was impossible, because either garbage collectors or reference counts were used.
Rust is much safer than C compiled with its bad default compilation options, but it did not bring much in comparison with other languages.
Even in C++, with appropriate rules, restrictions and discipline you can write programs that are guaranteed to be at least as safe as any Rust program, but unfortunately very few use C++ in this way, i.e. by strictly avoiding the features that are obsolete or unsafe.
The biggest innovation of Rust is bringing some of the good ideas from functional programming to low level programming. I'd also say that partially exposing data flow analysis to a proframmer is new.
Rust package management is quite good, and also not by any means an invention.
I am still not a fan of all the ugly macro programming systems and verbose syntax in the language.
The borrow checker in Rust is frankly novel. Cyclone had something somewhat similar, but not the same.
The broader ML-like type system in Rust is not novel, but the integration of the borrow checker -- and its move semantics more broadly -- with it... that form honestly is an innovation. And one I'd have a hard time living without at this point.
> Even in C++, with appropriate rules, restrictions and discipline you can write programs that are guaranteed to be at least as safe as any Rust program
If by discipline, you mean running something akin to the borrow checker in your head, that's essentially tautologically true. The issue with that is that it's mentally draining and/or you will still make mistakes sometimes.
No, I mean by using only custom types for things like arrays, strings and pointers, which do access checks and automatic memory release, and not using unsafe features like the built-in arrays, strings and pointers, or the incorrect integer type conversions inherited from C.
For maximum safety, beyond what Rust offers by default, in C++ it is easy to replace the built-in integer types with custom integer types, which check for overflows and allow only the correct type conversions. It is also easy to define distinct types for various kinds of physical quantities, for increased safety.
You do not need to run anything in your head. With appropriate type definitions, a C++ compiler will do anything that is required.
The problem is that because of the requirement for backwards compatibility, C++ is a huge junk collection. I think that more than half of C++ consists of obsolete features, which should never be used in new programs, and this is a serious difficulty for newbies. There are various C++ style guides, but in my opinion even most of those are not very inspired.
Despite of its defects, C++ still has the advantage of extreme customizability. It is easy to write programs that appear to be written in a language that has no resemblance with C++ (inclusively by having different keywords and what appears to be a different syntax), but nonetheless they are valid C++ programs.
Such a customized C++ variant can mimic any safer language.
No. C++ doesn't have a garbage collector, and it doesn't have a borrowck, as a result any reference types are subject to ordinary human error because it can't garbage collect the target objects and it doesn't know when to destroy them otherwise.
The work to try to address this for C++ 29, half-finished and untried as it is - is extremely restrictive, you'd likely hate it, and that's just to solve this, the relatively easy problem.
Thing is, Rust wasn't content just to solve that easy problem. (Safe) Rust also doesn't have data races. The C++ standard doesn't say very much about data races, can't help you ensure they don't happen - it just explains that if they do that's Undefined Behaviour, game over.
The rules can be enforced by a static code checker.
That is really not very different of rules enforced by the Rust compiler.
For someone who does a fresh start, using a Rust compiler may ensure safer programs out of the box, but that does not mean that the same results cannot be achieved by alternative means when using other languages, when the use of those languages makes sense for other reasons, and it is worthwhile to invest resources in making appropriate libraries and tooling.
In general, I recommend against the use of C++ in new projects, but I see much too often claims about things that are supposedly difficult or impossible to do in C++, which are just false.
> I suspect that Rust will start taking over as a dominant LLM output language.
I doubt it. I think most people will become more entrenched in their favored ecosystem.
> I also suspect that in short order we'll have entirely new languages that are engineered to be ideal languages for LLMs to generate.
This is already happening. A couple months ago I came across this language that is engineered for AI and human consumption https://www.moonbitlang.com/
I tried that, but the Rust build process was too painful, and agents seemed to burn a lot of tokens guessing how to get the code to compile. I rewrote my project in Elixir and it’s been going much more smoothly
Elixir is great, and I have recently started using it myself, but its not a substitute for Rust. Try writing device driver in Elixir, or anything CPU intensive.
> not limited to the very few architectures that rust supports
Of all the complaints about rust, this strikes me as one of weirdest. How much code do you actually write for architectures outside the Tier 3 support list?
Ok, but recently? I too wrote code for obscure platforms once upon a time, but not in, say, the last 15 years.
Now that PCs, game consoles, and mobile devices are basically all either amd64 or ARM, there's just not such a long tail of weird platforms to develop for.
(the embedded world I will grant you, still lots of bespoke toolchains running around in that space)
Thanks, I didn’t know what logs it opened, or how to open others. It had menus and drop downs but I didn’t understand what they were listing.
Need to read the manual I guess, not a big deal but it should be obvious for a log viewer. Why I recommended CUA, though I understand it is not so common on Unix.
> but there is no obvious way to exit. I tried Q,q
It's not very responsive during initial indexing, which is something I need to improve. Pressing `q` should work to exit in general, though. Pressing CTRL-C three times in quick succession will force quit it.
It would help to know which version you tried. Things have gotten better over the years.
> I tried `man lnav` in separate terminal - but no man page is provided.
A man page exists, but only contains basic information. The builtin help text is much more extensive and can be viewed by running:
> `ps` shows 3 processes which would not die with SIGTERM, have to `kill -9`.
Older versions of lnav would use readline for the prompt and had to run it in a separate process because of "reasons". More recent versions have a custom prompt and don't require the extra processes.
re: man page - It looks like there is no support for man pages from the snap infrastructure. So, there's not much I can do.
The "stable" version of the snap is really old (circa 2023) at this point because I have been shy about bumping it. The candidate and edge versions are more recent and should be more usable.
Yep, I would say the stiffest competition for lnav is the old tools[1]. I would just hope folks could have an open mind and give "new" things a chance (although lnav has been on github for 17 years).
Hmm, I think you might be mixing things up, goaccess is written in C and probably not written as an alternative to lnav.