In this episode, Conor and Bryce chat about graph algorithms, the 7 bridges of Königsberg, Hamiltonian paths, Eulerian paths, and more! Link to Episode 220 on Website Discuss this episode, leave a comment, or ask a question (on GitHub) Socials ADSP: The Podcast: Twitter Conor Hoekstra: Twitter | BlueSky | Mastodon Bryce Adelstein Lelbach Show Notes Date Generated: 2025-01-22 Date Released: 2025-02-07 Algorithms by Panos Louridas ICPC 7 Bridges of Königsberg Hamiltonian Path Eulerian Path Elements of Programming C++20 std::ranges::transform Haskell iterate C++20 std::views::iota From Mathematics to Generic Programming (FM2GP) PLVM Meetup 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.
ConorI was thinking that this is very similar to a Hamiltonian path. Uh and I just looked up what the difference is. So a Hamiltonian path is a path in a graph that visits every vertex exactly once. Yeah. Whereas an Eulerian path is a path in a graph that visits every edge exactly once.
BryceYeah. And it's sometimes also called an Eulerian path or an Eulerian walk. And an Eulerian circuit is a Eulerian path that starts and ends at the same node.
ConorWelcome to AD SP the Podcast, episode 220, recorded on January 22nd, 2025. My name is Connor, and today with my co-host Bryce, we chat about graph algorithms, the seven bridges of Konigsberg, Hamiltonian paths, Eulurian paths, and more. End of episode 219.
BryceSo um as I said before, I listed out some of the chapters, and chapter one of this book was What Are Algorithms? Um and uh he he he makes this one one other claim that that I didn't mention, um, which is that he says that uh that algorithms he claims that algorithms must terminate after some finite number of steps and that an algorithm cannot run forever. I'm not sure whether I agree with that. But in general, I I I like the way that he defines algorithms as as an algorithm is a set of steps to uh to do something, and it's a set of steps that you could execute by hand.
BryceI'm on board with that. But so chapter two is uh about uh graph problems, um, which is not something we've ever really talked about um amongst ourselves. Um do you know much about like graph problems?
ConorUh not really. I mean there was a point in time back when I was a budding ICPC uh programmer, and you had to then go and learn all the different shortest paths.
BryceYeah.
ConorI can't even remember the name of them, but like there's the beginner ones that ooh, let's look it up. I I want to say Mistral, but that's the name of an LLM, not the name of a shortest path. Uh let's go for actually, let's actually go to an LLM. Uh because I want to say one of them is like a three-letter as well. It's like Kim or Kip or something like that. List me just the names of twenty um very common uh graph algorithms, obviously Dykstra and A star, but I want um I think they're called like spanning, something like that, common graph.
BryceI do believe that this chapter covers uh it's got a section and shortest path, and I believe it covers the deikstra algorithm.
ConorYeah, Prim was the so I'll just read out uh the 20 that uh we're on ChatGPT. Am I on O1? I know they give you like a couple of O1 for freeze, and I think this is my first it's 10 a.m. on a Wednesday.
BryceYou know what I've used my O1s for? Travel planning.
ConorI was about to say travel planning. Um it's very good. Do you know that through Perplexity AI we have O1?
BryceOh, I did not know that.
ConorYeah. You gotta just register, email them.
BryceBy we here, Connor means we in video.
ConorYeah, yeah. I was about to explain to the listener, this is uh what do they call it, inside baseball talk. Uh I think I'm not actually sure for how long, but perplexity AI, which is a search engine, which gives you uh all the top models. Let's actually go. We're gonna we're taking a tangent, folks. First time ever on ADSP the podcast. And you can our first tangent ever. Yeah. First tangent ever. It it you can either set it so that it says uh default, which is optimized for fast search, and I think then it dispatches to any one of the models that it currently supports, and or you can hard code it to the one that you want.
ConorSo currently I have mine hard-coded to 01, but they also give you access to Claude 3.5 Sonnet, Sonar Large, uh GPT 4.0, sonar huge, Groc 2, and Cloud 3.5 Hiq. And I just became self-conscious of the way that I'm saying, is it cloud or is it claude? We don't know now. Anyways, forgive me, internet, for potentially mispronouncing uh cloud. Um and it's actually pretty pretty phenomenal. Like uh the stuff that I'm able to do now, half the time if I'm doing personal stuff, I'll just give it a whole script and I'll just be like, I need you to make these changes, and it goes and does it, and uh, it works.
ConorAnd it's fantastic. Anyways, Perplexity AI. I guess that's uh, you know, if if you are quote unquote sponsoring a company where some of the employees at that company may have uh podcasts, you may end up getting a free plug. I don't know how much it costs, because we get it for free for I think a number of months. Currently it's still letting me log in. So, anyways, uh I also have ChatGPT though, because uh ChatGPT gives you faster responses. So if it's something that I know that whatever the free model is will do okay with, such as listing me a list of 20 things that doesn't actually need to be correct, we just go to it for that, and uh we go to Perplexity AI if we're willing to wait five seconds for a response, which is uh quite a bit.
ConorAnyways, top 20 algorithms brought to you by ChatGPT, could be O1, could be whatever the free tier is. Number one. Actually, do you want to play a game? You want to see how many you can guess? Oh, of the top uh graph algorithms? I'll give you if you can get five, because I technically already said that number six was prims. That was the three-letter one, which actually turned out to be four letter, four letters. Well, there's there's actually there's And actually you think think of very generic, like not even named after someone or with a special name. Like number one and number two, you definitely know.
ConorThe question is whether you whether you will think of it. It's more of like a category of algorithms.
BryceWell, there's probably connectivity or like uh connected components. It's probably somewhere on there.
ConorYou're thinking too specific. Okay. You definitely learn this in school. Like everybody learns this as the two broad categories of uh graph algorithms.
BryceBreadth first, depth first.
ConorYeah, so depth, DFS is number one, BFS is number two, and then we'll just go down the list. Full points for Bryce.
BryceThere's gonna be like a shortest path.
ConorUm Yeah, Dykstris is number three, number four is Bellman Ford, five is Floyd Warshall, six prims, seven Kruskulls, eight A Star, nine Johnsons, ten Tarzan's uh algorithm for str for strongly connected components.
BryceYeah, I said I I said connected components. I that counts.
ConorEleven is Kosaraju's algorithm for also for strongly connected components. 12 topological sorting. Uh 13 Edmonds Carp algorithm for maximum flow. 14 Ford Fulkerson algorithm for maximum flow. 15 uh we're gonna mispronounce this one. It could be uh we're gonna say Dinich. D-I-N-I-C algorithm for maximum flow. 16. I'm gonna mispronounce this one too. It's getting rough, folks. Here Holzer's algorithm for or actually I'm guessing because it's named after Euler, it's gonna be pronounced Eulerian circuits.
BryceYes, that's actually we're we're gonna we're just about to talk about the Eulerian circuits.
ConorThat's one of the seventeen is Flurry's algorithm for an Eulurian path. 18 is the Chu Lyo slash Edmonds algorithm for minimum spanning arborescence. Uh 19 is Khan's algorithm for topological sorting, and 20. Thanks for staying with us, with us, listener. Warschall's algorithm for transitive closure. And uh Prims was the one I was thinking of. I think Kruskal. But here, let's ask one follow-up question.
ConorUm What is Prim's algorithm?
BryceIt's uh minimum spanning trees.
ConorI was gonna say minimum spanning tree. Uh an example of I I knew the word spanning was in it.
BryceI remember that from school.
ConorSo I got I got into the memorizing of these algorithms at a certain point in time. I think I got through all of the MSTs and and then I think you know, Dykstra, A stars, those are all pretty straightforward. But then I didn't get to uh I mean Floyd War Warszhall, I think I programmed once. I don't remember what the algorithm is, um, but all of these like hyphenated names, that's why I kind of like fell out of love with ICPC, is that like, you know, intelligence and intuitive thinking works up to a point, and then like you just need to start memorizing these hyphenated named algorithms for like specific things, which uh didn't really I didn't really fancy that.
ConorDidn't really fancy that. Anyways, over to you now.
BryceLet's talk about your Eulerian circuits, because that that's where this this uh book opens up on and I I had a f I had a flashback to my uh my college years.
ConorIt's pronounced Euler, uh, and I've I didn't I used to pronounce it Euler, and then I got raked in the YouTube comments. It's pronounced Euler, so I believe that means that it's an Eulerian, unless if they change the pronunciation of Euler's name.
BryceAlright, we can go with Eulerian.
ConorUm come at us, Internet, if it's wrong.
BryceUm So I I remember this example from from school. Uh are you familiar with the seven bridges of Konigsberg? Not not Kingsberg in uh Norway where we've been. Norway Konigsburg.
ConorI have I have not been, only you have been to Tech Town.
BryceUm, okay. All right, all right.
ConorFair enough. Um I am I am familiar with the graphic, but um I think it's just you need to cross all seven is the problem.
BryceSo it's uh is it uh yeah, it's seven bridges. Um uh you should you should go we'll we'll we'll put a link in the show notes, but you should go look at the uh look at the bridges. Um and uh uh there I put a link for you in there. Uh and and the the question is there's there's an island in the center that some of the bridges are connected to, and then there's two um uh I guess there's two islands, um, and then there's two uh mainland portions of the of the city.
BryceThere's there's uh I presume the north and south. Um and the question is can you visit all of the bridges and uh all of the islands without crossing a bridge uh twice uh twice. Yeah. And I'll save you and the listener the uh uh the entire lesson.
BryceBut uh uh basically this for some reason this got uh became a mathematical challenge in like the 1700s, I guess, to to spur the citizens, the local citizens, to think more deeply about math. And uh Euler uh came up with a solution, and his solution is that the answer is that no, you uh you you cannot do that. And if you if you look at this as a graph, if you form a graph from uh the two islands and the two mainland regions, you'll see that um uh every node is connected has an odd number of connections to uh uh all the other nodes.
BryceUm and so because of that uh that means that you can't visit every bridge, right? That if if um one of the nodes of this graph has three connections to it, um you know if it had two connections to it from somewhere else, two edges to it, then you could go visit it over one edge and come back over the other edge. But if it's got three or if it has any odd number, then uh you're gonna always be left either unable to visit that last one, or you'd visit that last one, and then you would need to go back on a uh a bridge that you've already taken.
BryceUmes that make sense, sort of the intuition there?
ConorI think so, yeah. I was thinking that this is very similar to a Hamiltonian path. Uh and I just looked up what the difference is. So a Hamiltonian path is a path in a graph that visits every vertex exactly once. Yeah. Whereas an Eulerian path is a path in a graph that visits every edge exactly once.
BryceYeah. And it's sometimes also called an Eulerian path or an Eulerian walk, and an Eulerian circuit is a Eulerian path that starts and ends at the same node. It's called an Eulerian circuit or an Eulerian tour. What what was your all though?
ConorWell, no, I'm just trying. I always thought that Hamiltonian paths weren't allowed to visit edges more than once.
BryceIt may it may be I think it may be that they can't visit an edge more than once, but they don't have to necessarily vi visit every edge.
ConorOh yes, yes, yes, yes, yes. Because I was gonna think, by definition, if you have like um if you visited an edge more than once, you're gonna inevitably end up at a vertex more than once.
BryceAaron Powell And I I think that I think that an Eulerian path, you just have to visit all the edges in exactly once. You don't necessarily you can visit the same node multiple times. Yes, yeah. That's the distinction. That uh Hamiltonian is uh visiting each vertex uh exactly once, and Eulerian is visiting um each edge exactly once.
ConorIt's in it's interesting because I definitely know that in my programming competitive programming days that I've solved Hamiltonian path problems, and I'm pretty sure the problem even stated that it was a Hamiltonian path problem, but I've never come across Eulerian paths. Maybe it was implicit in the problem, um, but I I've never come across the word, which is surprising considering that it is the sibling concept to Hamiltonian.
BryceYeah. So then a a uh uh you know these are called either paths or walks, and um uh when when you have a path or a walk that starts or ends at the same vertex, it's called a tor or a circuit. Um so, anyways, a little bit later in the book uh in the chapter, it starts talking about um uh uh cycles and graphs and graphs without cycles and um uh and the importance of DAGs.
ConorUm and that's when I was I was gonna say, are we gonna say DAG directed acyclic graph?
BryceWoo!
ConorIt's every everybody's favorite word. But that's acronyms.
BryceThat's when I pulled out um EOP uh because uh for some reason it just reminded me of well, I guess it's not shocking, but it reminded me of this part of EOP where um we talk where where Steppin' Up talks about orbits. Are you familiar with this part of EOP? I think it's like the second chapter.
ConorNope. I don't think I ever got past I'm not even sure if I got past the first chapter. But wait, hold that thought. I'm I'm gonna get a beverage. I'll be back in two seconds.
BryceOh, that was quick. That was actually two seconds.
ConorAlright, orbits, EOP go. Chapter two, apparently. Well, how many chapters did Zach say we're supposed to read? Because I thought I had read I think I think maybe it's just the first chapter. He said, like the first chapter is like 80% of the value. Uh so maybe I did I didn't make it to chapter two.
BryceSo chapter two, and I've always thought that it's an interesting topic for chapter two. Chapter two talks about transformations and their orbits. And when it says transformations, it's not, I think, the type of transformation that you would think, which would well when I say transformation in the context of Stepanov, what do you think?
ConorStood transform?
BryceRight. Um, which is taking a function and applying it to all elements of uh uh you know some sequence.
ConorUm what every other language actually not every, what almost every other language calls map. Some languages call it each.
BryceUh but my guess is maybe Stepanov would also call it map, because when he talks about transformations in EOP, uh that's not what what he means. In EOP, a transformation is taking a function, uh a function that's a homogeneous function. So it's a function that on some particular type T where it takes in a single T and it produces a single T. And a transformation is what happens when you take one of these functions and you apply it uh uh repetitively uh uh in succession.
BryceSo like you do like uh you know f of f of f of f of f of x um or f to the nth power of x. Does that make sense?
ConorYep.
BryceUm and so like in the case of um uh uh uh uh if your your operation is multiplication, then literally taking the the nth transf transformation would be taking the number to its nth power. Um and uh if your operation is is a sum, then applying the taking it to the nth uh uh uh uh transformation would be uh multiplying the number by you know n times.
BryceUm and uh so that's the notion of a transformation. And then he talks about orbits. And an orbit is uh a structure to a transformation. So uh a transformation uh has an let me see if I can get this right. Um its orbits are the elements that are reachable from a starting element by repeated application of the transformation.
BryceSo y is reachable from x under a transformation f if for some n greater than zero uh y equals f of n of x? Um so like if you can if by calling f of you know repeatedly on x, if you can get to y, then uh uh that's an orbit.
ConorUm so for for well Well actually here's a quick question. Does this transformation only apply to binary operations, or is this also applicable to unary operations? So like in the examples you use, like the transformation of repeated pluses leads you to multiplication, the transformation of repeated multiplies uh leads you to power, those are two different binary operations that lead to a different binary operation. Is the same concept applicable to the plus one function, a partially applied binary operation, which is a unary operation, in which case if you're starting from any value, plus one will get you to all the integer values?
ConorUh technically. If you're starting at one, then it's technically natural numbers.
BryceSo he when he's talking about transformations, he talks about um he he says that it can be applied to unary operations, but that you mostly we care about uh unary and binary. Um so I think it's meant to be applicable to both unary and binary. Although uh he defines transformations as being uh something on a unary function. So my my examples of uh addition and multiplication may have been poor ones.
BryceI guess I guess I guess what he means when he talks about he gives uh sum and product as examples here. The transformation of of sum or product would mean like f of x would be x plus x or x times x. So it'd be a unary operation, not a binary operation.
ConorI see. Makes sense.
BryceIn which case what I said before about it uh uh uh about what you get out of sum, it wouldn't be multiplication, right? Um because uh uh f of f of x if if it's sum would be uh what it's so I think this is classically um it'd be a power.
ConorBen and I chat about this all the time that like Haskell has this function called iterate, which is the repeated application of a unary function to a starting value. And like iota is a specialization of this where iota's unary function is plus one, and you give it a starting value, and then it just generates you a sequence of however many times you want to apply that function. But you can generalize this to any unary operation. If you want to create all the powers of two, you can create a multiply by two unary function and just start with one, and then you end up with one, two, four, eight, sixteen, etc.
ConorExactly. Uh and this concept is called different things in different languages, but uh I I think. Yeah, and Haskell is called iterate, and that's where we will.
BryceWhich is interesting because iterate is uh uh it's an interesting name. It stepnov calls it transform, Haskell calls it iterate, um huh. Uh yes, I think that's exactly the same the same idea. Um but uh yes, some of what I said before may have been not quite correct. Um but it's certainly when he's talking about it here, he means it as a as a unary operation, a unary power operation. So so we we explained what orbits are, and then um there's a notion of of an orbit can be cyclical.
BryceUm so uh an orbit is cyclic if there is some in such that x equals f of n of x. So if at some n some number of root applications of your function you get back the original x, then you've got a cycle.
ConorYeah, it makes sense. The easiest thing you could think of is uh modify your plus one unary operation with a another unary operation of modulus ten. Right. So if you start at zero, you'll go up to nine, and then when you hit ten, you'll shoot back down to zero, which is obviously cyclic.
BryceAnd then there's also a notion of uh terminal um uh transformations, which is um it's terminal if at some point uh uh you you reach a point you reach an element that is like not in the domain. So like at some point the sequence ends. Um and so if it's not terminal um and not cyclic, then it's infinite. So he says that there's he he defines four different types of shapes.
BryceThe first is infinite, if there's no cyclic or terminal elements, then there's terminating, which is if it has a terminal element, um, and then it's uh circular if it uh if it's cyclic, and then uh it is p-shaped if x is not cyclic, but its orbit contains a cyclic element. So think about the p as like there's some starting sequence that you never repeat, but then uh you start at like you know zero and then you go to you know five, and then there's a cycle from like five to ten.
BryceMake sense? Like if you start you start iterating the the the orbit and then uh uh eventually you enter a cycle. So you won't repeat everything, but that you eventually could go into a cycle, and that's what P shapes be.
ConorOkay.
BryceDoes that make sense?
ConorYep.
BryceThere's there are diagrams, hopefully.
ConorVery nice.
BryceAnd um and uh then he talks about some algorithms for uh uh determining the cyclic structure, determining whether uh where the the collision points would be, um uh where the the collision point is uh a a point at which you have two um uh two different like the the the the the point at which you have two different numbers of invocations that um uh give you the same value.
BryceSo basically this is how you you detect cycles or the length of a cycle. Um so it's like you know, if there's is there some f to the nth power um uh that is equal to some f to the you know nth power, um, or he says f to the 2n plus 1 power. Um and then he goes into some algorithms for um measuring for finding collision points um and for measuring uh the size of uh orbits.
BryceUm and uh it's a very interesting part of the book, and it comes up in a couple later parts of the book. I think it also comes up in the by forget uh coordinates section, which is uh chapter like 18, which I know well because uh when I was in college, Hartmitt uh I did an independent study class with Hartmet where we no, sorry, chapter seven is coordinate structures, and I did an independent study class with Hartmitt where we uh where we focused on coordinate structures.
BryceUm that's what made me remember some of these algorithms. Um but I was thinking about the this notion of uh uh transformations and orbits and cycles, and whether there's any relation to uh graph algorithms and cycles and cycle detection and graph algorithms. Um and uh I don't really know what the applications of this field of transformations and orbits uh uh is.
BryceUm I imagine that like random number generators and things like generation functions like a are one such case. Um, but I don't know what the other applications would be. But but reading this first book, this algorithms book, got me thinking about whether this part of EOP has some relation to graph algorithms.
ConorWell, this is just uh another reminder that I need to go read EOP to have something to say here. I mean, this is technically part three of this conversation, which is not gonna come out before next Monday, but I m I may. It is penciled in, although I probably should have announced it by now. It is penciled in that potentially the first meetup uh for the From Mathematics to Generic Programming book. I'm reviving the programming languages virtual meetup, which has been dormant for uh over a year.
ConorTechnically uh Leslie was running it for I want to say a year, year and a half while I was taking a break. Anyways, and that is the sibling book to EOP. But yeah, I mean it definitely sounds like there's uh definitely some what do you call it, similarities between those two things. Do you have like further thoughts other than when you were reading the algorithm or the the graph chapter that it was making you think of uh EOP? And that's what uh led you to reaching for it.
BryceThat's sort of the reach, that's sort of the point where my reading reached um that uh that I was like, hmm, I remember this section of EOP and maybe it's got some uh some relevance. Um yeah, that that was that was pretty much that was pretty much it. I I mean I suppose I I'm just curious as to what the application of these Stepanov transformations and orbits are. Like why why is this chapter two in his book?
BryceUm and I know it comes up in some of the later chapters, um, and maybe maybe that's the answer, and I just need to reread the whole book. But um but yeah. I I guess the the thing that but the thing is that that when he talks about um uh detecting cycles um and and measuring the length of cycles, um uh he's talking about it on not on sequences or on graphs, but he's talking about it on one of these transformations, which is just a a function that you call on you know one particular piece of data.
BryceUm and I mean that's useful. Um but is that something that can generalize to like I have a a graph or I have some not just like you know a function, but like I have some data, you know, some structure, some data structure. I think that's usually where we want to do like cycle detection, is like some structured data, not on like a function itself. Um and we so really that's that's the question that I had when I was reading the other book was that I'd remembered this part of the Stepanov book talking about uh uh you know this these techniques for um finding cycles and classifying the uh the the shape and structure of uh you know applying a uh a unary function to some a homogeneous unary function to some input.
BryceUm and I'd never thought before that, huh, maybe that generalizes to uh uh finding cycles in data structures or things like that. And I'm not sure that it does, but maybe it does. Yeah, maybe that may maybe if maybe if you have an algorithm to detect a cycle in some, you know, uh uh like function f of x, like imagine that I'm trying to detect like find a cycle like in a list.
BryceMaybe my function could just be uh you know, like advanced to the next element in the list. Yeah, I guess it could, right?
ConorLike imagine that my Well, so that's if you're trying to c compute some kind of like reduction on some iterate sequence, if you detect a cycle, you can just convert it into some kind of not constant time operation, but it's constant time once you find the cycle, right?
BryceWell, well, but so so I guess I guess I guess here's one of the applications of it. Imagine that my transform imagine that my input is some iterator, and that my my transformation function just calls next on the iterator, just advances the iterator. Right? So then I can use these Stepanov algorithms for detecting collision points and detecting whether something is circular and whether something is a cycle.
BryceI can use that to see if there is a cycle in my whatever thing it is that I'm iterating.
ConorRight. Yeah. Yeah. That makes sense.
BryceUm so then that's there there's the usefulness of this. I guess I must have may maybe that's something that's obvious. Um it doesn't seem particularly insightful, and maybe it was just so obvious that he didn't explain that particular application, or maybe it comes up in later chapters and I'm just not remembering that. But it seems like this generalization of the cycle detection to just an arbitrary function, it gives you the most general form of this.
BryceYeah. He really he really was very brilliant in his ability to get to the most generic form of something. Speaking speaking of meetups, speaking of meetups, and this is a good a good segue slash closing.
ConorWell, I was gonna say part four. Speaking of meetups, and then we're gonna get some 2025 hot takes. I mean, technically, the time this comes out, it's gonna be like mid-February. Be sure to check these show notes either in your podcast app or at adspthepodcast.com for links to anything we mentioned in today's episode, as well as a link to a 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, what is the tagline of our podcast?
ConorIt's not the tagline. Our tagline is chaos with sprinkles of information.
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:...
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:...
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...
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...
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:...