Show HN: Algorithm Cookbook in Rust

https://github.com/EbTech/rust-algorithms

by EbTech • 9 years ago

320 71 9 years ago

rustacean• 9 years ago

Thank you. But it seems not finished? I am trying to implement some basic data structures and algorithms in Rust, but the compiler nearly kills me. Is there any data structure & algorithms implementation tutorials for Rust newbies, like Learning_ Rust_With_Entirely_Too_Many_Linked_Lists[1]?

[1] http://cglab.ca/~abeinges/blah/too-many-lists/book/README.ht...

kazagistar• 9 years ago

Data structures and algorithms are possibly the hardest part of Rust to dive into unless you have a really good reason. The best advice for newbies is to learn with other stuff and use existing algorithms. Of course, if you are trying to implement stuff for a class or something similar then you might be out of luck, and might consider trying the Rust IRC and such to work through the errors as they come up.

freeordead5• 9 years ago

So what is the point of a language in which things are hard to implement?

EbTech4• 9 years ago

I made this cookbook specifically to show it doesn't have to be so hard. While designing an industrial-strength data structure library is an advanced skill, there are easier ways to implement the core ideas that should suffice for contest and coursework purposes. Very few people have to build actual libraries.

lgas• 9 years ago

Data structures and algorithms are possibly the hardest part of Rust to dive into

What else is there?

rectang3• 9 years ago

MIT license is good, but I always wish that sample code was available under CC0 -- and presented with explicit permission to copy and paste without preservation of notices. (A request to link back optionally when practical is fine.)

That makes it easier for sample code to achieve the ends of maximum code reuse and popularization of the underlying technology.

jay-anderson2• 9 years ago

Agreed. I once tried to submit a small snippet I found on stackoverflow to an open source project. Stackoverflow user content is licensed under creative commons, cc-by-sa (see https://stackoverflow.blog/2009/06/25/attribution-required/). I included attribution, but it was rejected. I totally understand why. CC0 is much preferred in this case.

peternicky• 9 years ago

Why was it rejected?

rectang3• 9 years ago

Your anecdote also emphasizes the requirement for open source projects to maintain rigorous intellectual property policies. Good on both you and the project for the appropriate handling!

Clean IP history is especially important for "universal donor" sample code intended to be copy/pasted. If there turns out to be an IP violation in library code, with luck it may be possible to excise it from the library -- and hopefully not too many downstream will have forked it. But when code is copy/pasted, the linkage is severed and the damage becomes increasingly difficult to repair.

For those reasons, I'd prefer to see something like a contributor list of public identities for collections of sample code, so that it is transparent who is making the licensing promises.

guelo• 9 years ago

FYI, newer Stack Overflow code is licensed under an MIT license with a link in code comments serving as attribution. See https://meta.stackexchange.com/questions/272956/a-new-code-l...

bogomipz2• 9 years ago

This is great. I am curious if anyone else knows of a similar cookbook for Golang.

peferron• 9 years ago

I have implemented quite a few algorithms and data structures for fun here [1]. It's not authoritarive by any means, and not everything is in Go (there's also some Rust, Swift, TS and JS), but maybe a few things could turn up useful.

[1] https://github.com/peferron/algo

bogomipz2• 9 years ago

Awesome. Thanks for the links.

urethrafranklin• 9 years ago

https://github.com/arnauddri/algorithms https://github.com/0xAX/go-algorithms

zeotroph• 9 years ago

A graph with the members first, next, and endpoints is not exactly self explanatory, something better than a "A compact graph representation" comment would be nice.

Especially since the ownership model of rust makes the classic graph representation of edges owning nodes impossible.

_0w8t• 9 years ago

It is only impossible using safe Rust. Unsafe code blocks should allow to represent this behind safe API.

sjroot• 9 years ago

Thanks for sharing this. I did a deep dive into Rust when I was tasked with implementing an interpreter[0] for school, and found the language to be very enjoyable to work in. While I have tried to convince other friends to try it, they still object citing a steep learning curve and the thought that it will never replace C++ or Java in mainstream development ecosystems. It is resources like these that will help lower that barrier to productivity. :)

[0] https://github.com/sejr/core-interpreter

gerty2• 9 years ago

To add to the list, here's my rust-noob implementation of the Kalman filter.

https://github.com/rbagd/rust-linearkalman

daniel_rh• 9 years ago

GPL3 means it's not very reusable in any projects that aren't GPL3. Have you considered a BSD or MIT license? or CC0 if you want to get code reuse? Or at the very least LGPL2

cyphar5• 9 years ago

Maybe that's what they want. Personally I use GPLv3 for any project I can, because my priority is user freedom and not code reuse.

gerty2• 9 years ago

Thanks, that's a fair point. I do value user freedom and also come from the R language where GPL rules among packages. It's true though that Rust crate culture leans much more to MIT.

I think GPL3/MIT as a dual license should be a decent compromise.

cyber1• 9 years ago

"My other goal is to show developers that C++ and Java kinda suck" -- really?!

sidlls2• 9 years ago

Yeah, it's a pretty poor way to represent Rust to an audience that is ostensibly it's target audience.

Buttons840• 9 years ago

Any developer who can't name something that "kinda sucks" about their language probably isn't ready to learn a new language.

EbTech4• 9 years ago

Fair point, it's fixed now.

tormeh2• 9 years ago

It's a worthy cause. At HN we might all be aware that Java and C++ suck, but others might not know. Assuming others know what you do is a common cognitive bias. To us, this might be a slightly childish thing to have as a goal, but it might be a revelation to some readers.

dsacco2• 9 years ago

At Hacker News we might all be aware that Java and C++ suck, but others might now know. Assuming others know what you know is a common cognitive bias.

The irony here is astonishing.

I am a frequent commenter on HN and I rather like C++. I acknowledge its shortcomings without resorting to the tribalism that is marring the reputation of the Rust community.

Please don't assume that all of Hacker News has a homogenous perspective on the utility (or lack thereof) of extremely common languages like C++ or Java. I've said this before in threads like this and I'll say it again: one language does not need to win.

partycoder• 9 years ago

It's not achieving it though. The code could be more idiomatic and the project could have a better file organization.

EbTech4• 9 years ago

Suggestions are welcome; I'm still learning Rust.

To put context around my provocative comment earlier: I write C++ professionally. C++ is the right choice for my organization, though for purely historical reasons. In the long run, I think we'd all be better off if Rust or similar languages replaced their predecessors. So my comment came from a place of personal frustration, the same that led me to find Rust in the first place.

The old wording was too negative; I'm just excited to show that contest programming, which one might imagine to be hard to translate, is not only practical but indeed arguably nicer in Rust.

jeeyoungk2• 9 years ago

Rust is one of my favorite programming language, and yet I learned how much I missed java after learning Rust. The fact that everything is dynamically dispatched and non-primitives are referential type allows you to program things without thinking too much. Plus, JIT & Hotspot can optimize these away - perform stack allocations for local objects and short-circuit dynamic dispatch for nonvirtual functions.

That said, my biggest Rust-gripe has been writing recursive / self-referential data structures. Borrow system is just not very convenient for that and you end up restructing the code to be a bit more Rust-friendly.

amelius• 9 years ago

Offtopic/meta. I'm wondering about a technology that allows us to NOT re-implement an algorithm cookbook every time a new language shows up.

freeordead5• 9 years ago

Ah yes, a unicorn.

Groxx2• 9 years ago

yea, I'm convinced it's possible to at least save a ton of work there. Probably not for optimized libs, but something is much better than nothing.

Rust though might be an example where such a thing wouldn't work (had it already existed), since the type system is legitimately different. But some things might work, and even a test suite you could develop against would be wonderful.

lmm• 9 years ago

An ABI that could express sum types would be a huge step forwards for the computing world.

.net and the JVM sort of achieve this, but at the cost of going managed for everything. That's mostly fine (and possibly entirely fine, to be honest; very few people who think they need non-GC actually do), but there are still a few diehards who insist on non-GC, and Rust is pretty much the first good language to have that option in a comprehensive way.

chc• 9 years ago

I'm pretty sure any technology that could do that would also obviate programmers for most tasks, since it apparently understands how to program and learn new programming technologies well enough to implement algorithms both correctly and idiomatically based on a specification.

Retra• 9 years ago

Like dynamic libraries and FFI? Reimplementing algorithms is the only viable way to get them completely in a new language. It's also the only viable way to improve them.

adrianN2• 9 years ago

But implementing algorithms is an enjoyable way of learning a language.

simplehuman• 9 years ago

Just a heads up: please put github tags

Show HN: Algorithm Cookbook in Rust

Related Stories