Rust 1.78.0: Last Rust News Video...ever?
Вставка
- Опубліковано 1 чер 2024
- New changes Rust 1.78.0! Sponsor me on GitHub for more videos: github.com/sponsors/CleanCut/
Links:
Rust 1.78.0 Blog Post - blog.rust-lang.org/2024/05/02...
Rust 1.77.2 security advisory - blog.rust-lang.org/2024/04/09...
ByteCode Alliance WASI 0.2 blog post - bytecodealliance.org/articles...
Rust's Google Summer of Code - blog.rust-lang.org/2024/05/01...
Artwork - foundation.rust-lang.org/img/...
0:00 Last video?
0:21 Intro
0:36 Catch Up
0:39 1. RustConf 2024
1:11 2. Rust 1.77.2
1:41 3. Changes to Rust's WASI targets
2:44 4. Default to rust-lld on nightly x86_64-unknown-linux-gnu
3:18 5. Google Summer of Code 2024
4:22 Rust 1.78
4:31 6. Stabilize #[diagnostic] and #[diagnostic::on_unimplemented]
5:23 7. Asserting unsafe preconditions
6:27 8. Deterministic realignment
6:52 9. Windows 10 is the oldest supported version
7:08 Other Rust Changes
7:18 10. Stabilize #[cfg(target_abi = ...)]
7:44 11. Make async-fn-in-trait implementable with concrete signatures
7:53 12. Matching NaN is now a hard error
8:08 13. Allow static mutable references to arbitrary types
8:33 14. Extend invalid_reference_casting lint to include references casting to bigger memory layout
8:59 15. New lint: non_contiguous_range_endpoints
9:25 16. Add wasm_c_abi future-incompatibility lint
9:44 17. Update indirect_structural_match and pointer_structural_match lints to match RFC
9:59 18. Promote lint to hard error: const_patterns_without_partial_eq
10:13 19. Split refining_impl_trait lint into _reachable, _internal variants
10:29 Cargo
10:33 20. Stabilize global cache database
11:01 21. Stabilize lockfile v4 format
11:20 22. Auto-detect when terminal supports non-ASCII Unicode characters
11:50 23. Officially support target.(triple).rustdocflags
12:09 24. Ensure users know a feature is being created
12:20 25. Improve error when adding registry packages while vendored
12:27 26. Collapse generated file list
12:46 27. Print "Creating" rather than "Created" Status
13:00 28. Cargo keys and definitions note moved from Cargo.toml to output
13:14 The End...?
14:24 So let's try this...
15:18 Today's Artwork
16:17 Outro
Visit ultimaterustcourses.com for Ultimate Rust Crash Course and other Rust courses or sponsor me on GitHub: github.com/sponsors/CleanCut/ - Наука та технологія
This is a lot like a “Behind the change log” that explains the actual point of the changes vs simply listing them. This is invaluable to the community.
Thank you so much! Explaining the actual point of the changes is exactly what I'm shooting for. I would settle for "valuable enough that community/organizations will support him so he can keep doing it".
Nathan! Please seek sponsorship from the Rust Foundation! They’re the right people to help you (and the community) here ❤️
I seriously do hope they'd help yeah!
It's ironic but they are the one foundation that I can't trust to come through on such matters, due to their obvious past idiocies.
I will try that! Any insights as to how to go about doing that?
@@NathanStocks I don’t know personally, sorry. I do know that @fasterthanlime is or was sponsored, maybe he can advise further
Didn’t fully see the video yet, but for what’s worth the first thing I thought about when a Rust update came was “I need to watch Nathan’s update video”. I actually waited for your video instead of reading the official blog post 🎉
Aww. You made my day! ❤
@@NathanStocks ❤. Now that I have finished the video you should definitely try to get your courses on PluralSight and LinkedIn Learning. Those are used heavily on enterprise and the ones from LinkedIn Learning are very mediocre. I know because I've seen them all and your on Udemy and I think you style of teaching than really shine there.
Hey Nathan. First time viewer here. Maybe using Patreon or an alternative platform might be helpful?
Indeed, same here. His videos are absolutely fantastic.
Just started watching the video.
I actually enjoy your Rust news videos with explanation of everything. It's the go to place for this kind of content, ever since I discovered it. I would be really sad if this goes away, because nobody does it like you (and this clear and well). With every update of Rust, the first thing I do is to check if you posted a new video about it. So please, if its possible, then keep doing this. Just for me, who cares about the other users. :D Thank you for all the efforts you did so far. Now I will continue watching the video. Have a good weekend.
Thank you for your kind words! That's very encouraging.
For what it's wroth, I actually did purchase your course videos. It was very helpful still while I was an "ok" rust dev to an "actually pretty good."
Thank you so much! It must have been on Udemy, because my self-hosted courses still haven't gotten a customer.
For what it's worth, yes these videos are (were?) a nice a way to keep up with ecosystem changes while doing other work than having to read the notes. But I don't have a better idea for monetization, and I can truly empathize with the position you're in. But yeah, I don't personally need a course, I just need time to use what I already know.
Side note, I had no idea you contributed to game engine development. There's information in there people could find useful and interesting. Kind of like what Asahi Lena does, but maybe without the vtuber graphics/audio, you are your own interesting personality already.
I'd love to expand the videos into actually covering the ecosystem. Think release videos for tokio, serde, bevy, crossbeam, criterion, etc. It all takes time, and right now I need my time to earn money. (sigh)
Hey Nathan, it’s hard to keep putting things out there with little coming back. I have a project out there and have never received a donation. It just seems people expect free labor most of the time.
I would suggest taking the production quality back a lot, and just doing a fun video when you’re excited about something.
But, no harm in just moving onto the next thing!
Sorry to hear about the single dad situation. That’s a huge burden. I would like to help, but not sure how.
Good advice. Cutting down the videos could allow me to keep doing them. I just...really like covering everything!
And moving onto the next thing may just need to be what I do.
Please don't stop with your Rust news
Thank you for your kind words. We'll have to see what happens!
i dont even read the rust official posts since there are such indepth update videos
Awww, thanks!
I haven't watched the video completely yet but I'd just like to say that this is very important piece of content. I often just can't aggregate all the changes and why they were made (for reading them later on). Your videos are definitely what fills this kinda space and I'm prettty sure it does for a lot of other people too.
Thank you! That is exactly what I'm shooting for.
This is my first time watching one of your videos, and i think that, while having a course is fine, you need to have videos here on youtube as teasers.
Ones that teach, but dont go in depth. Short, maybe.
Basically, i have no reason to even know you have a course unless i happen to stumble upon your detailed update videos and am otherwise bored.
Doing this will bring more people to your channel in general, and help the community as a whole. This will mean more video monitization, more sponsorships, and more viewership in general.
Alright, so that's another vote for focusing on courses.
Dear Nathan, we love the work you do, and I have some 'small' rust dev teams who learn a lot of the content you make. We just became a sponsor (Trust1Team), and would like to see you achieving your goal! I'm pushing your courses to new devs, as that speeds up the onboarding process for new consultants in our software engineering team.
You're amazing! My first organization sponsor. It wouldn't take many sponsors like you to hit my goal. ❤ Tell me why my Rust News videos are valuable to you!
First time viewer, loved the video. I really hope you will find a sponsor, and if it isn’t going to be too much of a hustle please update us in the community tab if you will manage to find one wish all the best to you!
By the way your content is tailored towards very advanced audience in the rust ecosystem as you are pretty much covering the bleeding edge of rust development, so it isn’t too surprising that no one is buying your course from the advertisements on those videos. One potential idea you can try is to make a low level video that will explain some simple features of rust, for one bad example is how does .len() on arrays work and why is it O(1) complexity. Making analogies with C in those cases will also benefit the community as you will be exposing the working of rust to people who are familiar with c but not rust, which can hypothetically get you costumers for your rust course.
Thank you! I'm glad you loved the video.
So in other words, make intro course videos to get people to buy courses? That's definitely one of the options that falls under "stopping doing news and working on courses". Not a bad idea, but not going to help me keep doing the news. 😆
@@NathanStocks haha fair enough, but ya I don’t think bleeding edge news will bring in new people to buy your course sadly. I was thinking of doing this along with the news but ya I can understand if you aren’t enthusiastic about doing that, wish best of luck to you really hope you will get a sponsor!
first video i watched of yours was your live Bevy coding in a room full of people and frankly i enjoyed it. since then i have been watching only these Rust update videos. i was not even aware of Udemy courses. you should talk about them during these videos or any content you put out.
Thanks for watching!
Hmm, I’m one of those who already did your courses, had them on Udemy. And because of the courses, I found you here 😀. Thank you for your work, and I agree with previous comment, I really wait your videos to learn about the changes in new releases
Thank you for your kind words!
To go off of what @Dygear mentioned, it would be cool creating a useful project or two that solves real problems and incorporates all the new features in the rust change log.
That way, you can document how each new feature was useful compared to other similar features in the language and why it was good/bad using that new feature along with covering the scenarios where the new feature is best applied in.
That would be super cool.
Don't feel any pressure to do anything you don't want to do. But I personally love these videos. Thank you for each and every one!
Thanks for your advice and kind words!
Hey Nathan, I got both of your courses on udemy just because of these videos, so please please please drop the new ultimate course! :D
Yes, I understand it is tough. Choose life, spend time with your child, take a break, and if at some point you come back, we'll still be here. Either way, take my (small) sponsorship.
People like you make a difference in the world. Thank you for your kind words and your sponsorship! All sponsorships mean a lot to me, regardless of size.
I think you should keep up with making these videos, though the views might not be much but I think that's because not many people know you posts youtube video, other than udemy courses. I honestly really like watching updates like this because you simplify a lot of things that the changelog didn't explain very well, which make me understand the reason for the fix and update that Rust pushes out, and I think that's also the same for people watching your channel
Thank you for your kind words! It takes me a lot of time and effort to understand the changelog myself. One of my favorite things is figuring out the reason something was important and sharing that with people who care.
I'm just... So people still write to this day code to programming languages standard libraries that accept unchecked user input? Just wow!
As far as I know its a bug in Windows that makes it unchecked. There was no way for the Rust team to know this. It's not just Rust that had this problem, but bunch of programming languages, including Java, Go, Haskell, NodeJS, Python and more. This was not an oversight by library developers, but by Windows developers.
Wait. I see what you mean. But if the operating system API documents to clients the idea that strings will be checked, it's NOT the Rust standard library responsibility to fix this. And it shouldn't really. It's up to either Microsoft to fix it, or developers to validate the input themselves.
Ok. Replying to myself. I had to go and take a look. This wasn't making sense to me. The issue is not that there is a bug on Windows. And the Windows API documentation is clear that it is the client responsibility to parse strings. The issue is instead that cmd.exe has its own string splitting mechanism and there was no way the standard library could work consistently on all cases. The fix was to return an error when it cannot safely escape the string. It's up to the client to do their own checking.
@@Marfig Thanks for digging deeper!
Suggestion: Make automated tools to create the videos, if already doing so, open source and let people help you optimize them. In essence, do what Manim did. Idk, something like IRC/Feed -> preprocessor (shell/grep) -> Slides/PDF (Sent) -> ffmpeg (video) will save a HUGE amount of time. You can contact me or publicly explain your workflow to help developers do that for you, no reason to code even.
Hmm. I mean...maybe that could help a little? Most of my time is spent understanding why something was done and what was so important about it and then stating that. Maybe AI could give that a try and reviewing and fixing it would be faster? I dunno. Not completely sold, but not throwing the idea out either.
@@NathanStocks Fair, just assumed the most time consuming is the video editing and gathering information. If your time goes on other things, then yeah this is not the best idea. Anyways hit me up if I could help in anyways.
Is that sponsorship thing us-only?
No! GitHub Sponsors is currently available in 35 regions. github.com/sponsors/
Cool! What's that color scheme? Can I get it on neovim?
pretty sure its monokai, or one of its derivatives
@@scarecat thanks! yeah after doing some searching i think its called "Molokai" so you were right :)
I did, indeed, begin with Monokai from Sublime Text back in the day. I've adapted it through migrations to Vim, IntelliJ, and VS Code. Here are my current settings you are welcome to adapt: github.com/CleanCut/dotfiles/tree/main/vscode
@@NathanStocks cool! Thank you!
Thanks for being honest! I've watched almost every video of you reviewing the Rust releases, and it's going to be really sad if you stop doing them. They are great. Is it possible to lower the standard and spend less time on the videos? I feel like I would gladly watch even a simple, unedited live video where you just sit and read the Rust release changelog and tell all you can about the described changes.
Thank you for your kind words! I don't think I could do a lower quality version (it just wouldn't feel good to me--I want to be proud of what I create), but being more selective about what I report on is definitely an option!
@@NathanStocks Depending on your target audience, a different platform and structure might be a way to feel better about that. Sketching an idea: you advertise that the Saturday after every release you'll do a twitch stream going over what's in the release at, say, 3 PM EST, which you time-box to something you're comfortable with. Then you can try to make it a bit of an event that people join for and chat with you and other excited fans about what's new, and aim to get bits donations (that are more efficient for small amounts than some other things are) or subscriptions to offset your effort. Because that sets a different polish expectation than a youtube video like this does.
Maybe that's a bad idea, or not how you'd like to do things, but just musing about different possibilities because I agree with you that on youtube I'm usually looking for something focused and nicely produced (like you've been doing) rather than something that can get more rambly.
@@pfeilspitze Thanks for the idea!
Jokes on them, I link my installed Rust packages with mold which is even faster (on Gentoo, so it is global)
Haha! Love it. You should definitely be able to micro-optimize your own projects far more than any defaults. I won't complain about a big improvement in the default, though!
@@NathanStocks for normal distros and packaged software, it's really difficult since everything is prebuilt and only binaries are installed. On Gentoo and similar distros, the packages are downloaded as source (if possible) and compiled in-situ, so per-package optimization is available.
For Rust, changing linker in Gentoo is as easy as defining your own RUSTFLAGS as "-C linker=clang -C link-arg=-fuse-ld=mold"
However, as you say, it's much easier for your own projects. You just have to put that in your config.toml instead as
[target.x86_64-unknown-linux-gnu]
rustflags = ["-C", "linker=clang", "-C", "link-arg=-fuse-ld=mold"]
It doesn't answer how to make more money but I just wanna say that this format was actually pretty nice too... Just put less in it, most of the changes may not worth covering... The amount put in this video was just perfect.
Interesting idea! Heavier editing and only reporting on the more interesting things. That could make the videos more interesting and somewhat quicker to produce. Only somewhat faster, because more than three quarters of my time is spent in the initial research, and I would still be doing most of that.
I think these need to focus more on what *you* can provide value about more specifically. I read the release blog posts, so while these videos are nice they're also not essential as-is, to be frank.
Maybe if you did more "here's one cool thing" videos? Like for this release, maybe you could do a "here's how the ub checks help you write correct code, and BTW if you're interested in that then here's my course about `unsafe` that you should buy!" kind of thing.
Being exhaustive feels like it must be so much work, but I don't know how much value most viewers get from you covering *everything*.
Covering things exhaustively is definitely a ton of work. Thanks for your suggestion!
The improvement of security is not news that I want to hear about programming, it's boring, there are many kinds of excited released which I expected to hear.
that's my perspective that is of course wrong at all.
Thanks for your feedback.
Looking forward to your Rust Advanced courses, I've enjoyed both your beginner and intermediate courses.
Thank you so much! I'm glad you enjoyed my courses and want more! This is a vote _against_ making more news videos. 😂
@@NathanStocks😂 I can’t keep up with ‘real’ news! 🗞️ 😂
@@GaryChike Oh, neither can I. I gave up on keeping up with "real" news 10 or 15 years ago. Rust News is far more wholesome and rewarding.
Re 13: @8:31 - You're not my real dad.
Your real dad would warn you against using static mutable references, if he were me. 😂
@nathan can I get a discount on Udemy video?
Sure. Here's about 87% off of my Ultimate Rust Crash Course on Udemy. www.udemy.com/course/ultimate-rust-crash-course/?couponCode=ABD7539FAD05201FDBE8
The exact discount depends on the Udemy country / price region you are in. It expires 5 days from now. Anyone can use it.
@@NathanStocks Much appreciated, I bought it.
People pay to attend a programming language conference, what?!!???
Conferences cost money to host => them not being free is entirely reasonable
Yep!
I agree with your decision to cancel this Rust news format. Make it 3 minutes long.
Look at your video statistics - how many people watch entire 16 minutes? I am afraid you are making high effort video for a very small group of people.
The ~30 minute videos get 15-25% watching until the end. The shorter videos tend to get about 20-33% watching to the end. The shortest videos get about 50% watching until the end. The longer videos get more total people watching, interestingly enough.
First time here and not a rust programmer (at most a beginner).
I could not understand the appealing side of this video: looking for $1000/month for Rust News?
Looking at the video list, it's averagely 2 videos per month, and that will be about 48 hours per month!
I don't understand what your question is. Would you like to try my Rust crash course? I posted an 87% discount link in one of the other comments.
@@NathanStocks Not asking a question. Just a comment on your appeal for support - that you should keep it real.
Also, thanks but I was not looking for a rust course. This video was recommended by UA-cam.
I really like these videos, but 1000usd for 20 hours of work seems like a lot, no? I understand that you want to be paid for your work and I definitely support that, but it seems to be a bit out of proportion. Regardless, I suggest contacting the rust foundation, they sponsor these kinds of projects.
You are kinda right but also no. Notice 1000$ is not THAT much if you are Canadian (probably, the heck do I know I earn 100$ or less a month coding in my country so yea). Also, pay attention, he is not saying 1000$ for 20hours, he is saying 1000$ a month. He may definitely work more than 20 hours. Also, there is another flaw in your logic. One may work 10 years, (free lets assume, like education and practice) to do something worth 1000$ in 5 mins which somebody without that experience does 10 hours. If you hire the 5 minute guy, you must pay him for 10 hours, even maybe more. If he is taking less time, he actually borrowed his past time to the future, in other words, in some sense, he already worked those long hours, you just don't know. Yes, I know making videos is not heart surgery but if I was him, and I was also you at the same time, commenting this and reading this, I'd be furious as the editor because I know for a fact that me as a critic, cannot realistically do the same job with the same efficiency and for less money. I, Yas, personally find making UA-cams and scavenging data a boring and (frankly) stupid hell of a job. So yeah, sorry for the long response, since your comment had thoughts behind it and you seemed like somebody that actually may want to see other points of views, I wrote this.
Quite frankly, no, it's not a lot for me, especially given my negative starting balance from investing $25,000 in equipment and hundreds of hours into making free videos for 15 months. It's a _minimum_ that could help me justify to myself spending more time making these news videos that I love. But if I can't, I'll pivot that time (and equipment) towards other pursuits that people don't expect to be free, like my courses.
@@Yas-gs8cm Actually, @holly_hacker is not right at all. $1000/month for two years won't even pay for the existing equipment I use, much less pay me anything for my time or cover any ongoing costs. It's just...a start.