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 125: NanoRange with Tristan Brindle

Apr 14, 202328 minAudio

In this episode, Conor and Bryce chat with Tristan Brindle about collection oriented programming and NanoRange. Link to Episode 125 on Website Discuss this episode, leave a comment, or ask a question (on GitHub) Twitter ADSP: The Podcast Conor Hoekstra Bryce Adelstein Lelbach About the Guest Tristan Brindle a freelance programmer and trainer based in London, mostly focussing on C++. He is a member of the UK national body (BSI) and ISO WG21. Occasionally I can be found at C++ conferences. He is also a director of C++ London Uni, a not-for-profit organisation offering free beginner programming classes in London and online. He has a few fun projects on GitHub that you can find out about here. Show Notes Date Recorded: 2023-04-05 Date Released: 2023-04-14 C++ On Sea Conference ACCU Conference cpp.chat Podcast CppCast Podcast Phil Nash on Twitter CppCast Episode on C++ Uni (with Tristan Brindle) C++ London Meetup Flux Flow D Ranges C++ Ranges Rust Iterators Collection Oriented Programming Functional Programming Smalltalk Java 8 Streams SETL Thrust Library cuCollections Swift Sequences Swift Collections Ranges TS Range-v3 Library NanoRange Conquering C++20 Ranges - Tristan Brindle - CppCon 2021 CppCon 2015: Eric Niebler “Ranges for the Standard Library” What a View! Building Your Own (Lazy) Range Adaptors (part 1 of 2) - Chris Di Bella - CppCon 2019 What a View! Building Your Own (Lazy) Range Adaptors (part 2 of 2) - Chris Di Bella - CppCon 2019 cmcstl2 Intro Song Info Miss You by Sarah Jansen

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.

Tristan BrindleI was talking talking to him about this once, and he was like, that is the definition of feature creep, right there. Like you started off wanting to call sort, and you've ended up with 18,000 lines of code that that implements the whole of the range of CS.

ConorWelcome to ABSP the podcast episode 125, recorded on April 5th, 2023. My name is Connor, and today with my co-host Bryce, we interview Tristan Brindle about collection-oriented programming, his NanoRange library, and more. Alright, we will just hop straight into things. Are you in London?

Tristan BrindleI am. Well, actually a little town just on the outskirts of London.

ConorAnd Bryce, are you going to C on C?

BryceYeah. Yeah. I'm going to London three times in the next. Wait. Three times in the next couple months.

ConorI'll be in London for like a day before. Are you going to C on C too, Tristan? I am. I am. Okay, well, we'll see. We'll see each other there then. So I was going to say we should meet up in London.

BryceThis was dangerous information to give me because now I have another name of a person whose couch I can try to crash on every time I come through London.

Tristan BrindleUm yeah, you can try.

BryceYeah, I'm going to be there for um on my way to Core C for CPP on C, and then in two weeks for ACCU.

ConorWhy do you count your layovers?

BryceThat's not like in this case, it's I'm actually there for a day. I fly in and I have to No, that was cool.

Tristan BrindleSo yeah, this is the problem. I don't have any cool stories like Sean, obviously. Um I've never met Steve Jobs.

ConorUm I was never in the army, so I just I don't know where my stories are gonna come from, but we'll uh I feel like you probably everyone, I think that's the thing is everyone doesn't think their stories are like amazing or awesome.

BryceYeah, but everybody has.

ConorAnd then they will just like incidentally tell someone some story, and then you'll be like, what the heck? You that and that happened?

BryceAnd you'll be like, oh, yeah, I mean, I guess I seemed normal at the time, but I guess from a third point party point of view, uh I think it's pretty rare to meet people who are actually boring. Like I I re I really struggle to think of somebody that I know who does not have interesting facets.

Tristan BrindleYou've met me a couple of times.

BryceSo and I think you're an interesting person. I'm sure I'm sure.

Tristan BrindleYou did invite me onto this for some reason. Yeah, yeah, sure. You're under the spotlight.

ConorI know, I know. That's the problem. Yeah, we are uh we are an ever-growing, we're the fastest growing, and we're we're now officially oh we never brought that up. So for the longest time, depending on what podcast app statistics you look at, we were behind CPP chat, which is and now discontinued, I think forever discontinued. I actually don't know the status. Phil Nash is is always too busy skipping from hot podcast to podcast. And now that he's on CPP.

Tristan BrindleYeah, he's well, he's doing CPP cast now.

ConorSo he's like a pirate, he's like a podcast pirate, you know? He's on one podcast pirate ship, and then he and then he jumps to another podcast pirate ship, and then he saw this like slowly sinking CPP cast, uh, you know, pirate ship, and then he was like, I'm gonna take that back and bring it back to life.

Tristan BrindleUh well, uh he and Timo are doing a really good job with with CPP cast. I think I've uh really enjoyed their content since it's come back.

ConorYeah, they've had some they've had some interesting guests. Um but let's okay, now I feel like we're done warming up. Uh whether or not we'll see what makes it into the Go for it. Go for it into the actual final cut.

BryceYou you have a topic in mind, Connor, right?

ConorUh yeah, I well, I'm not sure. Did you see it? Well, first of all, let's introduce Tristan Brindle as our guest today. This uh interview is kind of happening because let me bring up the uh email. I did link to both of them. So how what was what's the date on this? February 9th, 2023. Uh Tristan, who I mean I I assume most listeners know that you know we know a lot of the people that we're bringing on, but we've all met obviously at multiple conferences and committee meetings.

ConorUm actually, yeah, when was the first time we met Tristan? Maybe it was CPPCon 2019? Probably.

Tristan BrindleThat I I was yeah, that was that sounds about right, yeah. Uh I don't rem I've never been to C now. I'd love to go, but uh yeah.

ConorSo uh prolific speaker, uh many topics. We'll link a bunch of them in the show notes. Uh I think ranges is primarily like the majority of the talks that I've seen by you.

Tristan BrindleYeah, yeah, that that's that's uh kind of the subject that I've I've given a few talks on.

ConorUh and actually a couple of conferences. Well, before we get back to the tweet that brought this interview/slash conversation to you today, I think the very first time I came across your name was you were on CPP cast as a one of three guests, correct me if I'm wrong, that was talking about the CPP London University. And uh I remember just being like in awe of the amount of work that it would take. Because you were saying that you basically were like marking assignments and like giving feedback and conducting this kind of like online C course.

ConorAnd I was just, I don't even think at the time I had connected when I first met you that you were one of those three voices. I knew you had done the um uh your sort of nano, was it nano ranges? Is that what it was called? Nano range, yeah, yeah. So I knew that you had implemented a bunch of ranges, like range like stuff in the range library that I think has been quite popular. And then at some point I I re remembered that you had mentioned that you were doing this CPP London University thing, and then I was like, I listened to that podcast, and then I went and looked, and I was like, holy smokes, you were one of those three voices.

ConorYeah, yeah.

Tristan BrindleSo um that was uh I think that I I started with uh well uh it was started by a guy called Tom, Tom Breza, who um it was an offshow. There is a uh C meetup in London that's called C well it's called uh C London. Uh it's run by Phil Nash, um and it's roughly monthly. Uh and uh a guy called Tom who was learning C started going along to these meetings, but you know, they're that they're for established programmers, you know, that's for people who use C all the time.

Tristan BrindleAnd he was a learner, and he was like, it would be really cool if there was some sort of meetup for people who were learning C. So Phil like emailed round everybody who was on the list for the for C London. He said, like, you know, Tom is trying to start this. Would anybody fancy uh like giving a like maybe a couple of classes? Um maybe like five classes. We maybe we can do it in five, maybe we can go through all of C in five like hour-long sessions.

Tristan BrindleI was like, well, I I I don't know about that, but that that sounds that sounds like fun. That sounds like something I could do. So then I kind of signed up. Okay, I'll do I'll do five sessions. Um and uh then very quickly another guy got involved uh called uh Oliver Din, Ollie, and uh so then the three of us, this just sort of uh just sort of snowballed from there, and we ended up uh giving these classes, uh weekly classes in C in London for uh a couple of years.

Tristan BrindleAnd then this was it was all going pretty well, and then unfortunately COVID hit and there was lockdown and we weren't able to meet up anymore. And we started doing them online, like right at the start of of all the COVID shutdown. But at that point, like everyone had massive Zoom fatigue, you know, because we were all spending so long all day uh in these Zoom meetings, and uh it just we so you know it it sort of tailed off because we're all we all kind of lost enthusiasm for for being on Zoom in a couple of hours in the evening.

Tristan BrindleUm so yeah, unfortunately, uh COVID sort of put paid to uh to to that. Um it would be nice to get it going again. Tom and Ollie, if you're listening to this, um get in contact because I'd I'd quite like to get it going again.

ConorYeah, I remember listening to that thinking that that's such a cool thing. And also just like I've run virtual meetups before, and virtual meetups are like a lot of work and just like preparing the content and the slides, and um, it's just a lot of work to run that. But on top of that, you were you were like doing marking, and like I was just like, wow, that's like sounds like a Heraculan, was it the word Heraculan effort? And uh Herculan, there we go. Thank you. Uh it didn't sound right. Uh but yeah, so that you know, if potentially connected some dots for some listeners that uh may know of Tristan's talks and have also listened to CPPcast for a long time.

ConorUm and you so you did nano ranges, you also were gonna this is gonna come up at some point, um, have a library that I don't think you've really formally announced at any point called Flux, but it's I've known about it for a while because you've I follow you on Twitter and you've tweeted a couple things, and then I've been like, What is this? And you were like, Oh, you know, it's my unannounced library. So I feel like I'm saying it's unannounced, and yet yet we have like several thousand people that listen to this podcast, so it's gonna get a this is the announcement! Yeah, here we are.

Tristan BrindleYes, yes. All right, let's um yeah, we can we can talk about that. So I've been um working on this uh for a few months now, working on uh a library called Flux. It's actually a follow-up, sort of a follow-up to another earlier library called Flow. Um, but uh so Flux is uh a library for sequence-oriented programming, I guess would be a good way to put it. Yeah, yeah. I don't mean in the in like the APL sense, like don't get too excited, Connor, because I like I know nothing about APL or BQN or any of those uh array languages.

Tristan BrindleWhat I mean is sequence-oriented programming in the same way that you would get with C ranges or Python iter tools or Rust iterators or you know, D ranges or or what have you. Um it's a library for doing that. Um it shares pretty much the same goals as C ranges, but it takes the things that I have learned from working on C ranges from inter uh from implementing C ranges, and it does some things slightly differently. Um and I think these are you know it's an experiment.

Tristan BrindleI don't know that it's uh you know it's gonna work. Um, but I I'm quite excited about it, and I think it's uh I think it's I think it's pretty good really.

ConorAll right, I've got one comment and one question. So the the name of the paradigm, the reason why I went is because I think I maybe mentioned it once or twice. There's a paradigm called collection-oriented programming, which is literally what you just said. Like it is ranges in C, iterators in Rust, and it's it's got some overlap with functional programming. Like I feel like spiritually, functional programming and collection-oriented programming are like cousins, but it's the things that I actually like previously associated with functional programming, like having functions that operate on sequences or you know, even in small talk.

ConorLike small talk is not a functional language, but it is a collection-oriented programming language because it has all these different containers that just have operations, aka what they call messages, that you basically just like you build up this kind of like data flow where you're just chaining a bunch of things together and you're never writing a for loop, you're never worrying about like indexing and stuff like that. You just sort of like build up this sequence of operations, which is exactly what C 20 and 23, 26 ranges and all these types of like Java streams fits in in this model. Anyways, there's gonna be a talk at some point that I'll call collection-oriented programming, but uh it's like my favorite way to program, whether that's in Haskell or in C Rangeland or in Smalltalk, or even in like I think there's a language called SETL C-E-T-L that I haven't looked at, but apparently it's like specifically kind of like the APL small talk kind of thing, but it purely works on like sap sets and maps.

ConorI could be wrong about that. But anyways, collection-oriented programming, fantastic.

Tristan BrindleYeah, yeah, yeah, yeah. So I I I agree with you. So I previously used the term like functional style programming. The trouble is like functional programming means a lot of things to different people, you know, referential transparency and yeah, all that sort of stuff. Uh whereas I I'm more focused on on like you say, the bits where you you have a sequence and you manipulate it in some way, you build up a pipeline of operations. And um, yeah. So we're not not kind of full-on Haskell style functional programming, but uh yeah.

ConorWe've got it. One of my goals over the next one or two years is to popularize this term collection-oriented programming because it is like the perfect description of what you just like described. And like it it like goes across like APL, small talk, functional languages. Like APL is kind of functional, but like the thing that I love about it is this composing operations or like a pipeline of operations, and then you just feed it your data and poof, you get your answer. Like, and that is my preferred way to program in general. Like in C that means like doing as much range-like stuff as possible, but uh in other languages it it just means like using the language as it's designed.

Tristan BrindleYeah, yeah, yeah, definitely. Okay, so I my my term was sequence-oriented programming. If collection-oriented programming is the uh is the the proper term, then I'll I'll start using that instead.

BryceWell, to me, collection suggests an unordered um set of things, and sequence is a thing that has how does collection mean unordered? We literally in this field, when I think collection um versus sequence, that's just what comes to mind to me.

ConorYou're forgetting, I don't know how long ago this was, probably a year or two ago. You made me, or you made me, you nerd-sniped me into doing an opposition analysis of what languages call data structures, because we were trying to figure out a name for the thrust equivalent of like a data structure library and NVIDIA. And we ended up coming across the terms basically containers, collections, data structures, and maybe there was one over one other less popular one. And we ended up going with collections, and that's why KUCO, Q collections, is called collections because it's um what many languages call their quote unquote data structures.

ConorSo I'm surprised that now your semantic attachment to collections is not just like data structures in general, but an uncle.

BryceI mean, I I think that that's that's one of the um semantic attachments that I have to it. But uh if I'm comparing the term with sequence, um the important property that I get out of um sequence uh is is that it's a it's a thing that I'm getting in in where where order is significant.

Tristan BrindleThat's true. So I I'll I'll add a data point on this, which is not something I know a huge amount about, but uh I know uh but uh I know that in Swift Um there are in the Swift standard library there are both sequences and collections. They have sort of two iteration APIs, and sequences are single pass and collections are multi-pass. So this was um as far as I know, uh Dave Abrahams uh came up with this, I believe. Um so he know he knows an awful lot about these sort of things.

Tristan BrindleSo uh yeah, so what sequences and collections.

ConorDo you have like an example of what it means to be a multi-pass on a collection? Like uh I think the sequence single pass thing makes pretty it's pretty straightforward. You're mapping over an operation, you do it whatever, left to right or in some sequential order.

BryceWhat what does it mean to I think it's just about capability, about whether you can consume it multiple times.

Tristan BrindleYeah, yeah. So the equivalent of a forward range in in C land.

ConorI see, I see. That makes sense. Interesting. Well, yeah, there is it. That is another data point then. Uh I don't not sure if Swift was a language we looked at of what they called their uh called their data structures. Alright, so the that was the comment. The follow- the question is let's talk about the the differences and everything you've learned. Uh well, so we'll link for uh you know the curious listeners, I'm pretty sure nanorange still lives online. Um even though it was kind of designed as like for people that are in previous versions that don't have access to C20, you can use this in the interim.

ConorSo I think if I recall it, it's usable in like C 14?

Tristan BrindleUh 17, 17 uh nanorange. Yeah, so just to rewind a bit, so back in the range's TS days, so uh before this was uh standardized, there was um I don't think we do them anymore, do we, Bryce?

BryceBut the uh like uh TSs were was this idea that we would we'd sort of try out some new uh libraries with the intention that they would eventually we've um we've stopped doing what I like to call grab bag TSs, which are TSs that are just a collection of things without really a specific purpose. And the the reason for that is that there is a um there's a high overhead to publishing something in a TS versus an IS. And if we publish it in a TS and then want to put it into the IS, we basically have to pay the standardization cost twice.

BryceUm, because oftentimes we have to go through and do a re-review of it because things have changed or evolved um since then. And so we we still we'll still pursue TSs, but um uh I think the the general vibe on the committee, at least in library evolution, is that we only want to see focused TSs. T S is where we have a very clear goal for why this is why we want to pay the additional cost of putting this out into a technical specification, um, what questions we're hoping that it's gonna answer.

BryceUm I mean, another reason that we've been become um more TS pessimistic is that um uh we have found that we don't really get the usage experience that uh in the deployment experience and the implementation experience that we we had hoped with um TSs, um especially because we are now shipping on a three-year cycle pretty reliably. Um, you know, TSs were something we came up with when we were in this long after after this long decade uh uh that led to C11, and we thought that we might it might be a while before we put out another standard revision, but now these days we reliably put them out every three years.

BryceUm and uh because of that, what often happens is we go through all the trouble to publish a TS and then almost immediately before it even really has any time in the wild, somebody um advocates for us putting it into the IS. Um and so we just we don't end up getting any experience from it, and um we just pay a lot of extra effort.

Tristan BrindleMakes sense. Um Yeah, so to so well ranges was I I think ranges was uh you know it was a pretty focused uh TS. So uh Eric Niebler came up with uh Range V3, uh and then this was this went through the the T S uh process to be uh standardized, eventually found its way into C20. And back at the time when the Rangers TS was published, I went and um I wrote a like a from scratch implementation of the TS for I started out with C 14 and then I I moved to 17 later on because it had some uh really useful features like ifcons, textprint stuff.

Tristan BrindleUm but yeah, so I I wrote basically my own implementation, it was called nano range. And the idea was that you could like as you say, Connie, that you could use uh get started with nano range, and then whenever C20 came along, uh you could you could switch to that without having to use and you know you you could use range v3, but range v3 is kind of huge. Um and so if you wanted something that was smaller, like a nano-sized, um, you could use nano range instead. Um so that um to be honest, it never quite got the UCs that I was hoping for.

Tristan BrindleI had a few people interested in it, and I believe I've seen somewhere that it's used in the Intel thread building blocks, or parts of the bits of bits of the nano range code are used in there, which I thought which I was I was quite pleased about. Um but uh perhaps that's not the case anymore. Uh but uh yeah, so there was so there was nano range, uh, and I I learned so much about uh ranges and um how all these things are implemented and how all these parts fit together uh from doing that.

ConorBefore we move on, uh uh how like when I think about implementing a TS, I don't know if it's my imposter syndrome, but like to me that seems like a a very like impenetrable wall. Like and I've I've also seen a lot of like the internal range v3 code. I've also watched your talks, I've watched Christabella's talks. Also, too, we should mention the Eric refers to Eric Niebler if if you're this is your oh sorry, if this is your first episode. I mean, we've had Eric on, so hopefully our our listeners know when we say Eric who Eric is.

ConorBut uh like so I've I've seen Eric's talks, I've seen your talks, I've seen Christabella's talks. Christabella has like a two-part CPP con talk where he walks you through how to implement like a stride um uh a stride range adapter, and just like you know, I I if I ever had the thought I've had the thought of like, oh, it would be cool to implement my own like mini ranges, like similar to what you did, like only what the standard says, not a bunch of action, extra stuff that's in range v3, and then immediately my brain is just like, yeah, probably. not like that would be so much work and like you'd you would immediately hit some like language lawyer wall of like having to digest like the wordies of the standard anyway so my question after that long unnecessary monologue is like what gave you the confidence or like and how difficult did it end up being having to implement that you know taking it from a TS to basically having your own library?

Tristan BrindleOkay so to start off with I did not uh by any means intend to start out and just like start off with an empty file and go, I'm gonna implement the ranges TS. What happened was that um I was using Range V3 and I was like this is really cool but like I've got to be honest like the compile times at that time were a real drag. And I was like what I really want is to be able to call STL algorithms but just pass the range. I don't so I don't have to call dot begin and dot end you know just sort brackets vec.

Tristan BrindleAnd I was like okay what I'm gonna do is I'm just gonna write a single header and it all it's gonna do is it's just gonna have some overloads of um you know all the well all the useful algorithms and you're just gonna be able to pass in a vector as well or pass in a range directly as well as these dot begins dot begin dot end and I'll just forward those calls to the um to the the the existing STL right um and then the thing is if you start doing that it turns out that quite quickly you'll find out that there will be ambiguous overloads so if you think about the example of sort uh you can pass a custom comparator right so if you have a call to sort that has two arguments you've got to have some way of deciding are these two things you've given it are they a begin and an end iterator or are they a range and a comparator?

Tristan BrindleRight? Right. So in order to do that you need to start using so well nowadays you could use concepts but uh we didn't have concepts yet then. So I was starting to do some metaprogramming and it's like how do I work out whether this object is an iterator? How do I work out whether this object looks like a range? I was like you know what some very smart people have already worked this out and it's all written down for me in this ranges TS document. So I started off like using enable if and all that horrible stuff we had to do at the time all this template metaprogramming to work out okay this looks like a valid range this looks like a valid iterator and you know it it it's really pretty useful as well because it it allows you to reject calls when uh you know iterators don't uh meet the requirements and all that so it I I didn't I really didn't start out intending to do the whole the whole TS I I was like it was just going to start out being really simple just what I needed and then um I was like okay well now I need to be able to differentiate between iterators and ranges and um all that sort of stuff so I'm going to going to take what it says in the ranges Ts and do that.

Tristan BrindleAnd I was like you know what this is really fun. I was in you know I I actually quite enjoy this so I'm I I just ended up doing the the whole thing. And the other thing that I I really couldn't have done it without the help of uh Casey Carter who is um one of the two sort of godfathers like the fathers the fathers of ranges along with Eric Nieber. And uh Casey was uh we were I was on Slack at the time the CPP Lank Slack pretty much every day just asking Casey the the dumbest questions and he was very very patient with me uh explaining stuff um we found a few bugs in the spec and uh yeah so um casey for listening thank you very much for your help uh at at that time I I couldn't have done it yeah he has his own version there's range v3 there's nano range and then there's also his which is like CM CMCSTL or something like that.

Tristan BrindleCMC L STL too yeah um which I I I think that that was sort of the uh so that used the GCC concepts TS implementation at the time.

ConorYeah. I think that's online somewhere so we'll I'll I'll link to that as well. And what I was going to say is that this is uh it's very inspiring to hear because as soon as you said oh the I only we just wanted to provide an overload for like existing algorithms is like that is all I want to do is call sort bracket adapt. That sounds like my brain when I think of like oh if I wanted to do that my brain goes oh yeah you can definitely provide just like an overload that dispatches to the existing thing. And then from there you basically just was like oh well I just I need to know this one little thing you went to the the TS and then things snowballed from there.

ConorAnd I could you I definitely know that like I've written things in the past where like the end thing that I ended up writing seems like a behemoth of like you know how would have I done this to begin with but like the initial thing that I wanted to do was just like build some small little stack machine for a small toy like calculator that ended up being you know something way way bigger. So I guess that I guess the word of advice for the listener here is that if you're ever intimidated by some ultimate thing you want to achieve just choose like a very small easy part that your brain knows that you can do and go from there because that seems like it's exactly what you did and you ended up with a whole implementation of the of the TS.

Tristan BrindleYeah I'd say Gashwell Altzman I don't think it's saying I was talking to about this once and he was like that is the definition of feature creep right there. You started off wanting to call sort and you've ended up with 1800 lines of code that that implements the whole of the range's TS. Yeah. So yeah. Not what I started out and you know intending to do but uh somehow somehow ended up with that. So yeah that that was but that was a few years ago now uh now we have uh ranges in um in the standard in the standard library as well it's uh you know and I and I I love ranges I love this collection orientated stuff um I I you know like you it's um I re I really enjoy that style of programming.

ConorBe sure to check your podcast app or adspthepodcast dot com for links to any of the topics or things that we talked about in this episode as well as a link to a GitHub discussion where you can leave any thoughts, comments or questions that you have. Thanks for listening. We hope you enjoyed and have a great day.

Open the transcript file

Hosts and Guests

Guests

  • Tristan Brindle Guest Inferred

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