WEBVTT 00:00:00.010 --> 00:00:05.430 As we close the chapter on 2025 and celebrate our second year of Rust in Production, 00:00:05.730 --> 00:00:11.170 it's time to reflect on the highlights of the 17 episodes since our last holiday special. 00:00:12.130 --> 00:00:16.690 We looked at Rust from all angles, from cloud infrastructure to embedded systems 00:00:16.690 --> 00:00:19.310 and from robotics to satellite technology. 00:00:19.750 --> 00:00:25.110 One thing that all of these stories have in common is the passion and dedication 00:00:25.110 --> 00:00:29.870 of the Rust community to build faster, safer and more reliable software. 00:00:30.830 --> 00:00:35.610 In this special episode, I want to thank all the guests who shared their experiences 00:00:35.610 --> 00:00:41.190 and insights with us, as well as our listeners who have supported the podcast throughout the year. 00:00:41.710 --> 00:00:43.330 Let's look back at some of the 00:00:43.330 --> 00:00:46.650 memorable moments from the past year and celebrate Rust's achievements. 00:00:47.030 --> 00:00:51.010 This goes beyond the case studies we've covered. It's about the Rust community 00:00:51.010 --> 00:00:55.070 as a whole and the state of the Rust ecosystem at the end of 2025. 00:00:55.590 --> 00:00:58.910 My name is Matthias Endler and you're listening to Rust in Production. 00:01:00.630 --> 00:01:03.170 Chapter 1. Crossing the Chasm 00:01:03.170 --> 00:01:06.370 One of the recurring themes in our episodes this 00:01:06.370 --> 00:01:11.370 year has been the idea of Rust crossing the chasm, from early adopters to mainstream use. 00:01:11.530 --> 00:01:14.270 This is a significant milestone for any technology. 00:01:14.590 --> 00:01:18.970 It is also very rare for a programming language to achieve this level of adoption 00:01:18.970 --> 00:01:21.510 maturity, let alone in such a short time. 00:01:21.770 --> 00:01:25.630 Rust is no longer a language for enthusiasts and early adopters only. 00:01:25.630 --> 00:01:30.570 We're seeing more and more companies and projects adopting Rust for mission-critical applications. 00:01:30.890 --> 00:01:34.570 We've heard how Microsoft mandates Rust for new components in Azure, 00:01:34.890 --> 00:01:37.850 even going so far as adding it to the Windows kernel. 00:01:38.190 --> 00:01:41.970 We've heard from Volvo using Rust inside ECUs for their vehicles, 00:01:42.250 --> 00:01:48.110 and from Prime Video using Rust to power new components in the UI layer of their streaming service. 00:01:48.760 --> 00:01:54.000 But it's not just large companies. Some of my clients managed to build impressive 00:01:54.000 --> 00:01:55.920 products with small teams using Rust. 00:01:56.200 --> 00:01:59.780 I gave a presentation at this year's CodeTalks conference in Hamburg, 00:02:00.000 --> 00:02:02.020 where I shared a few case studies. 00:02:02.320 --> 00:02:06.740 For example, how a team of just three developers built a backend service for 00:02:06.740 --> 00:02:11.280 hundreds of thousands of students in Africa who learn math and English online 00:02:11.280 --> 00:02:13.240 through low-bandwidth connections. 00:02:13.600 --> 00:02:18.960 After the service went live, they had zero downtime and were able to scale to meet the demand. 00:02:19.480 --> 00:02:24.240 The maintenance dropped to almost nothing. All they focus on now is adding new 00:02:24.240 --> 00:02:25.920 features to help students learn better. 00:02:26.740 --> 00:02:29.720 Success stories like these are why I love working with Rust. 00:02:30.140 --> 00:02:34.140 It's hard to convey just how transformative Rust can be for a project until 00:02:34.140 --> 00:02:35.100 you see it in production. 00:02:35.320 --> 00:02:39.020 I wish it was easier to demonstrate this to potential new users before they 00:02:39.020 --> 00:02:40.660 commit to using Rust for their projects. 00:02:41.120 --> 00:02:45.760 Rust is and will remain a day-two language, and the benefits become more apparent 00:02:45.760 --> 00:02:47.520 at the later stages of development. 00:02:47.800 --> 00:02:50.860 That is still a significant barrier to adoption for many teams, 00:02:51.100 --> 00:02:55.520 as they have to make a leap of faith to trust that Rust will deliver on its promises. 00:02:56.140 --> 00:03:00.280 That's why as a community, we should not stop telling the stories of Rust in 00:03:00.280 --> 00:03:04.700 production, the good and the bad, to help other teams make informed decisions 00:03:04.700 --> 00:03:06.920 on whether Rust is the right choice for them. 00:03:07.080 --> 00:03:11.160 Speaking of sharing stories, it comes as no surprise that Rust has been named 00:03:11.160 --> 00:03:16.080 Stack Overflow's most admired language every year since our 1.0 release in 2015. 00:03:16.380 --> 00:03:21.280 And what people love the most about Rust is the sense of empowerment and versatility it provides. 00:03:21.580 --> 00:03:25.500 You can write everything from the top to the bottom of your stack in it. 00:03:25.860 --> 00:03:27.500 Here's Anton Lazarev from Brave. 00:03:27.880 --> 00:03:33.000 I just feel that Rust solves so many of our problems in software development that we've run into. 00:03:33.220 --> 00:03:37.720 And it really just works everywhere from embedded to desktop to mobile to even 00:03:37.720 --> 00:03:38.960 on the web with WebAssembly. 00:03:39.780 --> 00:03:43.380 I guess my take is that you better have a really good reason not to use Rust 00:03:43.380 --> 00:03:45.640 as the go-to language for your next project. 00:03:46.200 --> 00:03:51.200 Julius Gustavsson from Volvo and Charlie Marsh from Astral both echoed this sentiment. 00:03:51.970 --> 00:03:57.070 Yeah, empowering is definitely the key word. 00:03:58.070 --> 00:04:07.770 We have been super, super happy with this project and it really shows that Rust has a bright future. 00:04:08.830 --> 00:04:15.670 The hurdles that we mentioned before when we started the project and which made 00:04:15.670 --> 00:04:20.630 it basically a non-starter for most of the ECUs that we wanted to use it, 00:04:20.630 --> 00:04:23.750 those are all coming down and or have 00:04:23.750 --> 00:04:27.030 come down already so now you have qnx support 00:04:27.030 --> 00:04:30.330 for example you have infinium tricore support there 00:04:30.330 --> 00:04:33.190 was recently an article on how to run rust 00:04:33.190 --> 00:04:36.170 together with autostar classic autostar classic is 00:04:36.170 --> 00:04:39.650 the is the common automotive software 00:04:39.650 --> 00:04:45.450 framework that most automotive software runs in and now that is becoming available 00:04:45.450 --> 00:04:52.430 for for us as well there is work being done on various different automotive 00:04:52.430 --> 00:04:57.210 platforms and components so for automotive it's, 00:04:58.250 --> 00:05:00.850 definitely a bright future. 00:05:00.850 --> 00:05:03.810 It's kind of amazing i think that i only started 00:05:03.810 --> 00:05:06.950 writing rust like a few years ago and now we've shipped i mean 00:05:06.950 --> 00:05:09.830 along with a great team like we've shipped two of these tools 00:05:09.830 --> 00:05:13.070 that are having i i think at least a huge impact on 00:05:13.070 --> 00:05:15.870 python which is like the most popular or the second most 00:05:15.870 --> 00:05:19.110 popular programming ecosystem on earth so 00:05:19.110 --> 00:05:21.770 if you think about it like rust is kind of in a lot 00:05:21.770 --> 00:05:24.570 of ways rust is kind of like powering python at least to you 00:05:24.570 --> 00:05:27.930 know if i have say about it at least rust is powering is 00:05:27.930 --> 00:05:30.690 powering python and so i don't know i've always just 00:05:30.690 --> 00:05:33.490 felt i again i never considered myself 00:05:33.490 --> 00:05:36.250 to be like a systems programmer quote unquote and most of 00:05:36.250 --> 00:05:39.770 my career i was writing typescript python 00:05:39.770 --> 00:05:42.790 i mean i did some java professionally but like 00:05:42.790 --> 00:05:45.510 i had never except for like a course in 00:05:45.510 --> 00:05:48.770 college done any c i really hadn't done any c++ and 00:05:48.770 --> 00:05:52.350 like in the span of a few years i like learned 00:05:52.350 --> 00:05:55.490 to build this kind of software so i don't 00:05:55.490 --> 00:05:59.490 know i've had just like great experiences with the community and and being welcomed 00:05:59.490 --> 00:06:04.530 into it and learning the language and i think that's should continue to be a 00:06:04.530 --> 00:06:08.810 very important part of of rust is like welcoming people in and helping them 00:06:08.810 --> 00:06:12.390 learn because the impact that we can have by building this kind of stuff is 00:06:12.390 --> 00:06:13.810 just huge even outside of rust. 00:06:13.810 --> 00:06:18.450 Finally Andrew Tinka from Scythe highlighted the confidence that Rust gives 00:06:18.450 --> 00:06:19.530 you when building software. 00:06:19.830 --> 00:06:26.130 Rust as a language sits at a very valuable point in the language design space. 00:06:27.590 --> 00:06:33.730 Really is something special in terms of the guarantees it gives, 00:06:33.970 --> 00:06:40.190 the confidence you can have when building complicated systems with it, 00:06:40.490 --> 00:06:47.290 knowing that the language has struck such a valuable trade-off between competing 00:06:47.290 --> 00:06:52.190 concerns of safety and efficiency and expressiveness. 00:06:53.210 --> 00:06:58.950 It's a technology I'm glad to use every day that I use it, and I hope that its 00:06:58.950 --> 00:07:02.730 development toward these goals continues as long as it can. 00:07:02.910 --> 00:07:07.690 And there are few, perhaps no, other languages that can make such bold claims. 00:07:07.870 --> 00:07:11.650 I certainly struggled in the past when I had to struggle multiple languages 00:07:11.650 --> 00:07:13.270 for different parts of my projects. 00:07:13.530 --> 00:07:16.430 Besides navigating different ecosystems and switching contexts, 00:07:16.770 --> 00:07:19.610 I had to deal with different build systems, package managers, 00:07:19.710 --> 00:07:23.710 and toolchains. With Rust, I can master one language and use it everywhere. 00:07:23.910 --> 00:07:27.150 It feels like this is a very common sentiment among our guests. 00:07:27.370 --> 00:07:32.610 For many, Rust became a gateway into domains that would otherwise be inaccessible, 00:07:32.770 --> 00:07:36.550 to break down the barriers to entry and make larger parts of the software stack 00:07:36.550 --> 00:07:37.770 accessible to more people. 00:07:37.990 --> 00:07:41.710 And when the time comes to integrate Rust into an existing project, 00:07:42.130 --> 00:07:47.250 it happily takes a backseat to other languages, as it doesn't require a runtime or garbage collector. 00:07:47.250 --> 00:07:52.930 You can incrementally adopt trust in your codebase one component at a time and 00:07:52.930 --> 00:07:54.210 gradually leap the benefits. 00:07:54.490 --> 00:07:59.290 This is a very appealing proposition for many teams who want to slowly and carefully 00:07:59.290 --> 00:08:02.630 oxidize their codebase without the risk of a complete rewrite. 00:08:03.050 --> 00:08:05.970 Some of my clients do exactly that to great success. 00:08:06.310 --> 00:08:07.210 Chapter 2. 00:08:07.630 --> 00:08:13.270 Fear, Uncertainty, and Doubt It is hard to build systems that must not fail. 00:08:13.730 --> 00:08:15.590 Or at least do so gracefully. 00:08:16.360 --> 00:08:21.120 And when Rust fails, it makes the headlines. Many listeners will have heard 00:08:21.120 --> 00:08:23.620 of the first CVE in Rust for Linux recently. 00:08:24.060 --> 00:08:28.240 People were up in arms about it, claiming that Rust in the Linux kernel was 00:08:28.240 --> 00:08:30.920 a mistake, because Rust has bugs too. 00:08:31.140 --> 00:08:34.940 But if you take a closer look at what actually happened, you will learn that 00:08:34.940 --> 00:08:38.940 there is no known remote code execution or privilege escalation. 00:08:39.240 --> 00:08:45.200 Yes, Rust had a bug and crashed the kernel, but it did not lead to a memory safety issue. 00:08:45.880 --> 00:08:53.160 For context, 159 other CVEs were raised the same day, in the C parts of the kernel. 00:08:53.540 --> 00:08:59.240 In comparison, Rust had one CVE in five years in an unsafe code block in the 00:08:59.240 --> 00:09:01.100 implementation of a tricky data structure. 00:09:01.400 --> 00:09:04.160 That's not perfect, but it's pretty damn close. 00:09:04.480 --> 00:09:08.020 I often hear people spread the narrative that mistakes can also happen in the 00:09:08.020 --> 00:09:11.760 Rust compiler and the standard library itself, but fail to acknowledge that 00:09:11.760 --> 00:09:16.380 a bug in the compiler is a one-time cost that affects all Rust users, and once fixed, 00:09:16.660 --> 00:09:20.600 will never happen again for anyone rebuilding with the new version of the compiler. 00:09:21.000 --> 00:09:23.620 Isn't that a massive improvement over other languages? 00:09:24.200 --> 00:09:28.180 I believe we'll have to live with the fact that there are people who will jump 00:09:28.180 --> 00:09:32.040 on any opportunity to criticize Rust. It makes for great news headlines. 00:09:32.660 --> 00:09:37.400 However, I ask you to be critical of the narrative and read past the headlines. 00:09:37.400 --> 00:09:41.980 On the other hand, when people find bugs in popular software written in C, 00:09:42.200 --> 00:09:48.040 such as Redis or Curl, the reaction is often that mistakes happen and there is no holy grail. 00:09:48.720 --> 00:09:53.460 Yes, mistakes happen in every language, but just because there is no silver 00:09:53.460 --> 00:09:55.920 bullet doesn't mean we shouldn't strive to do better. 00:09:56.740 --> 00:10:01.700 C and C++ are littered with foot guns and undefined behavior that are impossible 00:10:01.700 --> 00:10:03.860 to fix without breaking backward compatibility. 00:10:04.880 --> 00:10:09.740 If you fix a bug in Redis, that fix only applies to Redis, but not to the countless 00:10:09.740 --> 00:10:12.240 other C programs that are vulnerable to the same issue. 00:10:12.520 --> 00:10:16.940 As an example, Redis had a remote code execution vulnerability in the Lua interpreter 00:10:16.940 --> 00:10:18.780 that was enabled by default. 00:10:19.720 --> 00:10:23.580 The bug was a use-after-free combined with a null-pointer dereference. 00:10:24.040 --> 00:10:31.300 It went unnoticed for 13 years. In Rust, this code would not compile. In C, it ended in a CVE. 00:10:31.800 --> 00:10:35.960 If you can't prevent safety issues before they happen, why isn't that a big leap forward? 00:10:36.500 --> 00:10:41.480 Rust should not be rejected because part of the problem still exists after adopting it. 00:10:41.740 --> 00:10:45.520 We need to stop comparing actual security problems to an unrealistic, 00:10:45.820 --> 00:10:48.740 idealized reality where all software is bug-free. 00:10:49.400 --> 00:10:53.820 While we may never get to 100% bug-free software, we should still adopt the 00:10:53.820 --> 00:10:59.040 tooling that provably reduces the number of bugs and security vulnerabilities in our codebases today. 00:10:59.320 --> 00:11:05.100 And no, writing more unit tests or just being more careful when writing C code is not enough. 00:11:05.400 --> 00:11:10.060 The irony is that teams that adopt Rust often are more rigorous about testing, 00:11:10.320 --> 00:11:15.420 not less, because their goal of following engineering best practices is what 00:11:15.420 --> 00:11:16.980 brought them to Rust in the first place. 00:11:17.860 --> 00:11:22.760 In code reviews and when writing tests, you can focus your testing efforts on 00:11:22.760 --> 00:11:27.100 the bugs that actually matter and not undefined behavior or memory safety issues. 00:11:27.340 --> 00:11:32.060 If you put the type system to work, you have more headroom to focus on logic 00:11:32.060 --> 00:11:34.360 errors and things no compiler can catch. 00:11:34.580 --> 00:11:40.440 Another popular narrative is that replacing stable C or C++ code with Rust is a waste of time. 00:11:40.800 --> 00:11:44.640 Why replace perfectly good software that works just fine? 00:11:44.920 --> 00:11:47.860 But of course, it's not fine, and it's not stable. 00:11:48.100 --> 00:11:53.140 The myth that C code, once written and working, never changes, is simply not true. 00:11:53.820 --> 00:11:57.860 For instance, Daniel Stenberg, maintainer of Curl and former guest on this podcast, 00:11:58.220 --> 00:12:01.760 posted that he's been working on this codebase for more than two decades, 00:12:01.980 --> 00:12:04.560 and that they get one bug report every three hours. 00:12:04.880 --> 00:12:09.000 More than half of Curl's source code has been changed within the last four years. 00:12:09.320 --> 00:12:13.560 Only 1011 lines from before the year 2000 remain untouched. 00:12:14.740 --> 00:12:18.120 The reality is that programmers continuously rewrite, refactor, 00:12:18.320 --> 00:12:22.680 and introduce new bugs while doing so. Here's Adam Hendel from Tempo on the topic. 00:12:23.100 --> 00:12:28.060 There's really no such thing as finished software or a finished programming language. 00:12:28.420 --> 00:12:32.820 It's not like software is distributed in the mail where you burn stuff to a 00:12:32.820 --> 00:12:36.240 disk and mail it out, and it just runs that way forever. 00:12:36.900 --> 00:12:43.760 Software is living, breathing organisms now, so things have to be constantly fed. it. 00:12:43.980 --> 00:12:47.820 One of my personal highlights this year was when Jon Seager from Canonical 00:12:47.820 --> 00:12:51.460 came on the show to talk about their experience shipping a replacement of the 00:12:51.460 --> 00:12:53.400 GNU core utils in Rust for Ubuntu. 00:12:53.760 --> 00:12:57.660 If you've got a really interesting Rust project and you think it's a great fit 00:12:57.660 --> 00:13:01.020 for some of the work that we're doing and you want to have a conversation, then reach out. 00:13:01.480 --> 00:13:04.980 I'm open to ideas. I don't know all of the high-quality Rust projects out there. 00:13:05.220 --> 00:13:09.580 And if there are implementations of things that are very widespread that you 00:13:09.580 --> 00:13:11.420 think we should be using, then let me know. 00:13:12.080 --> 00:13:17.180 Another highlight was Google's update on their Android OS work to integrate Rust into the platform. 00:13:17.400 --> 00:13:23.040 For the first time ever, memory vulnerabilities fell below 20% of total vulnerabilities thanks to Rust. 00:13:23.300 --> 00:13:28.820 For reference, the industry norm is around 70%. But their biggest surprise was 00:13:28.820 --> 00:13:30.660 Rust's impact on software delivery. 00:13:30.900 --> 00:13:34.220 Rust changes have a four times lower rollback rate. 00:13:34.600 --> 00:13:39.920 Just imagine the eye-watering cost savings on a platform like Android with 3.9 00:13:39.920 --> 00:13:41.800 billion active devices worldwide. 00:13:42.080 --> 00:13:46.060 The safer path is now also faster and more cost-effective. 00:13:46.560 --> 00:13:50.800 But Rust isn't the right choice for everyone. For instance, we had Richard Feldman 00:13:50.800 --> 00:13:52.340 on the podcast talking about ROC. 00:13:52.760 --> 00:13:57.080 His team moved from Rust to SIG for their compiler work, mainly because of slow 00:13:57.080 --> 00:13:59.180 Rust compile times. And it makes sense. 00:13:59.460 --> 00:14:03.020 When you're iterating rapidly on a compiler, waiting for Rust to compile can 00:14:03.020 --> 00:14:06.220 be a real productivity hit. That's a legitimate trade-off. 00:14:06.820 --> 00:14:09.500 gives you faster iteration at the cost of safety guarantees. 00:14:09.880 --> 00:14:11.920 For their use case, that made sense. 00:14:12.440 --> 00:14:16.320 The lesson isn't that Rust failed. It's that different problems need different 00:14:16.320 --> 00:14:19.600 tools. And that's okay. Let's hear it from Richard himself. 00:14:19.860 --> 00:14:22.080 Rust is such a big language with so many different features. 00:14:22.440 --> 00:14:26.140 And there's so many different competing concerns you can have in terms of like, 00:14:26.180 --> 00:14:30.900 I need to balance like making this safe and also making it well tested and well 00:14:30.900 --> 00:14:32.140 structured and easy to maintain. 00:14:32.760 --> 00:14:34.940 There's all these different things that you could be thinking about. 00:14:35.140 --> 00:14:40.460 And I think focusing on what's the end user going to get out of my program. 00:14:40.720 --> 00:14:43.040 That's the most important thing. That's got to be the North Star. 00:14:43.560 --> 00:14:47.860 And even for teams that stuck with Rust, compile times remain a common pain point. 00:14:48.480 --> 00:14:52.180 Here's Tom Hacohen from Svix and then Alexandru Ene from Prime Video. 00:14:52.560 --> 00:14:55.820 I think we got to fix the compilation time. I don't know what we can do. 00:14:57.120 --> 00:15:01.640 It's really, I think it's, it is like a bottleneck for a lot of people. 00:15:01.840 --> 00:15:05.560 It is, it is a pain and it's, it even affects, you know, like ID, 00:15:05.920 --> 00:15:09.540 like, you know, language servers and all of that, like when it's, 00:15:09.660 --> 00:15:12.100 when things are slow. so we got to fix that thing. 00:15:12.750 --> 00:15:18.470 One thing I wish we were focusing on more in the ecosystem is build speeds, 00:15:19.010 --> 00:15:21.770 like iteration times, just getting those down. 00:15:22.050 --> 00:15:29.110 I think right now we more or less accepted that Rust would be slower to compile or whatever. 00:15:29.390 --> 00:15:36.030 But I think there's examples out there in other languages where people have gotten great results. 00:15:36.610 --> 00:15:40.310 Like Zig, for example, is pushing a lot on the build speed and the iteration 00:15:40.310 --> 00:15:46.910 time. And I think we all win if we can iterate in our codes super fast, 00:15:47.050 --> 00:15:50.070 like press whatever button you're pressing to run your code. 00:15:50.290 --> 00:15:53.370 And if that's like two seconds, it's just magical. 00:15:53.670 --> 00:15:58.150 Once it gets past a certain threshold, my productivity just, I get distracted. 00:15:58.470 --> 00:16:03.350 So I would just work a lot and investing on, if I were like... 00:16:04.790 --> 00:16:08.410 If I had the power to invest in anything in the Rust community, 00:16:08.670 --> 00:16:12.910 I would say this would be a really good place to invest in iteration times and 00:16:12.910 --> 00:16:14.530 build speed to decrease them. 00:16:15.410 --> 00:16:22.430 Chapter 3. Outlook for 2026 Rust is here to stay, at least in the Linux kernel. 00:16:22.810 --> 00:16:26.250 What started as an experiment is now an official part of the project, 00:16:26.570 --> 00:16:31.050 with Google's binder driver and the Nova GPU driver being some of the most prominent 00:16:31.050 --> 00:16:34.150 examples of Rust code merged into the mainline kernel this year. 00:16:34.670 --> 00:16:38.930 As Danilo Krummlich mentioned in his episode, new contributors are excited to 00:16:38.930 --> 00:16:40.630 work on Rust code in the kernel. 00:16:40.870 --> 00:16:45.070 We now reached 200 billion downloads of crates from Crates.io this year. 00:16:45.810 --> 00:16:49.350 200,000 crates are published and there is no sign of slowing down. 00:16:49.570 --> 00:16:51.370 The ecosystem is fantastic. 00:16:51.730 --> 00:16:55.830 The maintainers are working hard to keep up with the growth of the ecosystem 00:16:55.830 --> 00:16:57.830 and that is highly appreciated. 00:16:58.390 --> 00:17:00.950 Here's Kevin Guthrie and Edward Wong from Cloudflare. 00:17:00.950 --> 00:17:09.450 Yeah, no, like we there's a bunch of HTTP ecosystem, things that there's there's 00:17:09.450 --> 00:17:15.890 a great maintainer for all of it is like open source, the HTTP, 00:17:16.530 --> 00:17:23.570 literally HTTP crate, H2, you know, a lot of those are our core dependencies for Pingora as well. 00:17:23.990 --> 00:17:29.350 And the maintainer, Sean is incredible at what he does. so. 00:17:30.280 --> 00:17:33.940 Yeah, the thing I was going to thank the Rust community for is for being so coherent, 00:17:34.300 --> 00:17:38.860 especially around HTTP things like the hyper ecosystem, the H2, 00:17:39.180 --> 00:17:45.280 all of those things are so ubiquitous that it makes integrating with existing projects much easier. 00:17:45.280 --> 00:17:50.060 Specifically, I was working with a ClickHouse client that is an official ClickHouse 00:17:50.060 --> 00:17:53.880 client that the ClickHouse team puts out, but I needed to add a new feature 00:17:53.880 --> 00:17:58.220 for rotating MTLS certificates, which obviously their client does not support. 00:17:58.400 --> 00:18:04.560 But because they expose access to the hyper-HTTP client under the hood, 00:18:04.760 --> 00:18:06.060 it made it an easy thing to do. 00:18:06.060 --> 00:18:08.980 It's just such a good 00:18:08.980 --> 00:18:12.340 experience to come to like if you need a feature you already have the tools 00:18:12.340 --> 00:18:17.660 necessary to add functionality to tools that are published by other people in 00:18:17.660 --> 00:18:20.620 a coherent way something that you don't get in java something that you i don't 00:18:20.620 --> 00:18:24.380 know if you get in go that's not my ecosystem but as a former and recovering 00:18:24.380 --> 00:18:27.460 java programmer it's very nice the. 00:18:27.460 --> 00:18:32.840 Ecosystem though i'm sure there are gaps from time to time generally if you 00:18:32.840 --> 00:18:34.860 are looking for a particular, 00:18:35.740 --> 00:18:38.480 pattern or thing you will either find out that 00:18:38.480 --> 00:18:41.300 it is hard to do so or that someone else 00:18:41.300 --> 00:18:45.820 has already tried to at least some extent to do it and has a working very much 00:18:45.820 --> 00:18:50.080 like you know if not production ready nearly production ready implementation 00:18:50.080 --> 00:18:56.560 of it so the rust ecosystem in general has has just been kind of the amount 00:18:56.560 --> 00:19:00.780 of excitement that folks have within the community, 00:19:01.480 --> 00:19:05.280 is a great sign of promise. 00:19:05.740 --> 00:19:10.620 We're also looking forward to a year packed with Rust conferences and meetups around the world. 00:19:10.860 --> 00:19:14.400 To name the currently announced events, RustNation in London, 00:19:14.540 --> 00:19:20.200 England on February 18, Rust in Paris, France on March 27, Rustikon in Warsaw, 00:19:20.360 --> 00:19:24.920 Poland, March 19, RustWeek in Utrecht, the Netherlands, May 18, 00:19:25.440 --> 00:19:28.540 RustConf from Montreal, Canada, September 8, 00:19:29.160 --> 00:19:33.880 Oxidize in Berlin, Germany, September 14th, EuroRust in Barcelona, 00:19:34.100 --> 00:19:38.740 Spain, October 14th, and RustLab in Bologna, Italy on November 1st. 00:19:39.080 --> 00:19:43.440 Oh, and if you're bored during the holidays, why not code a little Rust? 00:19:44.220 --> 00:19:49.420 For example, I recently finished recording my Rust HTTP server course for CodeCrafters, 00:19:49.460 --> 00:19:50.980 and I will link to it in the show notes. 00:19:51.700 --> 00:19:54.960 Chapter 4. What will Rust 2026 bring us? 00:19:56.220 --> 00:20:00.880 I personally look forward to Cargo Script, which will allow you to run Rust scripts 00:20:00.880 --> 00:20:02.680 without the need for a full cargo project. 00:20:03.080 --> 00:20:06.300 This will make it easier to experiment with Rust and share small, 00:20:06.500 --> 00:20:07.920 self-contained code snippets. 00:20:08.100 --> 00:20:11.780 Simon, the editor of this podcast, has been using it heavily over the last year 00:20:11.780 --> 00:20:15.840 with the nightly releases, and I'm planning to use it for all my future blog posts. 00:20:16.690 --> 00:20:20.790 The async ecosystem still has some rough edges but i'm confident that we will 00:20:20.790 --> 00:20:24.790 see steady progress in this area better pin ergonomics could help here. 00:20:24.790 --> 00:20:27.970 I think my primary message to 00:20:27.970 --> 00:20:30.830 the rest community is just polish up 00:20:30.830 --> 00:20:34.150 async get it to be the best 00:20:34.150 --> 00:20:37.150 experience it can ever be there are some pitfalls 00:20:37.150 --> 00:20:40.830 now even though rust 2024 edition stabilized 00:20:40.830 --> 00:20:43.710 the async closures very happy 00:20:43.710 --> 00:20:46.730 about that but there are still some questions around just 00:20:46.730 --> 00:20:50.530 observability of what is happening within an async 00:20:50.530 --> 00:20:53.370 context and how do you navigate that and and just 00:20:53.370 --> 00:20:56.130 yeah and getting to getting to 00:20:56.130 --> 00:21:01.210 the bottom of issues related to to as i said the blocking issues we have and 00:21:01.210 --> 00:21:06.530 just cancellation safety and and drop safety and async drop and all these paper 00:21:06.530 --> 00:21:11.910 cuts that that just are not completely answered yes that that would be my message 00:21:11.910 --> 00:21:13.430 to really polish up that. 00:21:13.690 --> 00:21:16.670 That would make selling Rust to others much easier. 00:21:16.930 --> 00:21:21.870 That was Vegard Sandengen from KSAT. On top of that, we need even more learning 00:21:21.870 --> 00:21:25.050 resources for beginner to intermediate and advanced Rust sessions. 00:21:25.570 --> 00:21:29.990 Andrew Burkhart from 1Password agreed with me on that point in his episode when he said. 00:21:30.130 --> 00:21:32.910 Thank you to the Rust community for keeping things open, right? 00:21:33.010 --> 00:21:36.410 Like I said, there's so many books out there and the Rust book and the brown 00:21:36.410 --> 00:21:39.310 version of the Rust book is amazing with the quizzes and things. 00:21:39.430 --> 00:21:44.150 That's all great. I think we also need to continue to or increase how often 00:21:44.150 --> 00:21:45.810 we relate to people, right? 00:21:45.930 --> 00:21:49.970 Meet people where they're coming from, understanding, you know, 00:21:50.030 --> 00:21:50.990 what they do and don't know. 00:21:51.330 --> 00:21:54.650 Obviously, with someone coming from my background, that was really critical 00:21:54.650 --> 00:21:59.530 that people here did that, you know, and there's a lot of, you know, 00:21:59.730 --> 00:22:04.830 acronyms or computer science knowledge. And there's so many people in Rust that are so smart. 00:22:05.210 --> 00:22:10.210 It can be difficult. So there's a lot of that that I think we can continue to improve on. 00:22:11.290 --> 00:22:16.170 Finding paths to not just go from zero to Rust, which I think we've started 00:22:16.170 --> 00:22:20.070 on, or we have a really good path to with the Rust book, or to go from Haskell 00:22:20.070 --> 00:22:22.210 to Rust or C++ to Rust or whatever, right? 00:22:22.670 --> 00:22:26.370 I think there is a missing middle gap of how do you go from TypeScript to Rust 00:22:26.370 --> 00:22:28.770 or this or that, where it's like, I understand variables. 00:22:28.890 --> 00:22:32.450 I understand all that. So the whole beginning of Rust book feels boring and getting lost. 00:22:32.910 --> 00:22:36.690 And then we go straight into memory or, you know, async and really how async 00:22:36.690 --> 00:22:39.290 works. And there are these missing pieces sometimes. 00:22:39.890 --> 00:22:42.170 There's things like rustlings, which is incredible for like, 00:22:42.670 --> 00:22:43.670 exercise based learning. 00:22:44.150 --> 00:22:48.130 But at the end, I didn't see exercise six mesh with exercise 10. 00:22:48.350 --> 00:22:51.050 And now I understand that those are two maybe related things. 00:22:51.210 --> 00:22:55.110 And so, you know, a lot of people say, right, build stuff, that's that is the greatest way. 00:22:55.370 --> 00:22:59.370 But that's not not realistic for everyone. You know, I was working before I 00:22:59.370 --> 00:23:02.550 became a developer, I was working, you know, 50, 60 hours a week, 00:23:02.690 --> 00:23:05.450 and then, you know, coming home and trying to keep, you know, 00:23:05.550 --> 00:23:06.830 my apartment clean and all these things. 00:23:07.010 --> 00:23:09.410 And I did not have that much time left to just build things. 00:23:09.510 --> 00:23:13.270 It took me a year or more just to try and figure out, you know. 00:23:13.970 --> 00:23:15.590 React or whatever it was at the time. 00:23:16.230 --> 00:23:20.310 And if it were Rust, I was trying to learn, I had no idea how long it would have taken, you know. 00:23:20.450 --> 00:23:27.410 And so I think just finding ways to bridge those gaps and make cohesive learnings 00:23:27.410 --> 00:23:30.690 that make concepts actually make sense holistically. 00:23:30.690 --> 00:23:37.030 I think that's a really, really huge area that will help a lot of people close the gap of Rust, 00:23:37.090 --> 00:23:39.890 which will help Rust adoption you know I've seen that even here where it's like 00:23:39.890 --> 00:23:44.290 we may have something we build and go because ramping that team up on rust is 00:23:44.290 --> 00:23:47.730 just not feasible in the time frame we have well if there's better resources 00:23:47.730 --> 00:23:51.210 to get from go to rust that becomes a non-issue you know and I think we'll see 00:23:51.210 --> 00:23:56.870 rust grow the more we are open to bringing people in from more intermediate spaces into rust. 00:23:57.460 --> 00:24:00.480 Apart from that, my wishlist is pretty short. For the most part, 00:24:00.620 --> 00:24:03.020 I wish for faster compile times and more clippy lints. 00:24:03.520 --> 00:24:06.500 My personal hope is that we will see slow and steady growth. 00:24:06.740 --> 00:24:08.840 The next 10 years aren't guaranteed. 00:24:09.860 --> 00:24:14.260 Rust's success depends on maintaining the delicate balance between growth and stability. 00:24:14.620 --> 00:24:19.900 We don't need to be everywhere all at once. We don't need to be a solution to every problem. 00:24:20.220 --> 00:24:23.660 And we need to bring people along thoughtfully, not hastily. 00:24:24.100 --> 00:24:27.680 It's important to collaborate with other programming language communities, 00:24:27.680 --> 00:24:31.760 such as C++, to work on safer interoperability. 00:24:32.160 --> 00:24:37.580 Work on this has already started. There are billions of lines of C++ code representing 00:24:37.580 --> 00:24:40.980 trillions of dollars of value stored in C++ codebases. 00:24:41.200 --> 00:24:45.180 It would be naive to assume that all of this code will be rewritten in Rust 00:24:45.180 --> 00:24:49.360 anytime soon, so we need to find ways to make Rust and C++ work better together. 00:24:49.640 --> 00:24:52.040 Let's hear Victor Ciura from Microsoft. 00:24:52.400 --> 00:24:54.420 In order for Rust to succeed, 00:24:55.270 --> 00:24:58.150 C++ does not mean it has to die. 00:24:58.510 --> 00:25:05.470 So it's not a zero-sum game. It's very much about many, many years from now 00:25:05.470 --> 00:25:07.270 where we need to learn to coexist, 00:25:08.450 --> 00:25:13.910 both in terms of software interop and in terms of community interop. 00:25:14.110 --> 00:25:21.290 So I think my message would be, let's spend more time on improving these bridges 00:25:21.290 --> 00:25:28.670 and making sure the languages work well together so that we can successfully 00:25:28.670 --> 00:25:32.090 have these and as communities learn to collaborate better. 00:25:32.390 --> 00:25:37.870 Other initiatives like BorrowSanitizer, a work-in-progress LLVM instrumentation 00:25:37.870 --> 00:25:43.210 pass for detecting aliasing violations in multi-language Rust applications, also received funding. 00:25:43.390 --> 00:25:47.190 I will add a few links in the show notes for those interested in learning more. 00:25:47.470 --> 00:25:52.830 Looking ahead to 2026, the Rust project has 41 active project goals. 00:25:53.190 --> 00:25:56.550 13 of them are designated as flagship goals. 00:25:56.830 --> 00:26:00.890 They are working on making references more ergonomic, exploring things like 00:26:00.890 --> 00:26:05.010 field projections, they are making compilation faster with parallel frontend 00:26:05.010 --> 00:26:10.010 work and build std, they are improving async Rust, but there's still a lot of work to do there. 00:26:10.170 --> 00:26:13.510 The next generation trade solver is getting closer to being production ready. 00:26:13.770 --> 00:26:17.510 They're stabilizing Polonius for better borrow checking, getting Rust for Linux 00:26:17.510 --> 00:26:21.790 into stable Rust by stabilizing the compiler and language features the kernel needs. 00:26:22.270 --> 00:26:25.330 Let's hear it from Danilo Krummrich from the Linux kernel team. 00:26:25.650 --> 00:26:27.050 When it comes to... 00:26:28.230 --> 00:26:31.870 The things that the kernel needs 00:26:31.870 --> 00:26:37.010 from the Rust project, and there are definitely things that are needed. 00:26:37.430 --> 00:26:44.390 I also want to say thank you for the great collaboration. I'm not that involved myself. 00:26:44.930 --> 00:26:49.710 I definitely want to make that clear. We have other members from the Rust for 00:26:49.710 --> 00:26:57.290 Linux team who are regularly talking to people from the Rust core team and from 00:26:57.290 --> 00:26:58.690 the Rust project in general. 00:26:59.710 --> 00:27:03.670 And there is great collaboration already ongoing. 00:27:04.010 --> 00:27:09.850 So yeah, I'm very happy to see how things go. 00:27:10.250 --> 00:27:14.270 For a full overview, check out the Rust project goals update on the Rust blog. 00:27:14.570 --> 00:27:16.310 I will link to it in the show notes. 00:27:16.630 --> 00:27:20.550 And now we wish you all happy holidays and a prosperous new year filled with 00:27:20.550 --> 00:27:22.630 safe and reliable software written in Rust. 00:27:23.210 --> 00:27:28.110