News for hackers. Stuff that matters.
Re:Improving Rust compile times to enable adoption of memory... by Bjartr on Friday February 03, 2023 @09:00AM (#34641116)
It's funny, any other post on HN about improvements to Rust I've seen are chock full of comments to the effect of "I guess that feature is nice, but when will they improve the compile times?" And now many of the replies to this post are "Faster compiles are nice, but when will they improve/implement important features?"

The Rust dev team can't win!

Re:Improving Rust compile times to enable adoption of memory... by jerf on Friday February 03, 2023 @09:09AM (#34641207)
This is when it is important not to model the comment section as some sort of single composite individual.

Since it is impossible to mentally model them as the number of humans they are, I find it helpful to model them as at least a few very distinct individuals, or sometimes just as an amorphous philosophical gas that will expand to fill all available comments, where the only question is really with what distribution rather than whether a given point will be occupied.

Re:Improving Rust compile times to enable adoption of memory... by thechao on Friday February 03, 2023 @09:47AM (#34641659)
I used to be a grad student adjacent to Bjarne Stroustrup; he has a quip he's used a bunch: if no one's complaining, no one cares.

I see all of these complaints — both the volume, and the count — as great indicators of Rust's total health.

Re:Improving Rust compile times to enable adoption of memory... by guhidalg on Friday February 03, 2023 @10:00AM (#34641863)
So true. I use a similar quip at work: "Take the shortcut and if we're not out of business when it becomes an issue, then it will be a good problem to have".
Re:Improving Rust compile times to enable adoption of memory... by rascul on Friday February 03, 2023 @12:59PM (#34645130)
I guess that depends on the work. By day I repair houses and taking shortcuts can mean I could have an even bigger problem to solve in a few months. Luckily I have yet to be in such a situation myself but I've fixed other's shortcuts a number of times.
Re:Improving Rust compile times to enable adoption of memory... by guhidalg on Friday February 03, 2023 @05:28PM (#34648687)
Yes, I should specify I’m talking about software development. Physical products and work rarely have the luxury of making mistakes or taking shortcuts, the universe is a harsh place.
Re:Improving Rust compile times to enable adoption of memory... by capableweb on Friday February 03, 2023 @09:51AM (#34641709)
1) HN is maybe one organism if you zoom out enough, but it consists of people with wildly different opinions, you'll have capitalists arguing with anarchists here, any post is bound to have both sides, no sides and every side, all on the same page 2) it's easier to complain about stuff, somehow. Not sure why, or if it's extra prominent on HN in particular, but people tend to start thinking "Why am I against this thing?" and then write their thoughts, rather than "Why do I like this thing?". Maybe it is more engagement to write something that can be challenged, and people like when others engage with them, so they start to implicitly learn to be that way.
Re:Improving Rust compile times to enable adoption of memory... by jchw on Friday February 03, 2023 @09:56AM (#34641793)
I think pessimistic and cynical reactions are the literal lifeblood of news aggregator comments sections. It's been like this for as long as I can remember across as many aggregators as I've ever used.

Part of the problem is that news aggregates reward people who comment early, and the earliest comments are the kneejerk reactions where you braindump thoughts you've had brewing but don't have anywhere to put. (Probably without actually clicking through.)

Re:Improving Rust compile times to enable adoption of memory... by nindalf on Friday February 03, 2023 @10:37AM (#34642454)
I agree in general except HN also rewards quality a bit more. All new comments get a few minutes to bask at the top of their subthreads. So a really good, late comment can still get to the top and stay there.
Re:Improving Rust compile times to enable adoption of memory... by jchw on Friday February 03, 2023 @10:52AM (#34642704)
To a degree, the comment ranking algorithm helps, though long/fast threads do often leave brand new replies buried upon posting.

Still, I believe what makes HN unusually nice is just the stellar moderation. It is definitely imperfect, but it creates a nice atmosphere that I think ultimately does encourage people to try to be civil, even though places like these definitely have a tendency to bring out the worst in people. Having a deft touch with moderation is very hard nowadays, especially with increasingly difficult demands put against moderators and absolutely every single possible subject matter turning into a miniature culture war (how in the hell do you turn the discussion of gas ranges vs electric ranges into a culture war?!) and the unmoderated hellscapes of the Internet wrongly painting all lightweight moderation with a black mark.

I definitely fear for the future of communities like HN, because the pressure from increasingly vile malicious actors as well as the counter-active pressure from others to moderate harder, stronger, faster will eventually break the sustainability of this sort of community. When I first joined HN, a lot of communities on the Internet felt like this. Now, I know of very few.

Re:Improving Rust compile times to enable adoption of memory... by smolder on Friday February 03, 2023 @12:28PM (#34644487)
Another part of it is just psychology. People seem much more inclined to join discourse to make objections than to pile on affirmative comments, which generally an upvote suffices for.
Re:Improving Rust compile times to enable adoption of memory... by yadoomerta on Friday February 03, 2023 @01:30PM (#34645703)
It's also partly the site's culture. Not saying it's wrong, because it adds some noise and not much new info, but I've been downvoted before for posting comments like "Thanks for saying this!"
Re:Improving Rust compile times to enable adoption of memory... by boredumb on Friday February 03, 2023 @10:43AM (#34642557)
Rust is amazing, I truly believe a large number of people are intimidated by it and so go out of their way to shit on it and pretend like it's only for some niche IOT device... when it's just as easy to write out a full crud application in Rust as any other language at this point.
Re:Improving Rust compile times to enable adoption of memory... by crispinb on Friday February 03, 2023 @07:12PM (#34649727)
I wish extreme Rust solipsists would stop blankly stating as a fact that working in Rust is 'easy' just because they find it so. If you do, good for you, but your experience is not universal. Many don't.

I find Rust extremely difficult and slow to work with. It takes me a long, long time to get anything working at all. Easily 5x more than any other language I've used (and that's many, and a good handful professionally), and I've been learning Rust for over a year.

Figures are hard to come by of course, but anecdotally I'm the only person in my circle who's continued using Rust. All the others have dropped out because they just find it too hard to get anything done in. Not sure if this is true, but I heard on a podcast the other day that one of the big surveys showed Rust to have the biggest learning drop-out rate of any mainstream programming language. That wouldn't surprise me, and comports well with Rust being the 'most loved' (people tend to love skills they have gained with much effort!).

Re:Improving Rust compile times to enable adoption of memory... by steveklabnik on Friday February 03, 2023 @08:31PM (#34650364)
Do you remember which survey? I’d love to see it!
Re:Improving Rust compile times to enable adoption of memory... by crispinb on Friday February 03, 2023 @09:48PM (#34650876)
No I was half-listening and think it was mentioned in a loose way, and I think I'm conflating it with some other mention in a different recent podcast so that's more than a few levels of looseness.

I've just scrubbed back through - the podcast was https://syntax.fm/show/571/supper-club-rust-in-action-with-t..., which follows the pleasing practise of providing chapters (yay). It was Tim McNamara speaking from about 12'40": what he actually said was that about half of Rust learners who do drop out fall away because of the purported difficulty of the language. Very different from my faux summary so apologies for the grievous misrepresentation.

I still find Rust as hard to use as others I've spoken to do though. I'm kind of keeping at it for reasons specific to projects I have in mind, as well as a certain dense stubbornness.

Re:Improving Rust compile times to enable adoption of memory... by timClicks on Saturday February 04, 2023 @04:02AM (#34652953)
Hey Steve, I was referring to the official Rust Survey in the podcast interview. I'll track down a specific link, but the stat comes from one of the PDFs rather than the analysis blog posts. I'm having difficulty finding it currently.
Re:Improving Rust compile times to enable adoption of memory... by steveklabnik on Saturday February 04, 2023 @09:17AM (#34654651)
It’s all good, I know where that is, thanks!
Re:Improving Rust compile times to enable adoption of memory... by steveklabnik on Saturday February 04, 2023 @09:19AM (#34654661)
Thank you! It’s all good, I just hadn’t heard that before and was curious how it was figured out! The Rust Project has long acknowledged that Rust is tough to learn. That’s why I got to have a job back then! And there’s still new good work going on in that space. I don’t think the nut has been fully cracked yet.
Re:Improving Rust compile times to enable adoption of memory... by crispinb on Saturday February 04, 2023 @08:15PM (#34660497)
> there’s still new good work going on in that space

Is there anything there someone like me (ie. having difficulty with Rust) might usefully contribute to? I'm a little overwhelmed right now trying to keep a roof over my head, but am compiling a list of things I might like to help with when the current storm has passed.

Re:Improving Rust compile times to enable adoption of memory... by steveklabnik on Saturday February 04, 2023 @08:27PM (#34660587)
I don't know what's good to help with and what isn't, because I haven't been involved with Rust development for a pretty long time at this point.

Personally I think contributions work best when you're trying to solve a pain that you personally have or at least have some sort of connection to, so I'd encourage you to consider what/how/why you struggled to learn, and then try to fix that. I know it's vague, but it's the best I've got right now!

Re:Improving Rust compile times to enable adoption of memory... by crispinb on Sunday February 05, 2023 @12:30AM (#34662243)
Fair enough, cheers. I've been keeping notes on and off. I may be able to turn them to use at some point.
Re:Improving Rust compile times to enable adoption of memory... by marcosdumay on Friday February 03, 2023 @11:07AM (#34642975)
That's how winning looks like.
Re:Improving Rust compile times to enable adoption of memory... by lambdaxymox on Saturday February 04, 2023 @12:17PM (#34656091)
Some problems are good problems to have. If different groups of people are taking the time to substantively and constructively complain about different shortcomings of Rust and its ecosystem, it shows that there's a wide enough audience of serious users out there. My take is that the Rust devs are doing something right, even if it seems like they can't win.