Follow


I was thinking about these drew devault posts, the one about rust and the one about wayland, and thinking how it feels like he was arguing the same thing on two different sides, and it seems someone else had the same feeling and actually did it :)

boringcactus.com/2021/02/09/an

@sotolf
Particularly these lines stick out to me from the original (devault) blog post:

> Rewriting your code in Rust is always going to introduce new bugs, including security bugs, that wouldn’t be there if you just maintained the C code. Maybe there are undiscovered bugs lurking in your C codebase, but as your codebase ages under continuous maintenance, that number will only shrink.

No? That implies that no new functionality is added.

(1/?)

@sotolf And that library dependencies aren't upgraded. And that there aren't refactorings to better support new/improved behavior.

What world is devault living in?! What they're saying isn't true for _any_ codebase except for *maybe* the most minimal libraries (and even there, standards and algorithms change).

(2/?)

@sotolf It's absurd to say that a codebase only undergoing maintenance will only decrease in bugs because in reality, that's a stupidly restrictive scenario only applicable to a *very small* selection of programs and libraries. In reality, most programs add new functionality or get refactored or update due to dependencies or change algorithms. And all of those *necessarily* could introduce bugs, in *any* language.

(3/?)

@sotolf So yeah, it's weird that devault makes the claim that stagnant codebases reduce in bugs over time.

(4/4)

/endrant

@sotolf HAH just read devault's Wayland post and got the reference of this post now... 😛

Yeahhh...holy shit, the cognitive dissonance is strong on this one.

@sotolf The wayland rant was about a specific type of person who complains how "you can't do $X on wayland" without actually trying it.

There are people who will say rust is terrible without ever writing a line of it or even interacting with it, for whatever reason, but he is not a part of that group.

If you can equate those writings as opposites, then you're probably taking them out of context, and it's a shame he didn't provide that context in the articles… would have reduced conflict.

@sotolf e.g. people complain "haha rust has to use `unsafe` to make a doubly linked list, and it's meant to be a safe language! how terrible!" and those people have no idea what they're talking about, same as those who say "wayland doesn't support x11 forwarding so it breaks my workflow" have no idea what they're talking about.

@sotolf THIS!!!
Thank you so much for sharing! It has really struck a chord with me.
All those people who say "Rust sucks" and "C over everything" and bla bla have the following in common: They are just arrogant people, who are afraid that they loose their "elite" position as "systems programmers", because #Rust now "empowers everyone to build reliable and efficient software"! 🦀

So Rust is an "equalizer" in this regard and is a great enabler for diversity in this space!

#Diversity #Programming #Arrogance

@janriemer Rust lacks architecture support, moves too fast without a specification, has only one implementation, etc. What works today can easily break tomorrow. That's why C programmers don't like it, it is not really reliable as a language.

To be fair, I'm not anti-Rust, the language is awesome but I fully understand people who don't like it.

Ref: drewdevault.com/2021/02/09/Rus

@sotolf

@yerinalexey @janriemer Drew seems to be pretty anti-rust, and hypocritic when you read through some of his stuff.

@sotolf But he at least provides reasons why Rust is not perfect and not just "rust sucks, period" which seems to be covered by earlier reply

@janriemer

@yerinalexey @janriemer That's true, but giving reasons doesn't make it be less one sided, if he doesn't like it he should just use what he likes and move on, we who use rust knows it has many flaws, things like compiling being slow, the borrow checker sometimes disallowing safe constructions, the unsupported systems and so on, we don't need someone on the outside screeming it out in addition ;) unless he's wanting to help fix it that is ;)

@sotolf He does and he wants Rust to be a somewhat good language. It's not typical "Rust bad, write in C" thing

> To the Rust team: it’s time to calm down. Slow down the language, write a specification, focus on improving your tier 2 and tier 3 targets, expand to more platforms, and work on performance, stability, and accessibility. Invest more in third-party implementations like rust-gcc.

@janriemer

@sotolf This article is written so nice!

I laughed hard!

But I think, to be fair, that we shouldn't start throwing mud in each others direction. The Rust community is better than that!

@musicmatze It's literally just a parody on this article

drewdevault.com/2021/02/02/Ant

since the same guy also wrote this a couple of days earlier

drewdevault.com/2021/02/09/Rus

With that background maybe it doesn't sound as shitty ;)

Sign in to participate in the conversation
Fosstodon

Fosstodon is an English speaking Mastodon instance that is open to anyone who is interested in technology; particularly free & open source software.