Skip to content
PodSubs
ShowsPeopleGuestsJobsNewsEvents
More
BlogPodCred
Menu
ShowsPeopleGuestsJobsNewsEventsBlogPodCredNetworksSign in
Join nowSign in
  • Shows
  • People
  • Jobs
  • News
  • More
    • Guests
    • Events
    • Blog
    • PodCred
    • Networks
    Join nowSign in

ADSP: Algorithms + Data Structures = Programs

ADSP: Algorithms + Data Structures = Programs artwork

Episode 165: 2023 Retro! LLMs, Rust vs C++ and More!

Jan 19, 202435 minAudio

Link to Episode 165 on Website Discuss this episode, leave a comment, or ask a question (on GitHub) Twitter ADSP: The Podcast Conor Hoekstra Bryce Adelstein Lelbach Show Notes Date Recorded: 2024-01-18 Date Released: 2024-01-19 ADSP Episode 164: Are We Going to Run Out of Credit Cards? ADSP Episode 163: Sean Parent on Tilt Five, Metrowerks & Be Inc. Acquired Podcast S10E5: Nvidia Part I: The GPU Company (1993-2006) Acquired Podcast S10E6: Nvidia Part II: The Machine Learning Company (2006-2022) Acquired Podcast S13E3: Nvidia Part III: The Dawn of the AI Era (2022-2023) Acquired Podcast: NVIDIA CEO Jensen Huang ADSP Episode 150: Is C++ Dying? Rust Programming Language Hylo Programming Language Circle C++ Compiler Carbon Programming Language MLIR Mojo Go C++ Community Survey LambdaDays Run For the Fun of It Episode 14: YΓΌBaΓ­ Joins BlackToe πŸ₯³, 5k & Half PBs πŸ”₯ & 2024 Goals 🎯 CppNorth Intro Song Info Miss You by Sarah Jansen https://soundcloud.com/sarahjansenmusic Creative Commons - Attribution 3.0 Unported - CC BY 3.0 Free Download / Stream: http://bit.ly/l-miss-you Music promoted by Audio Library https://youtu.be/iYYxnasvfx8

Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.

Transcript

From the show. Tap a time to play from there.

ConorSo, the point being, you were right, Bryce. I was wrong.

BryceI'm gonna, I'm gonna need, I'm gonna need a clip of of of that little part of the conversation.

ConorWhat the you were right, I was wrong.

BryceYes.

ConorI'll put that as the cold open. How about that? So uh episode 165, you'll be able to go back to at any point you want.

BrycePerfect.

ConorWelcome to ADSP the podcast, episode 165, recorded on January 18th, 2024. My name is Connor, and today with my co-host Bryce, we look back at some of the highlights of the year 2023 in our annual retro LLM's Rust vs C and we also look ahead to the year 2024.

BryceI did I did call you with the intention of start early, end early. So let's get this going.

ConorAll right, 424, folks. We're gonna time box this to 30 minutes, so we will be done five minutes before the hour. Welcome to Happy New Year. Because I don't, I'm not actually even sure. You so quickly were eager to ask me your credit card number question last time. I don't think we exchanged any pleasantries, any Happy New Year's.

BryceHappy New Year.

ConorHappy New Year. Hope I mean, we did chat a little bit afterwards, but anyways, none of that made it because I mentioned on the last edit of episode 164, Microsoft screwed us two different ways. A bit of it's my fault. I think I actually do have Audacity set up to back things up. I went and checked it, but it did not back things up when the computer restarted.

BryceI was just reading a thread about that, about Audacity uh not backing things up when Windows does forced updates because I was trying to figure out how to prevent it from happening to me.

ConorYeah. And and the thing is, like 75% of the time when that has happened, it's only happened a handful of times, my uh recordings are recovered. So they typically are recovered. But in this one case, I had three different Audacity projects. Technically, one of them was outtakes from our conversation with Sean, where we were talking about an individual in the C community having gone to a new employer, and we had to cut all that out. And I was gonna insert it in one of these episodes at some point, but I'll have to go back and find that if we're going to do that.

ConorAnyways, point is Microsoft, Windows restarted, then Teams crashed. We're still using Teams. Why are we still using Teams, Bryce? It's a good question. Uh because we're cheap and we don't have another alternative.

BryceLook, you know, you know the reason that we used Zoom? Because at the time I was an officer of an ISO committee, and so I had a Zoom account paid for by ISO, and so I just used that for everything. But now I don't have any paid Zoom account, so there's we're limited to an hour or 40 minutes or something when we have more than two people. And so yeah, we're we're cheap. Now we're using Teams because that's what NVIDIA uses, and so once again, we're we're cheap. I'm cheap at least.

ConorYeah, I agree. I mean, we we at some point we we're gonna get merch. 2024 is the year of merch. We'll have I will have stickers by the time I'm going to conferences. Here's a quick question though. Take your mic. The reason I bring this up is because I mentioned this at the top of one episode 164. This was uh an edit that I added in. So Bryce did not hear this, but I recently came across my new favorite podcast called Acquired, also known as Acquired FM. And I saw you tweet about that. They did four different episodes, three specifically on NVIDIA, and then they did an interview with Jensen.

ConorIt is amazing, folks. I already plugged it once. Even if you don't like podcasts, which now that I'm saying that out loud, you're clearly listening to ours. So you must have at least some familiarity with them or desire to listen to them. It's so good. And the reason I bring it up is not to plug it again. But their research and their auto quality and their production value is just so good. It makes me sad about my podcast. Because my audio is good, but Bryce's, you know, Bryce is all over the place.

BryceBut hey, hey, hey, we did now I got this thing.

ConorIt's good. It's pretty good now, but that's why I was asking because it did sound a little tinny. But, anyways, we're doing our best here, folks, and long-term goals. This podcast is gonna be going on for decades. We are going to, we're gonna find, you know, the first class, world-class experience. We're still we're not gonna take ourselves as seriously as acquired takes themselves because I will, and this is a great way to transition, I will, I I have misunderstood the value or what the purpose of our podcast is. I you know, I love the motto, chaos with sprinkles of information.

ConorHowever, I thought it was a little bit too chaotic last episode. A little too chaotic. And I asked for feedback. I asked the listeners, please go to Twitter or to GitHub, and we got an overwhelming amount of feedback. Ten different people, eight different comments, and I will stand corrected. Bryce was overwhelmingly correct. I thought it was our worst episode ever. I will read from a few of the uh comments that were left on the GitHub discussion from Pascal754.

ConorActually, let's see if I click on the profile. Do I get a full name? I do not. Although they say they're from South Korea, so shout out to our South Korean listeners. This episode was hilarious. Uh, and then they linked to a problem from a Harvard course. Uh, if we scroll down a little bit more, I'm not gonna read them all. Kent F. Let's double-click on their profile. This person is uh located in Oslo, according to GitHub. Could be right, could be wrong. Kent F. So Kent says, Connor, sadly, I have to agree with Bryce, this episode was gold. As they say about streaming services slash cable TV, come for the movies, stay for the TV shows.

ConorYour podcast is come for the algorithm, stay for the banter. A discussion with Sean Parent. Episodes are always platinum grade content. Thanks. The next one says from Pet Hill. Where's Pet Hill tuning in from? No information, but the full name is Peter Hilgenfeld. And uh they mentioned this was fun. The high-pitched forwarding noise really, really hurts my ears. Please stop it. And I have to apologize because I usually only 10x Bryce, but when we went from minute four to minute 17 where my audio was still back online, I had to, I think it was 50x it, and that really made it really went high squeaky at that point.

ConorUm, we've got another one from Prof RS. Prof. I'm gonna mispronounce this, Scabarassi. Skyabarassi? I apologize, Robert. From Valhalla, New York. Which I actually don't, where's Valhalla? Is that close to New York, New York? Or do you know where that is?

BryceUh you know, all those towns that have names like Valhalla are like real, real, real deep upstate.

ConorI love how we're just is this is this isn't technically considered doxing, right? If this info is on their GitHub profile, what does mine say?

BryceSo Val Valhalla is actually um it's not that far north. It's um I think it's like Westchester County. It's just north of the city.

ConorThat definitely sounds right because uh it looks like they work for, or at least it says next to a little building icon, Westchester Community College. And anyways, yeah, mine says Toronto, Ontario, so I don't think this is doxing. We apologize if people wanted to stay anonymous.

BryceUm but hey, if you're the listener who lives in Valhalla, like you know, we can get coffee sometimes. We live nearby.

ConorBut yes, this individual, if I can get back to the comments, said Bryce, it will definitely be one of my favorite episodes of the year. And to Connor, listenable content-wise, it was awful. However, the 90 seconds of 10x audio was outstandingly hilarious. You even 10x'd yourself. I did. Uh apologize to those whose comments I didn't read. We appreciate all of them, but the point is, overwhelmingly, uh the listeners gave feedback and I was I was wrong. I thought it was a terrible episode. And the listeners, even though maybe they didn't learn a ton, they were clearly entertained.

ConorSo the point being, you were right, Bryce, I was wrong. And uh we're not yeah, and to get back to acquired, we're not gonna go as professional as them, but we're gonna try and give you the best.

BryceI'm gonna I'm gonna need I'm gonna need a clip of of of that little part of the conversation.

ConorWhat the you were right, I was wrong. Yes. I'll put that as the cold open. How about that? So uh episode 165, you'll be able to go back to at any point you want. Anyways, that's I think like almost 10 minutes out of the way. Today's episode is a re-recording of what we already recorded, but Windows destroyed, Microsoft destroyed, and it is the 2023 recap and 2024 look ahead. We've already done this once, so will it be as amazing as it was before? Nobody knows. We're gonna throw it over to Bryce though first.

ConorWhat were your highlights, top five, top three? I mean, I already asked you this, and you didn't give me a top five, top three. Will he this time? Will he not? Go ahead. Go ahead, Bryce. Tell us, reflect back on 2023. What are you what are what stands out to you?

BryceSo I'm not gonna give him any particular order, but I'll say to me it was the the year of large language models. You know, huge theme throughout the tech industry. I think it's gonna be a transformative thing that we'll see slow adoption of over the next five to ten years. So I don't think there'll be as much excitement in the next few years because uh, you know, it's a it's an exciting new technology, but as I always like to say, momentum is king in uh the software industry, so it takes a long time for things to be adopted. It 2023 was also the year of C23.

BryceGreat standard. Um really proud of the work that uh that we did on that one. You know, even even though I'm no longer uh chairing on the C committee, I'm uh really proud of the team that I worked with when I was there. Uh I had a great time. I think we produced a great standard. The I this time I get to mention the thing that I uh left out the the last uh time we did this, which was of course a huge highlight for me, was the ADSP Slovenia road trip, which I thought was just just a blast.

BryceAnd I, you know, we got to do one of these every year. Um, you know, being out on the road, going to visit. It combines my love of travels. Travel with uh my love of talking shit about programming languages. And let's see, what other highlights come to mind?

ConorI'm trying to think. I think that was mostly everything that you said last time. LOL, C. Oh, yeah, I think I asked you the question. So this is the hard thing, is I've already asked and have had these answered, but the listener hasn't. What you know, at one point, episode 150 probably will go down as our most popular episode of 2023. That was uh is C dying? But you've kind of flip-flop when we talk to Zach. Sounds like Reflection is making it now, senders and receivers. What is your current, what is the state of the Bryce's opinion on successor languages in C even a broken clock is right two times a day.

BryceUm reflection is great. I'm super excited about it. I really hope it makes it into C26. I think it'll be a big win for the language. Even if Reflection makes it in, that that does not, that's not a vindication of the broken standardization process. You know that that that the there's there are underlying issues in not C ⁇ itself necessarily. Like set aside that. Set aside whether issues there are with C ⁇ . Um and set aside like the question of how good C ⁇ is today.

BryceIt's a question of acceleration, not velocity. It's a question of where is C ⁇ going and how quickly can it can it can it get there? And the evolution process is just simply not sufficient. It's broken in many ways, and without real substantial reform, I think that C is a very questionable future. And I've spent uh a good seven, eight years of my career um trying to bring about that reform, and I'm proud of the changes I was able to make, but it's just simply not enough.

BryceSo I I still think that C is in trouble. But back to 2023, I think that 2023 is in many ways the year of Rust. I think it's really been a year that's been a focal point in Rust going from being a challenger to becoming more um established and more um of a uh a part of the um uh the an incumbent. You know, it's gone from being the the challenger to p potentially the challenged.

BryceYou know, now we see Rust in the Linux kernel. We're increasingly seeing Rust used in a wide number of projects. Um uh even within my own work, I hear more and more people talking about Rust, interested in Rust. Um I think it's become clear that Rust has succeeded in uh the space of next generation systems software languages. And I I've always said for a long time that that to me, almost one of the least interesting things about Rust is the memory safety aspect.

BryceUm uh and I mean I I I think it's a compelling part of it, but that's not what interests me about Rust. What interests me about Rust is uh a systems programming language with a substantially better package management and build ecosystem and a better process and model for evolving the language. Um, obviously there's been a lot of drama and controversy in the Rust leadership over the past year or two, uh, but I do think that that's present in any uh community and environment. And I am hopeful that those are just growing pains. Uh, because I think Rust's uh, at least until recent drama, Rust's community and evolution process has been substantially better than a lot of established languages.

BryceAnd um uh I had a lot of hope that it would um uh uh become a platform for the next generation of systems programming languages and then the next place that we can um uh all build our system software around.

ConorSo any predictions in that space, Rust versus C ⁇ versus other Langs over 2024? Or do you think it's just gonna be more the same Rust on the slow rise and C ⁇ on the slow decline?

BryceI I think that what we're going to start to see with Rust is that we are going to start to see Rust experience and overcome various growing pains. You know, I've already heard of some growing pains with cargo and with that ecosystem. I'm aware of some other uh growing pains, but I think that as as Rust starts to get more and more up to scale, and now that it has to serve a broader and broader set of um users, including now, you know, kernel developers, uh and I don't say that as if like the kernel, like being in the Linux kernel is like a game changer thing, but but I say that as an example of like there's there's there's a hundred use cases like the Linux kernel that have very specialized requirements that are now using Rust that weren't using Rust a year ago.

BryceAnd the more and more of those users that have very specialized requirements that Rust adopts, the harder and harder it'll be to uh make changes and to evolve. And the real test for Rust is can it can it um maintain its its agility and velocity as a language as it builds up a real user base and as it starts to build up um a code legacy? You know, uh Titus Winters loves to say that uh uh like software engineering is is like programming like uh uh integrated over time or something like that.

BryceYou know, it it's not just about writing code, but it's about how that code is maintained over the uh a long period of time. Um and those sorts of software engineering challenges, Rust is gonna increasingly start to face those. And I and I do think that this is gonna be a year where we will start to see more and more of those growing pains as Rust increasingly is adopted in more and more places.

ConorIt'll be very exciting to see this unfold. I mean, it's it's already been exciting over the last year, two years, especially with carbon, hilo, circle. So much going on in space.

BryceUh I do also think that another area where we're going to see a lot of interesting developments um in this year is the continued growth of MLIR, which is a um a higher level um uh uh compiler representation that's part of the LVM project that can be used to build domain-specific um compiler representations. And I think that over the next few years, we're going to see more and more compilers start to use and adopt MLIR and build all sorts of programming language and domain-specific optimizations and pipelines with MLIR.

BryceAnd I think in many ways, MLIR may democratize compiler technology. So a lot of things that get done today in libraries um or in sort of like bespoke weird, quirky code generators or pseudo-compilers will start to move more and more to um MLIR. Interesting.

ConorDo you think uh Mojo, because Mojo's being built on top of MLIR as well, do you think that'll see any success, or do you have thoughts there?

BryceI mean, I wouldn't bet against Chris Lattner.

ConorYeah, that's true. He he has, for those that don't know, has had a number of very successful projects, including LLVM, Clang, Swift, etc.

BryceAnd he is the person behind Mojo. I I think that it's gonna be hard for Mojo to succeed, though, because um it's not, at least from what I see, it's not a community effort. It's very much their thing. And they're the the company behind Mojo is, uh if I uh understand correctly, they're a hardware startup, more or less, yes.

ConorI actually am not I have not familiarized myself with their business model. I've just seen the talks that they've they've pitched.

BryceHang on, I'm looking right now. It's uh uh I mean What hardware are they selling? Um it's the it's the company's called Modular. I mean, even Wikipedia call not that's that's rough. Wikipedia calls it a proprietary programming language. Even if it is, that's a little bit rough to have that be your first line on um uh on uh on Wikipedia. Um uh modular, I think they do like AI infrastructure and platform.

BryceI don't know if they have hardware, but I for some reason I thought that they did. Um maybe it's maybe it's hardware agnostic. Um but they it is still it's not designed as a at least my impression is that it's not an open thing where they're inviting you know other contributors and other companies to participate in the development of it yet. Um and I think until that happens, um uh it looks like they don't make their own hardware, but but that they're it's just like uh an AI platform thingy.

BryceUm I I think almost all languages have to go through this test of where they they need to see whether they can gain broader adoption than just their one initial champion. Um if it's just one person, then it looks like they may not have enough steam behind them. And if it's just one company, then people are always gonna be worried about like, oh, is just this just their company, this one company's thing, like, oh, is this just Google's language or Facebook's language? Um and you can't really become a successful language until you grow beyond that. Um, you know, for Rust, it it was originally like this Mozilla thing, um, and it took a few years to grow beyond that.

BryceBut but do you think of Rust as Mozilla's language now?

ConorYeah, not really, no. It's yeah.

BryceUm in fact it w when I talk about uh Rust, when we talked about Rust, I don't even think it's come up that that Rust was developed at Mozilla Research. I I I I bet if you search through every previous episode of this podcast, I don't even think we've mentioned it that that Rust started at Mozilla.

ConorIt might have been it might have been brought up on the episodes where we interviewed Jane.

BryceOh, yeah, yeah, that may have come up there. Um but but uh I the the fact that I can't like that neither of us can remember whether we talked about its origins, that that speaks well for it. You know, Rust is used by a variety of companies now and it's got a foundation, and it's not just you know, it's not just uh you know one company's thing. Um and a language to be truly successful has to get grow beyond becoming one company's thing.

ConorI mean, unl unless you're Java and Oracle, in which case, you know, you can just it'd be interesting to get, because I know we've got plans not to spoil our, although, you know, spoil is generous because we've we've had these people on our list for like uh a year or plus now. We plan to get uh you know Andrew Kelly of Zig on, some Russ folks on. Might be interesting to get uh some Go folks on. I mean, even Rob Pike, because Go is a fascinating language in that, yes, it was championed by Google, but it's definitely grown to a point where like it's not identified as like the Google language anymore.

BryceLike that's I I I do still kind of think of it as the Google language.

ConorUm I mean it it's the origin. The origins were in Google, but like it's it's a language that just entered uh the top ten on some programming language rankings, and it's it's like ranks above Rust, but you don't hear nearly as much buzz or hype about Go uh than you do, or at least I don't, compared to Rust. And yeah, I find it curious. And also it's a language that as like a programming language enthusiast, like no PL theory people or PL, you know, that that kind of space academics, they're not really big fans of it because it didn't do anything really interesting or exciting other than maybe like their concurrent.

ConorModel. And everything else was either like already done before, not novel, or like pretty boring. And the fact that they didn't have generics until like 1.18, you know, a lot of PL enthusiasts were just like, well, the language is done. Like it like to I'm not going to use it if it doesn't have that. Anyways, it's just interesting that like a language that compared to many other languages is kind of boring and simple, has done even better than like a language like Rust.

BryceIt would be interesting to talk about and to look at Go because uh, you know, our colleague Andre, he likes to say that you know a language has to be 10x better than the competition in some way, shape, or form to succeed. And so there must be some 10x aspect to Go that's led to its success.

ConorI feel like maybe it's like a bunch of like two to five X improvements on a bunch of small things, which like totals, because like its compile times are incredibly fast. It produces like a single executable, like you don't have to link stuff. It's like it's just much simpler. All the code looks the same. Anyways, we'll get a GoExpert on. We're running short on time. We've got less than 10 minutes left. Uh any other 2023?

BryceI I think it's gotta have some I don't think that a bunch of two X's to five X's are um uh uh is equal to a 10x in even one dimension. So I just linked you um uh an an interesting thing about 2023, which is it just came out the uh JetBrain's annual C ecosystem survey, which I consider the the best survey of the C ecosystem and community. And uh I don't know, maybe we can look at it and see if there's some interesting trends going on.

ConorI read it and it was kind of boring.

BryceThanks, buddy. You know, I was one of the three people that they asked for feedback on it.

ConorI mean, like, what is what is there to say? You know, people are moving to you know more recent standards, python's the next most popular language that c devs know. Uh VS Code is the most popular after that.

BryceIt's it's shocking to see a drop in the the survey numbers of uh of CMake, although I suggest that means that there's just a margin of error in that CMake has saturated at about 50% market share.

ConorYeah. It's hard to look into these numbers, you know. Anything you want to remark on it?

BryceUm well, if I remember my let me see if I can remember my remarks from when I uh that I originally sent to them. So I I I mentioned that that there's uh we we saw a large jump in adoption of C20, which is which isn't shocking because you know it usually takes two to three years for implementations to really um uh pick up the the the latest you know standard. Um and I think this is not just a C ⁇ thing. I think it's true that like for almost anything, it takes a while for uh uh adoption to really pick up.

BryceUm uh you know, once once you put out a new thing, it takes a while for it to saturate in the in the ecosystem. But but you know, C20 um usage going from 23 to 29%, that's a pretty big jump, uh especially relative to the other uh uh dialects. There was I was I was surprised to see that that almost 50% of people plan to use modules in the next 12 months, um, which is interesting because that's one feature where we haven't seen a ton of adoption yet. But um just recently, CMake launched their support for C modules in a non-experimental way.

BryceIt had been experimental for a while, but now it's fully baked. And so maybe we will see a lot of adoption of modules in the next year. Um in particular, we may see a lot of adoption of using the standard library module. Um, even if you're not making your own library uh and your own code into modules, you might just use the module from your standard library, the the std module, um, which can give you some significant uh compilation uh time wins. Um so I think that that's the the the first way that we'll see people using modules is not providing their own modules, but just using the std module.

ConorYeah, I've tried every four months for like the last couple years, because I've seen talks and I've seen demos and I've seen articles about how it's usable. And I've even seen like someone with a Microsoft Visual Studio use import std, and it just never works. Like uh they've always you gotta always check a bunch of boxes and experimental, blah, blah, blah. And the thing is is I think a lot of these they ship with some you know starter experience for modules, but like standardization of these uh standard headers or whatever, that's a C23 feature, which as far as I know, no library has done the work for yet.

ConorAnyways, hopefully people can start to use modules.

BryceUh and now I got I got two or three more interesting things here. So so one, um, you know, survey results indicate people still not using package managers. Um very little adoption, only like uh 11% adoption of VC package, which is the most successful one. Um, and the other ones account for like maybe 12% uh combined, maybe a little bit more than that, but only like 25% of people are using um some form of uh of package manager. And uh, you know, that that doesn't that doesn't necessarily surprise me.

BryceUm, but uh I wonder whether whether we'll start to see more and more adoption and more uh centralization around one of them, like in particular, Fece Package seems to be the leader. Um and in terms of IDEs, this is the first year of the survey where Visual Studio has been tied with uh uh another IDE for the top spot. Usually Visual Studios shows up as the most popular IDE, but this year it's uh it's tied with the the JetBrain C lion IDE, uh, which is pretty interesting because uh you look at back when uh uh C Lion came out in the market, and really there was only just one IDE, uh Visual Studio, and then I guess if you were on Mac, there was Xcode.

BryceUm but uh but now we're seeing uh a lot more diversity in uh developer environments for uh C. I think that's pretty much it from the survey. Um but uh I do think that there's some interesting, interesting insights to be gleamed from it. Yeah.

ConorLink will be in the show notes if you want to read it yourself. All right, we've got three minutes left, four minutes left.

BrycePredictions for 2024.

ConorYeah. So I well, not predictions. I gotta do my recap of uh 2023. Uh I mean, technically, you already heard it. Last time we recorded, but then we lost it. Anyways, I'll rattle through them quickly. I'm not even sure if this is the right order, because I had it down before and then deleted it. We'll come in with number five, which was hanging out with Sean Parent at C on C back at the tail end of June. That turned into three different episodes, was an absolute blast when Sean was on. We made a reference to it. Hopefully, we'll get to do something like that similar at a different meetup or conference in 2024.

ConorNumber four is the two top conferences, in my opinion, for the year were KXCon, which is the conference put on by First Derivatives that focuses around the Q language, and Lambdays, the functional programming plus plus conference that takes place in Krakow, Poland. Hope to go back to both of these conferences in 2024, but time will tell if that works out with my schedule. Number three, similar to what Bryce said, Slovenia 2023 trip. I still have fond memories thinking about wandering through the streets of Venice, recording podcasts, getting donuts, eating snacks.

ConorFantastic. Number two was not tech related. It was all the running. Check out the running podcast, link in the show notes. We're not going to say more about it here. And number one is to be announced at a later date in 2024. All right. Any anything else? Well, we got like one minute left. 2024 predictions, Bryce, go.

BryceUm, 2024 predictions. Didn't we already do these a little bit? Well, okay, so so so one prediction. I don't think that there will be as much exciting large language model news uh this year as there was last year. Um, because last year was like the the the you know the big hype curve, and now we've got the slow road to graduate to adoption. I think that there will be breakthroughs and developments, um, but I don't think that there will be quite as much excitement as there has been. I think that we're on like a five to ten year road of productization.

BryceUm I think that reflection will probably make significant progress in C. Um uh I'm hoping to get more involved in the Rust world uh this year. I may try to go to RustConf um and uh uh doing some other like compiler things that are pretty exciting that I don't think I can share many more details about other than that they're exciting. Um and uh I'm I'm going well I'm going to Ireland with my mother for uh uh right before uh ACCU in a couple months, so we're excited about that.

BryceUm if you live, if you happen to live anywhere between uh uh Dublin and Dingle, which is as far as I can tell on the westernmost side of Ireland. Um and you want to hang out, my mom and I are gonna be driving from Dingle back to Dublin because we have a friend who lives in Dingle, and the most convenient way to get from there to our flight is to drive. So we're gonna have a little bit of an Irish road trip. Um so if anybody wants to hang out, meet my mother, let me know.

ConorAny other conferences? So you just you just alluded to the fact that you're going to ACCU. Any other conferences that you're planning at the moment or T TBD?

BryceYeah, I submitted to CPP North, but they were they didn't accept me last year, so who knows? I may not be good enough for them. Um I I may go to C now. Cpp on C and may go to the dates are a little tricky for me. Let's see, other conferences. Uh I'm actually I'm still in the process of putting together list. I'll probably be a code dive. Um uh and I'll be a core C. And probably a couple other ones thrown in there.

ConorAwesome. I think, yeah, of the ones that you said, I've submitted to CPP North. Yeah, announcement if you didn't know. CPP North's call for proposals is open until I think mid-February.

BryceUm, but don't but nobody submit because I want my talk to be accepted.

ConorDefinitely submit if Bryce can't, you know, get his talk sub uh accepted. That just gives us fodder for the podcast. Yeah. And uh what are the other ones? I submitted to Lambda Days, KXCon's details isn't out yet. Uh PLDI, which is gonna be in Copenhagen. If there is going to be an array workshop, I would like to submit a paper there. But they have not posted on the PLDI website whether or not that's gonna happen or not yet. That's in June. And I think, oh yeah, PyCon, I submitted a talk to.

ConorAnd also I submitted my talk to an NDC, NDC conference. So yeah, so I don't think I'll get accepted to all of them, but I'll definitely be once once we're accepted, if we're accepted, we will let the listeners know. And a promise to close out episode 165, recap of 2023, a look ahead to 2024. We will have merch, well, not like purchasable merch. Maybe we will, but we will definitely have stickers. For those of you that have come up and said hi and said kind things about the podcast at conferences in 2023, we didn't have stickers.

ConorThis year is it's an ADSP promise. We're going to have ADSP stickers. So be sure to ask if you bump into us at one of the conferences and you'd like one. Any other closing thoughts?

BryceSo wait, so wait, now I have to figure out how to make room in my standard luggage for stickers.

ConorAll right, I'm gonna rephrase what I just said. I, Connor Hookstra, co-host of ADSP, will have ADSP stickers.

BryceI will have ADSP stickers as well.

ConorIf you would like one, come up to me. Bryce may or may not. It is unclear depending on how packed his uh his his his luggages are. All right, folks. Episode 166. Are we gonna have a guest? Maybe, maybe not. We'll figure that out in between now and the next time when we record. Be sure to check the show notes either in your podcast app or at adspthepodcast.com for links to anything we mentioned in the episode, as well as a link to the GitHub discussion where you can leave thoughts, comments, and questions. Thanks for listening. We hope you enjoyed and have a great day.

BryceLow quality, high quality. That is the tagline of our podcast.

ConorThat's not the tagline. Our tagline is chaos with sprinkles of information.

Open the transcript file

Hosts and Guests

Nobody is credited on this episode yet.

Credit the people on this episode

Be the first to rate this episode

More episodes

All episodes
  1. Episode 313: C++26 Reflection vs CRTP, AI, Kids & More with Zach Laine

    Nov 20, 202647 minAudio

    In this episode, Conor and Bryce chat with Zach Laine about C++26 Reflection vs CRTP, AI, kids & more. Link to Episode 313 on Website Discuss this episode, leave a comment, or ask a question (on GitHub) Socials ADSP:...

    • Apple Podcasts

    Credit the people on this episode

  2. Episode 312: The C++26 Contracts Fiasco

    Nov 13, 202652 minAudio

    In this episode, Conor and Bryce chat with Zach Laine about the C++26 contracts fiasco. Link to Episode 312 on Website Discuss this episode, leave a comment, or ask a question (on GitHub) Socials ADSP: The Podcast:...

    • Apple Podcasts

    Credit the people on this episode

  3. Episode 311: Career Changes, GPUs & More with Zach Laine

    Nov 6, 202628 minAudio

    In this episode, Conor and Bryce talk with Zach Laine about his career and the career changes he has made. Link to Episode 311 on Website Discuss this episode, leave a comment, or ask a question (on GitHub) Socials...

    • Apple Podcasts

    Credit the people on this episode

  4. Episode 310: Leaving NVIDIA for... ?

    Oct 30, 202624 minAudio

    In this episode, Conor and Bryce talk about Bryce's next chapter and where he is leaving NVIDIA for. Link to Episode 310 on Website Discuss this episode, leave a comment, or ask a question (on GitHub) Socials ADSP: The...

    • Apple Podcasts

    Credit the people on this episode

  5. Episode 309: Default Construction & CppCon 2026 with Jason Turner

    Oct 23, 202629 minAudio

    In this episode, Conor and Ben chat with Jason Turner about default construction in C++. Link to Episode 309 on Website Discuss this episode, leave a comment, or ask a question (on GitHub) Socials ADSP: The Podcast:...

    • Apple Podcasts

    Credit the people on this episode