People who were happy with Rust didn't have same need to criticise Zig in online spaces as Rust is the established player in the C alternatives space. (Though Rust is on the other side when compared to C once you expand the space to "all low level programming languages").
Also for people who don't care about the space at all, Rust has had years of exposure to promote fatigue, while Zig hasn't (yet).
> People who were happy with Rust didn't have same need to criticise Zig in online spaces as Rust is the established player in the C alternatives space. (Though Rust is on the other side when compared to C once you expand the space to "all low level programming languages").
I think that this effect is...more real and intentional than I wish it was or think is healthy.
There's no denying that Rust benefited greatly from heavy "language evangelism" from around 2014-2022 or so (rough subjective range of dates, ending with when I feel like Rust became sufficiently mainstream that "comment section evangalism" is no longer the primary driver of the language's growth). The atmosphere that surrounded the language evangalism during this era -- from my recollection as someone who was aware of Rust but explicitly not interested in it at the time -- felt like one of curiousity and exploring new innovation in the programming language design space, and it didn't come across as pushy or some kind of moral imperative.
Additionally, most of the evangelism was not driven by the Rust project itself, but more of a grassroots thing from:
- Early users
- People who thought the language was interesting without having used it
- The attention that came from prolific early Rust projects like Servo and ripgrep
- The attention that came from the language's technical innovations
(The only person I remember regularly seeing in Rust threads who was actually associated with the project was Steve Klabnik, and I consider him a stellar example of how to represent a project online in a way that's respectful and not pushy or combative.) And eventually I had a programming task where, thanks to those deep, insightful, and respectful comment threads, I recognized that Rust was the right tool for the task I was trying to accomplish, learned the language myself, found it a joy to use, and began using it more regularly.
But once Rust became mainstream, the eternal September kicked in. Now there seem to be quite a lot of people who are more interested in "language wars" than in healthy, technical discussion of programming languages, and it's exhausting. Many threads about C or C++ or Rust or Zig or Fil-C or whatever seems to get overtaken by these obnoxious, repetitive, preachy, ideological comments about why you should or shouldn't be using Rust, often with very ill-informed armchair takes.
I think this is just something that comes from mainstream Internet culture and is not very representative of "the Rust community", and especially not the Rust project -- most of the serious writing and comments coming from people actually involved with Rust have that spirit of respect, openness, and curiousity about programming language design that I remember from the 2010's. A lot of people who work on Rust are deeply interested in the tradeoffs and innovations of new programming languages and fundamentally excited about the possibility of Rust becoming obsolete because something better has come along. But now that Rust adoption is increasing and the broader Internet culture is being forced to consider it, you see a lot of unhealthy reactionary takes from people who act like programming languages are a zero-sum game and they need to pick a side.
Now Zig is where this gets a little bit odd to me. My impression (as an outsider to the Zig community as well) is that Zig's leadership has chosen to view adoption as a zero-sum language war. It's as if they considered:
- How Rust benefited from language evangelism during the 2010's
- How many other "C-replacement" languages over the past 25 years have failed to gain large-scale adoption
- How "language war" discourse is extremely effective for enagement on places like HN
...and decided that the best way to increase Zig adoption is by intentionally leaning into that high-engagement "language war" discourse. Maybe it's not intentional -- and there's certainly a lot of cool ideas and high-quality technical discussion that comes out of the Zig community -- but I've also seen a consistent pattern of snarky or hostile comments from Zig leadership towards Rust people (such as calling Rust users "safety coomers", or jumping into in-depth technical discussions of Rust's downsides like compiler performance with substance-less cheap shots). Whereas I can't remember ever seeing anything but respect and curiosity from people involved in the Rust project towards other communities.
I'm just tired of seeing people draw battle lines around programming languages instead of promoting a spirit of learning from each other and making our tools better.
I've looked at the major rust implements from big corp. It's all Arc, copy/clone, and people getting fired.
What I'm frustrated about is a general pattern in online discourse of treating programming language adoption as a zero-sum game and pitting the communities against each other. This is nothing remotely new -- holy wars over tooling are just about as old as computing itself -- but Rust seems to have become right at the center of it in the last few years, and I think this is likely just a result of it becoming "mainstream" enough for people to care about it outside of a niche circle. I consider that a community issue with Rust, but probably an unavoidable one given the scale of the Internet.
The thing that feels different to me about Zig is that the language's leadership participates in the "holy war", which is different than what I have observed from Rust's leadership. It's one thing if random netizens are trash talking the "opposing team" like a sports fan, but I would expect people who are deeply involved in programming language design and representing their community to know better and behave better than that. It's surprising and disappointing to me; it drags down the overall quality of discourse and sets a bad example for the language communities.
Maybe my own perspective on this is skewed, perhaps I'm looking at Rust's past with rose-tinted glasses. But I think my experiences are comparable. From 2015-2020 I was primarily working with Swift and closely following that language's design and development, and the Rust community (which I was not involved in at all) felt like friendly neighbors. The two communities shared similar goals and had a lot of like-minded people, and because of that there was a lot of overlap and amicability between the communities. Both languages' teams were frequently comparing notes and copying features from one another.
Now I'm primarily working with Rust and closely following that languages' design and development, and it seems to me like the Zig community (which I am not involved with at all) should be on the same friendly terms for the same reasons, but the vibes I see from them are "this town isn't big enough for the two of us", and I'm bothered by that.
> I've looked at the major rust implements from big corp. It's all Arc, copy/clone, and people getting fired.
I have no idea what this is supposed to mean. Are you saying you've seen projects implemented in Rust go poorly? If so, I don't know what to say without any context, besides that I don't think there's a "one size fits all" for programming languages.
Or are you saying big companies' contributions towards the Rust project itself have gone poorly? I don't really know what to say about that either. I have felt that the direction the project has taken over the past few years has represented me and addressed my needs well, so I have confidence in their technical direction and leadership for the time being.
Zig's BDFL was an early major supporter of Rust. It was Rust's core maintainer group who excluded him and warred against his language. The leaders decided any "memory unsafe" language was against "consensus"; first C, then Zig. They even posted on Twitter, calling Zig a "step backwards". This Language Supremacism was even worse in conferences and in private.
Today, you can see a lot of the core ex-leaders still sniping at at other language and even successful OSS projects based in those languages. The Zig community has its faults but Rust takes the cake for relentless gamesmanship.
Andrew and I literally got some tacos last week. I mean, I'm not in Rust leadership anymore either, but this sort of thing just isn't recognizable to me outside of, again, a few posts by one person who no longer even worked on Rust.
It's nice you got tacos with Zig's BDFL and I don't categorize you as a "hater". Rather, I think the Rust Core Group altogether was toxic. Unable to see their hate as hate, and having memories largely selective and largely untrustworthy.
EDIT: I hadn't fully gone through the comment section on this one yet and yikes it's worse about it than normal.
Non-negligible in that it happens in almost every thread about rust. As in it isn't a rare occurrence.
C language is strongly tied to Unix ecosystem. C language and current compilers give this illusion of "ultimate control" to the programmers (although C has been interpreted quite freely for the sake of optimizations). With GNU C compiler finding a niche (i.e. being a free as in beer C compiler and a userland that's not bounded by the original Unix or BSD licensing / patents) and taking off in many servers the beliefs were kind of validated.
Rust attacks those beliefs in multiple fronts:
First and foremost, Rust comes from an inherent distrust to the programmer's abilities to write secure software. Rust creates strong obstacles against writing "magic" programs that freely interpret memory. Programmers are not treated as gods but fallible mortals who needed to be guided. Many C programmers interpret this as an insult to their abilities to manually check and verify a program. Rust is no less capable than C, it has all sorts of escape hatches that can be used in performance critical parts.
Rust compiler itself and the most of the programs written in Rust are permissively licensed. This is to avoid possible issues with copyleft licenses combined with statically linked binaries. Without having a stable ABI or ability to incorporate third party libraries post-build, GPL and LGPL create hurdles that doesn't exist in the mostly dynamically compiled world of C.
Rust has a more equal regard of the operating systems, including non-POSIX ones like Windows. Despite most people in the world interact with non-POSIX APIs the most, accepting it as another step.
Rust community tries its best to create a safe space for marginalized parts of the society. Go to any Rust conference and you'll see an over-representation of LGBTQ people. This probably rubs majorly white male subset of strongly identifying C programmers up wrong way.
The community also welcomes a bunch of different ways of thinking and questioning the basics of every single decision we made when we were building the systems of today.
Zig, on the other hand doesn't attack any of those "base principles". It fixes the most annoying parts of C programming: fluid integer types, really weak type system, hostile dependency management. It still trusts the programmer 100%. The default behavior is still unsafe.
People who choose Rust are also particularly worried about the security of the programs and they would like to prove and validate the language's existence and goals. They do that by reimplementing very popular and very senior projects in Rust and compare its performance and safety against existing C projects too. Many such projects have enjoyed a long time of no competition and now there is one.
I note that, while different circles may call for "purity" (fans of, e.g., Unix, Lisp, or Haskell), I think honestly pursuing the purity of writing programs to achieve one's goals means rejecting dogmatism and accepting pragmatism. Adherence to beliefs is fine if, and only if, the beliefs are fundamentally justified....
Zig's appeal lies in its deliberate simplicity and explicitness, offering a different path to high-performance, low-level programming compared to Rust's safety-first, more complex approach. This divergence in philosophy and implementation leads to different reception within the programming community.
There is no need for a new fund, the existing funds should be directed towards development and there should be greater transparency about how funds are gathered and spent.
This post is devoid of details and action. It should read that the Rust Foundation is going to convert to a 501(c)(3) and ensure that current funds are spent on development. Otherwise, this is a shell game with the existing funds and how they are going to be spent. Understand that sleight of hand bodes poorly and suggest impropriety.
mperham•3mo ago
oli-obk•3mo ago
stefanos82•3mo ago