2025-06-05 - HackMD

--- date: 2025-06-05 url: https://hackmd.io/VWSTFnqdTki0q-48R8DMyg --- # 2025-06-05 People: Jack, Pete LeVasseur, Niko Matsakis, Mart, TC, Tomas Sedovic Notes: Tomas Sedovic ## Agenda Jack: We neeed more interviews. Particularly from people who tried Rust and failed. But more interviews in general. We discussed doing more coding (the finding commonalities from the responses kind). We're talking to a company that could do this for us. Tomorrow at 12 Eastern. Jack: Last week, Ernest [last name?] might have a slot at the last RustConf talk -- that may be a good target to get our next milestone on getting some coding then. He's setting up roundtable discussions with community groups. We're doing interviews in the meantime. Haven't decided as a group what interviews we're missing. Niko: I can attend that meeting tomorrow. We should establish the coverage we want. Jack: That's what I was trying to do with the survey. That's useful but maybe saying whom we need to interview is not needed right at this point? Right now, we haven't hit any given set group of people. Niko: Without some structure, it's too open-ended. Jack: The big one is people who don't use Rust. Niko + Pete: People who didn't choose Rust but were told to use it (at work). Jack: Another way: splitting out people based on how they learned Rust. Niko: Want to revisit our research questions -- are we happy with them? Are we getting the data we want? Niko: There's a script, but it's very loose. The reason is to give flexibility on following the person where they take you. Not bias the results by looking for what you expect to see. Niko: Want to talk about the Open Source community in Rust but we could drop it from the interviews. Jack: Maybe we poll the Rust project now about what they think is important and then contrast it with what our interviews show. Niko: Focusing on what people need form Rust is valuable. And following up on what Rust needs form People. Pete: The group of people I know is people in the automotive space. Niko: Get a sample, don't need to strive for a representative one a-priori. Pete: It's okay there are biases on whom we're interviewing and then adjust them by quantitative approaches. Jack: People often come to Rust for memory safety, performance, "modernism" Niko: Something that's missing: What do people love about Rust? Jack: Go back to the survey and look at the people who started using Rust recently question. Niko: Less value in people that know Rust well and have been using it for a long time. Jack: The idea of "people who know Rust well are us" -- not necessarily true. People using Rust in other industries etc. Jack: How to get the most data from experts without getting data that we already know? Niko: A lot of these questions are about how Rust is awesome. Ask where should we not use Rust. Jack: Python meetups: ask where is an overlap, where you wouldn't use Rust. E.g. quick prototyping? Unsure how to specifically target without going to other languages. Jack: Ask in our survey questions: if you haven't used Rust, what did you use? And why? Niko: Asking the same questions about other languages -- can be useful. Pete: I agree -- there was a similar question: "Why do you still use C", going to language meetups. Niko: On a mission at Amazon to remove the use of all unsafe languages. But also would write C if doing a lot of unsafe stuff -- Rust is painful in doing that, C is straightforward. Pete: (from an interview) if you bring something to the C committe and say this affects safety but will massively increase performance they'll go for it. If there's a negligible perf hit and has impact on security, it'll be a tough battle. If there's a performance hit and has a security impact, it won't happen. Niko: re C++ we could interview people from Amazon from Fire TV. Pete: We have an internal team that's porting Rust to a proprietary RTOS. Pete: Interview lined up with a member from the C++ committee. Jack is shadowing. Jack: Where do we want to end up today? Niko: If I would put two hours this week into this, what would I do? Only realised recently that I'm leading interviewees on. Asking questions and getting back generalities. Want to know what groups of people we want. And want to know what questions we're looking to answer. This "what do you love about Rust" thing is something we're missing. Jack: We should refine our interview and research questions. But also tell the whole team that future interviews should address this (what we're looking for, outlined below). Pete: Makes a lot of sense. Having a core set of questions we try to ask and then having some questions that we'd like to see and having them linked to these core questions. Jack: Yes, having the interviewers driving towards the questions we're interested in. Niko: Who's doing interviews outside of this meeting? Jack: We're not super active with interviews right now (outside of this meeting). Niko: What does each of us want to do next week? (nothing is okay) Jack: Probably nothing. Join this meeting next week and spend this hour brainstorming. Niko: Interview Pete. Think about research questions in advance of the next meeting. Maybe interview one other person. Pete: Got interview tomorrow and Tuesday next week. Put up feelers for people in Open Source doing Rust stuff. Jack: Pete's really good at interviews. Pete: Here's a [link](https://docs.google.com/document/d/1BggDUh20qEYU3KJvk9WMSNzWMd1vHWX4N_RLHdPzBR8/edit?usp=sharing) to the interview question bank I put together ahead of the interview I did last week. https://hackmd.io/eFdEtyGaSWapeCoy3SFuuA ### Categories * Students * People who don't use Rust * "Rust-curious": positive about Rust but not using it yet * What makes them feel positive? * What is stopping them? * People who recently learned Rust * Coming from Java, building network systems (Amazon) * Coming from C++, building embedded systems (Woven, Volvo Cars, ...) * Coming from C++, building consumer devices (e.g., Amazon FireTV) * tl;dr * ...? * People who know Rust deeply * How does Rust work for the domain they are working in? * People in safety-critical / embedded / C / C++ * People who teach Rust * Corporate training * (Pete running Comprehensive Rust training for C++ devs this week, possible to discuss) * Team experts ### Research questions #### Rust the technology > "How does Rust fit into the overall language landscape? What is Rust's mission?" Niko's answer: * Foundational software Jack's answer: * Memory safety * Performance * "Modernism" > "What do people love about Rust?" > "What do people find frustrating with Rust?" > "What brings people to Rust and why do they choose to use it for a particular problem...?" > "What would help Rust to succeed in these domains...?" * Async * Embedded * Machine learning * Scientific computing * Game development * GUI development * (others) > "How can we scale Rust to industry-wide adoption?" * "And how can we ensure that, as we do so, we continue to have a happy, joyful open-source community?" ### Rust the global project * "How can we improve the experience of using Rust for people across the globe?" * "How can we improve the experience of contributing to and maintaining Rust for people across the globe?" --- * "How does Rust fit into the overall language landscape? What is Rust's mission?" * "What do people love about Rust?" * "What do people find frustrating with Rust?" * "What would " --- Questions that we might ask * Think back to the last week of working with Rust -- tell me about it. * Was that typical of your experiences in working with Rust? * Think back to the last few people who asked you for help in Rust * Was that typical?