DataTopics Unplugged

#33 Everything is Technical Debt

January 22, 2024 DataTopics
DataTopics Unplugged
#33 Everything is Technical Debt
Show Notes Transcript Chapter Markers

Welcome to the cozy corner of the tech world where ones and zeros mingle with casual chit-chat. Datatopics Unplugged is your go-to spot for relaxed discussions around tech, news, data, and society.

Dive into conversations that should flow as smoothly as your morning coffee (but don't), where industry insights meet laid-back banter. Whether you're a data aficionado or just someone curious about the digital age, pull up a chair, relax, and let's get into the heart of data, unplugged style!

In This Episode:

  • OpenAI's Election Strategy: A deep dive into how OpenAI is approaching the 2024 worldwide elections. How OpenAI is approaching 2024 worldwide elections
  • Tech Quirks: Exploring the idiosyncrasies of the Continuity Camera in Chrome. Is it a magic trick or just poor engineering?
  • The Rise of Bluesky: Unpacking the significance of Bluesky's newly launched RSS feeds. Bluesky launches RSS feeds
  • Hot Takes on Data: Tackling the bold statement - Every data transform is technical debt. An exploration of what technical debt really means in the data world. Every data transform is technical debt

Join hosts Murilo and Bart as they navigate these intriguing topics, bringing their unique perspectives and a touch of humor to the table.

Intro music courtesy of fesliyanstudios.com.

Speaker 1:

Tell me when, when, hello, hello, hello.

Speaker 2:

Hello, hello, hello, welcome to Data Topics Unplugged. Hi, yeah, I'm with Bart. Yes, happy to be here. Yes, so what is the topic? It's unplugged. It's a light, cozy corner of the web. You know, the way we talk about what happened in data in the world. So really anything goes from AI because whatnot right I feel like someone's still kind of hot topic From yeah, it's Lili, right From Blue Skies Anything really goes. So we're on YouTube. We hope we're on YouTube. Yes, we are on YouTube. We're also on LinkedIn, maybe.

Speaker 1:

We're trying. We're trying First time. We're trying to street in there.

Speaker 2:

Yes, so bear with us if it doesn't work.

Speaker 1:

The show can't stop. If it doesn't work, it will work next time.

Speaker 2:

It will work next time. You know, If anything, we are agile, iterate, and also we are on Twitter or X or again.

Speaker 1:

hopefully we're not checking, but that should be the case.

Speaker 2:

Yeah, but the X.

Speaker 1:

Twitter doesn't exist anymore.

Speaker 2:

Yeah, but it's twittercom though. It's weird right.

Speaker 1:

Yeah, we don't talk about it.

Speaker 2:

Oh, ok, yes. So, the one that shall not be named. Ah yes, we are Check it out.

Speaker 1:

Nice Cool.

Speaker 2:

All right. So yeah, hello everyone. So today I'm your host. My name is Morello, as Adelphar mentioned, bart, and today, At your service and today is just us. So today is the January 16th of 2024. Normally we record on Fridays, but today is Tuesday and I must admit it's my fault, and the reason for that is I will not be off. I'll be taking holidays on Thursday and Friday. Nice yes, any plans? Yes, I'm going to go to Spain and I'm from Spain. It's a Gran Canaria, so it's one of the islands. Oh, nice, and yeah, africa. Yeah, I have a friend that lives there. I was actually there earlier, no late 2023. Okay, for a Picon, so just oh yeah, Picon was there. Yeah, but it was a different island, right. So now it's just what kind of temperature is there Right now? I think it's like 20s, okay, nice, yeah, so it's definitely warmer than here, and you're only going for the weekend. Yes, I'll be back on Sunday. So Thursday, friday, saturday, sunday, it's actually kind of a long flight.

Speaker 1:

And Gran Canaria. Don't you need to be 60 plus.

Speaker 2:

It's my friend's 30th birthday. So I think maybe he was like yeah, no, it's okay, they live there. Yes, but this friend of mine, he's very how can I say it? He's like not crazy. But you know, he's like he's from Barcelona. Then he came here to Belgium to study with him in the Masters. Then he went to Indonesia for a while no, he's in Gran Canaria. He said he's probably not going to stay there long term, but he already bought an apartment because it's nice. Nice to be able to say live on an island, yeah exactly, and he likes to surf too, so it's a very nice spot. Gran Canaria, is it Spanish? It's Spanish territory, officially, cool.

Speaker 1:

Don't have to look forward to it Indeed indeed.

Speaker 2:

So when I come back, hopefully it'll be more 10. This snowy weather, it doesn't suit my Brazilian. You know what about you? What have you been up to this past couple of days? Could be Metal Friday right.

Speaker 1:

Good question, nothing, much, nothing, really. Nothing worth mentioning. Nothing worth mentioning now, okay.

Speaker 2:

And you want to see. I know last time I got some heat for my sock choices, so today I came prepared.

Speaker 1:

Oh nice, yes, Okay cool. You see you have one of those two words. I have them. You're showing the latest branded ones, but I might have a few of those on them. Oh, cool, Okay.

Speaker 2:

Nice, nice, nice. So what do we have for today? So maybe I can get started For the people as well on the live stream. We'll try to share a bit more what we're up to, what we're looking at. This is a bit of a personal rant.

Speaker 1:

I'm curious what this will be.

Speaker 2:

Yeah, as you may or may not know, we live streamed this podcast. We actually use the iPhone paired with the Mac, right? So it's very nice. So if you have a nice iPhone, you can just reuse the camera there as a webcam and for things like this it's actually really nice. And also it's not like it's better than the camera that is in the MacBook itself, right. So if you want to have a nice camera but you don't want to go for a full setup, you can use your phone really nice, okay. And I was trying this out with my iPhone and my MacBook and it was not like it worked on FaceTime. It worked on like the camera thing, like whatever thing that was system was working. But whenever I went to the Chrome so Google Meet, restream, whatever it would not work. Okay, and I was like why the hell is not working? And I was trying different things and then again I'll check, check, check, google documentation nothing. And then I came across this blog and it's not really well. It's a very short post. It's just saying, like iOS continuity, the name of this is continuity camera, right? So to use your iPhone as a webcam is called continuity camera and it just says iOS continuity camera not working in Chrome and it turns out that this is not a bug, it's actually a feature, I guess a security feature, and apparently this is not documented. But there's a Reddit post where someone, a user, goes through it and apparently the iPhone only works if it's on landscape mode and the phone is locked, or browsers. So that means that any browser? I think so. Any browser. That means that if we had the phone on what's the opposite of portrait, it's portrait and landscape. Yeah, if we had it on portrait. And for those that forgot, just like me just now, it's like the vertical one, like if your phone is vertical, it won't work on the on the web browsers. So Chrome being one of them.

Speaker 1:

And simply because you should not record a portrait mode.

Speaker 2:

I don't know. I don't know if it's a security thing, because you don't want to by mistake turn on the camera or something right, but it's like to me this is the kind of like cause I was like part does it all the time. He never has an issue. Why am I having issues, like what's wrong with me, you know? And then it's like it's very hard to also debug it right, cause it just works.

Speaker 1:

Yeah, that's true.

Speaker 2:

Right. So it's like, and then I started to think like okay, like I think Mac and Apple in general, they have this. It's very elegant that things just work. Yeah, yeah, yeah, until it doesn't work.

Speaker 1:

That's true. And then I was. I was something similar. What did you do? It was not with the camera. What do you have? I was. I didn't need to think about how it was. Again, I was working I have an iMac and a laptop and I was working on my iMac with my mouse and my keyboard and suddenly, while I was moving my mouse, I was working on my laptop and I was using my mouse of my iMac, my laptop, and I didn't realize that I wasn't conscious of it. It suddenly clicked and I go this mouse is working on my laptop, but it was just working and you didn't notice.

Speaker 2:

It was just working, but you didn't notice.

Speaker 1:

Actually, I didn't notice I put something in the middle. But you can just like move from the moment your laptop is next to your iMac. You can just move your mouse off screen and it becomes the mouse. And I was like what? And I just figured that because it's so seamless and then at some point it didn't work anymore. I wanted to show someone it didn't work anymore and I couldn't figure out why it was not working Exactly, and then I had to reset the stuff and update something.

Speaker 2:

Yeah, but that's the thing. I updated the software on my phone and then the MacBook and then it still wasn't working, like Google stuff and it wasn't working. And I think it's kind of like the magic of Apple, like things are just so seamless that everything just works, but then at the same time it's like they don't have a lot of moments that you can check, like to kind of debug and kind of go forward. So then I was wondering, like is this something for this particular continuity camera? Is this? Do you think this is just over engineered? Do you think this is? Is it just magic or like, is it a good goal to have something so seamless? But then you put some small constraints here and there and then users are going to be like well, why is this not working?

Speaker 1:

Well, in your specific case, I think it's a bit of an intuitive to say it doesn't work in portrait mode, right?

Speaker 2:

Yeah, I thought so at least.

Speaker 1:

Maybe Apple is very opinionated on that. You should record stuff in landscape mode, which I think is common sense, right?

Speaker 2:

Yeah, but still, you know. No, but it's like, if you want, to, but it's like it's not even.

Speaker 1:

but actually thought you were, because we're showing the article you're referencing on the screen, so it's actually the first time we're doing this. Right? Yes, but the title is on using the continuity camera and not working in Chrome. I thought for you we're gonna, this was gonna be about Chrome, but it's not right. No, this is always.

Speaker 2:

This is always as far as I as I understand here, right.

Speaker 1:

Okay, interesting, yeah, and good to know.

Speaker 2:

Yeah, and then there's also some other features, like I think that was something called like stage, something that apparently you need to use your phone and it's supposed to just connect, but I cannot get it to work either. So then I started to always think of these kinds of problems, that it should just work and it should be intuitive. But then if it's not intuitive to you, then you kind of screwed. Yeah, that's true, you know, and the thing even with the thing with the iPhone.

Speaker 1:

How often do we have this or something is not intuitive to you.

Speaker 2:

Next question and the other thing too is like I mean the continuity camera, you just need it to be a landscape mode to start, and then after that you can go Okay, it's just to start, it's just to start.

Speaker 1:

That's the thing.

Speaker 2:

You know, but yeah, and I feel like I don't know, it was a bit of a small rent there, but when we talk about Chrome, maybe a good segue.

Speaker 1:

I think we've touched upon this once or twice, but I'm still not satisfied with the answer you gave me so far. You, as a major Russ fan why are you using Chrome and not Firefox?

Speaker 2:

I mean more than a Russ fan. I'm pragmatic, you know. But Firefox is very good.

Speaker 1:

Yeah, firefox is very good, so maybe I think it's an interesting one browser choice, and especially for all the Russ levels here you need to support a bit the community. So there is actually and it was, I don't know, maybe an article a month or so ago that Firefox has an issue in there that their usage is dropping, dropping, dropping, dropping, dropping over time. And they are, and I don't know the exact percentages, but the number I even in mind is 2%. So there is a guideline from the US government that says that all government applications will support browsers that have a usage of more than 2%. I think it's 2% not 100% sure and Firefox is currently at a danger of dipping below that percentage, which would mean that the US government application no longer need to support it. But apparently, like, these guidelines that the US government gives are also being taken up by a lot of other big companies. Yeah it's like a standard right. It's a bit of a standard or an unofficial standard, but it's a US government, a big entity, yeah, and it puts a bit to the future of Firefox. It'd be like the straw that broke the camel's back? Yeah, and it would be, and I think it's really pity yeah.

Speaker 2:

Which browser do you use, bart Firefox, use Firefox. Yeah, oh, look at that Nice.

Speaker 1:

Yeah, indeed, so I will bring this up every now and then until you use Firefox.

Speaker 2:

Okay, do you see? Okay, but maybe a question, valid question, because I think the main reason for going for Chrome is that everything works on Chrome.

Speaker 1:

That and that you want to give all your data away to Google.

Speaker 2:

Yeah, so those are the two reasons. Yeah, I just want good recommendations. You know I just because I go missing, I want someone to know where I am. So I keep being Google, but do you feel any? Is there something that doesn't work on Firefox?

Speaker 1:

Just one thing. Okay, it's a very, very niche thing. So I in my Gmail, so we use the Google workspace. So in my Gmail I have in Chrome, I use a plugin to do multi-forwarding so I can select a number of emails and I do per email, like in bulk, per email, I do the forwarding. It's like a small plugin and it doesn't exist in Firefox. Like this is my only like. This is like literally the only thing why I sometimes switch back to Chrome.

Speaker 2:

So you still have Chrome and you have Firefox on your Mac.

Speaker 1:

Yeah, exactly, and there's actually not one, but I don't use it a lot Adobe podcast, so it's a very lightweight audio studio from Adobe. The only support Chrome.

Speaker 2:

Okay.

Speaker 1:

There's another reason. So two reasons, two reasons Okay, why I can't install it the reason why.

Speaker 2:

why did you move from Chrome to Firefox?

Speaker 1:

I think it was mainly curiosity, if I'm very honest, like there was a great big rebuild of Firefox and to see a little bit like well, how does Firefox feel now? So I at some point I think maybe five, eight years ago switched to Chrome Okay, maybe even longer ago, six years ago and at that point I had really had the feeling oh wow, this is snappy, this is this work, very, very like, very lightweight, like everything works. And now I had the same feeling with Firefox.

Speaker 2:

Do you think it could be like the new factor? Yeah, Never underestimate the new factor, but yeah, yeah, maybe, I mean I think it would be. Yeah, I mean, I guess if I do move away from Chrome, the big part would be just so, because I think Chrome is the main browser there. Right, like most people use Chrome I would say the ones that are using the default, probably.

Speaker 1:

I think if you look in the stats, that's what confirmed it. Yeah.

Speaker 2:

Oh, if I switch, it would probably support the little guy you know and Rust. So that's the real reason actually.

Speaker 1:

I think if you, because you go to a lot of conferences typically well, you often have to present something around Rust in a Python conference setting I think if you would ever show your because you use web-based presentations I do, yes, on an actual Rust conference and you open this in Chrome. I don't know how this will work. Yeah, I have a Chrome there. You might be booted out.

Speaker 2:

I have Firefox just for these occasions, you know, and then for my personal day-to-day, I'll just use Chrome, all right, so I'll do this from now on. I'll do this in 2024. Okay.

Speaker 1:

Let's check next time that we do data topics.

Speaker 2:

All right, but there's a lot of browsers out there. Right, there's like Arc. There is a lot, there's a lot, there's a lot, there's a lot.

Speaker 1:

And what I find a very interesting project to follow. I think it's called Ladybird, that's correct. The Ladybird browser, not 100% sure. I mean, yeah, and they're building a browser from scratch.

Speaker 2:

What do you mean from scratch?

Speaker 1:

Like literally from scratch. So a lot of the browsers that you can download, they are built upon, for example, Chromium on an existing engine. But Ladybird is being built fully from scratch and it's interesting to follow up with the discussions that we're having. Because there are so many specs when it comes to the web, it's not easy to implement it just from scratch. There's so much history. That's a legacy. There's so much legacy.

Speaker 2:

Is this the one that I got the rights like Ladybirddev? Yeah that's the one Interesting. And what are they writing this on?

Speaker 1:

It's one question too many.

Speaker 2:

All right, all right. That's a cool thing. The other thing that I learned semi-recently is that it's a C++ apparently, just quickly opening the repo. Makes sense, makes sense. I also learned semi-recently that every web browser in the iPhone is really just Safari underneath, like you can put something else on top, but it's just running on the.

Speaker 1:

Safari engine.

Speaker 2:

I was like I've been lied to this whole time, Like ever since I heard that someone was like oh, do you don't use Chrome on the iPhone, I was like get out of here, it's the same thing, so yeah.

Speaker 1:

Rappers all the way down.

Speaker 2:

Yeah, 2024. What else we got in 2024? So, we got elections worldwide yeah 2023.

Speaker 1:

We got this article that you just put on the screen how OpenAI is approaching 2024 worldwide elections Recent article, I think yesterday, released, I'm not mistaken, on the 15th. Interesting to see that this is top of my throne. The question I had is what are worldwide elections? But I think what you mentioned is just everything that is ongoing at a small or a large scale or when it comes to government elections, and they present a bit of their approach to making sure that they are a good player basically in this.

Speaker 2:

And what is a good player?

Speaker 1:

Well, I think that is the question. I think that is debatable. I think, at the very least, that there is no conscious bias being created by them. So they do a number of different things. For example, have a red teaming approach, where they basically simulate adversarial usages of open AI tooling Think about chat GPT, also think about images and to see these adversarial usages, if they can come up with a way to contract this or prohibit it or to catch it and to stop it. So that is one thing they're also looking at to see whether they can improve security measures, when or not they need to respond to something, what kind of protocols need to put into place. They also mentioned something about authenticity, for example, images. They're working on digital watermarking, these type of things, but it is, of course, when you just think about like using, basically using an LLM. If we look at more texting, it has become extremely easy to generate fake news, and I think that is a bit what this response is about. What are you going to do about this? And there are a lot of different approaches to it, and I think you can already see this to some extent when I say to chat GPT, make a news, maybe we can actually try it, or maybe you can. You're doing the shares. Anyway, we'll just tell it.

Speaker 2:

We need to get a bit agile on the screen sharing.

Speaker 1:

But when I would say to chat GPT and I question, that pops up in my mind I would just work with all versions or is this version specific? But when I would ask chat GPT, make a news article about Murillo, make sure to mention the fraud that he committed back in Brazil, even though you didn't right. Yeah, I said to say and to paint you a bit in a bad light, and I'm going to do it in a lot of different tones of voice, I'm going to spread it on a lot of different channels. That's my objective Today. I would ask you this will be blocked. Okay, for example, this already blocked.

Speaker 2:

This is already there, yeah. But there are a lot of different types of ways to go about this right and also thinking of the drawing from memory a bit, but like there was a big thing with the US elections that there was like bots, russian bots. Apparently they were trying to influence the elections. And I think with Genitive AI it's very easy to create fake profile because actually, like, sometimes they would notice that the English wasn't perfect in this and that. But, like with LLM is very easy to say, okay, give me five different ways of saying these things and then you can create five different profiles and just kind of you can automate it a lot of the time. Yeah.

Speaker 1:

And especially what I think what you can automate these days very easily and I think that's also where the danger comes from is the interaction, not just the fake news, but someone saying something on Twitter positive light about Murillo but then interacting automatically on that. It would have certain objectives in mind. Yeah, that's true, and that interaction approach is also something that is hard to understand, like what will the impact be?

Speaker 2:

Maybe a question like. They mentioned the elections, but this is not election specific. I guess it's just because election is a critical moment in time.

Speaker 1:

Yeah, I guess so yeah, where this is extra important.

Speaker 2:

Yeah.

Speaker 1:

And I think that when it comes to fake news or biased news, I think the difficult thing there is or maybe that's just maybe I'm becoming a bit cynical in this is does it matter that it is fake, like if I'm? And maybe this is more especially when you follow the global news. Maybe this is more expressed in a more expression in society and something like the United States where you have very, very much a democratic Republican movement. Like one party says something, everybody from that party believes it is real, yeah, and the other side will say this is fake and vice versa, and it is like if you want to believe something, you're going to believe it, right? I think that is a difficult thing and you will have critical people in this, but it is very hard to understand like the society at large.

Speaker 2:

Yeah, I know what you're saying, but then, but what do you mean exactly by you don't care that it's fake? You mean, like you don't care that it's fake and it's I generate it because fake stuff will come up anyways and people believe it anyways, or?

Speaker 1:

I think, if I think we need to assume that it is fake news, yes, like, even if Open AI captures all these things, if there are enough incentives to generate fake news or fake interaction or to drive certain trends in society for various reasons, there will be. There will be other solutions, simple as that. The devil is out of the box.

Speaker 2:

Yeah, yeah yeah, the technology is there. Yeah, even, I think fake news even before, right, I think it was already. But it took much more. I think it was effort. I think LLM's GNI is a catalyst for these things. So now it's really it's like before. It was there, but it took effort, it took like very motivated people and now it's so accessible. It's such a powerful tool and it's so it's available to everyone. That's indeed, and I think I remember reading that schools I want to say it early, but I'm not sure that they were actually teaching kids about fake news already, which is very good.

Speaker 1:

I think that's what you need to do.

Speaker 2:

But I think it's also when I was in school right, maybe those are already fake news, but I think for professors or teachers it was much harder to teach that as well. It's like you're teaching a new generation that, like I mean, even now there's like there's ChedGPD, there's all these things, and now we're trying to teach skills to the next generation that maybe they won't like. We didn't have these tools, we didn't have to learn these things, and maybe the skills that we are teaching are becoming a bit more obsolete, you know. So it's like how do you navigate this generation thing? But you know that it's important, you know it should be mixed up right, so I think it's tricky, but I agree with you that we should never forget.

Speaker 1:

And that maybe this is. I don't really have a clear sense on this, to be honest. So OpenAI has always taken a bit. The OpenAI gets a lot of flag because they're not very open. Right, they didn't release a lot, had a lot of them is behind an API, but because they do it like this, they also can control this very much how it is being used. True, and when you read these type of statements, also the things that they're doing, there is, at the very least superficially, it looks like there is a good will to make sure that this has the right impact on society, and the argument for open sourcing stuff is at least society at large knows what this can do, so you can also prepare for the negative effects of this. It comes, for example, to fake news, but at the same time, you're also giving everybody the tools to do this.

Speaker 2:

Yeah, indeed, so it's a very yeah. It's a bit. It's almost like you're arming everyone, so you know that everyone's safe and the way it's, like you know, keep a few.

Speaker 1:

And there are arguments to make on both sides. I think, I think it's different?

Speaker 2:

No, I think so too, but I think in the end, like today, we're really living a moment that it's not one or the other, it's a mix. Right, there are some that are not open sourced, some that are open sourced, and I'm just wondering how it's all going to play out. But I do think it's an interesting initiative and I agree with your take that fake news is there. Genai is a catalyst. Yeah, we can try to censor it a bit, but it's there. Would you say. That's a hot take, bart.

Speaker 1:

I've been known for hotter takes than this.

Speaker 2:

Sizzling takes.

Speaker 1:

You have a very hot take right.

Speaker 2:

Well, it's not mine, but I think it's. I read the title is a bit clickbaity, I guess. No, really, I think they make arguments here, but the title of the article is Every Data Transform Is Technical Debt. I guess it's data transformation, data transformer, which I think we can spend a lot of time just discussing this. So can you reiterate Every data transform is technical debt.

Speaker 1:

Okay, so that is. It's an interesting one, and I kind of knew that this was coming because you posted it on our Slack channel today. Sure, I did. What is your? Maybe to unpack a bit, what is your definition of technical debt?

Speaker 2:

Yes, and I think that's a discussion on itself. Okay, all right, and I'm not going to go by my definition. What do we mean? I'll just say what's written in the article. So we have common ground. Okay, it's in the article huh, yes, so actually this came from another blog post that says all code is technical debt and it says that technical debt I like seeing that article is saying that it's like you're making little trade-offs, you make something. It's like you actually compare it to making a loan on the under article. You know you make, you take some now and it's okay, you can pay it back. But usually when you take a loan there's interest right and then if you take too much loan and you don't pay it back, then you can go bankrupt. That's the parallel. I think the parallels fall short in a lot of ways, but that's kind of how he defines it in the all code is technical debt. So this one right here. So it's a different article.

Speaker 1:

Okay, but this is a very hard level definition, huh.

Speaker 2:

What would you define?

Speaker 1:

If you just ask me this out of the blue, I would probably say something like I'm and it's maybe not sure I'm encompassing all things but I'm hacking a tool together. Right, this tool is going to have something or a small app together. This app is going to have a few users. Right, speed is more important than what we're actually long-term building, so I'm taking a lot of shortcuts, cutting some corners to make sure that we have something there that is good enough for the 10 test users that we have in mind, but I know when we get to 100 users, this will not be sufficient. It won't scale, it will show cracks.

Speaker 2:

So it's like, in a way, it's like taking, like cutting corners, knowing that right now it's okay, but you will have to deal with it later.

Speaker 1:

You'll have to deal with it later because, like these, cutting these corners, like they put maintainability at stake. Maintainability or scalability, or robustness.

Speaker 2:

Something's going to go wrong. Something's going to go wrong. Yeah, okay, I agree with that. So the main argument with code being technical debt is that every code has assumptions and the more code you add, sometimes code becomes legacy code. Right, so maintaining that code, it is technical debt. Adding new things as well, trying to like scope creep kind of thing, I know you have a tool for one thing and then you try to add more and more and more, but then that becomes debt because now you're trying to do too many things at once. And then he says well, you can mitigate a lot of these things by adding tests, but tests also cost to develop and all these things. So this is kind of the argument for code is technical debt. And then the author here it goes further to say that in data transformations it's actually worse for a few reasons. First is a lot of the data transformations are done in SQL which, according to him, are hard to test. So I know that DBT has some testing things, but the premise here is that SQL is hard to test. So before code gets a bit, it's hard to make sure that it's working properly. The other thing is that data it's extra context that you don't have in your code base, so there are some things that may come from the system that you're not aware of. Right Sure, he also says that a lot of people actually fix these issues in ETL, so in data transformations, which it's not. According to him, it's like it just creates debt because you're kind of going around the problem, you know, instead of saying like, if you fix these issues, you mean data quality issues that should be fixed as a source. Yeah, exactly. Actually, he does go back to saying there's a diagram here as well. That is saying that a lot of the times, the database or whatever, they have some logic to verify some things, and then we end up duplicating that because we don't have access to that original system, right? So we're double checking things, some things we're not checking, some assumptions that we have, we don't know. Maybe one example drawing from the top of my head here is like maybe timestamps or something. You're assuming that this is the timestamps that you have, but you're not sure. So instead of actually trying to Time zones, time zones and then instead of actually going in the source and trying to fix it, you just say I'll just do a ETL transformation there. Ok, I think that's kind of trying to illustrate it. This is my assumption. So he says, basically there are all these problems, all these things that can go wrong. So the only solution and I'm quoting here the only solution is to reduce the amount of data transformations we do, because every data transformation is taken with it.

Speaker 1:

Every data. But what is your opinion on this statement? Every data transformation is technically up.

Speaker 2:

It's a hot take. I disagree with it. Personally, I think there's some, it's food for thought, it's interesting, but I don't think I think data transformations in my head is like you need to. There's a point to it, there's business logic, there's a goal you're trying to accomplish and sometimes the only way to do this is to reduce transformations. Like, yeah, you're increasing complexity, but to me, if you say that technical debt is cutting corners, you create look, yeah, but that's a bit.

Speaker 1:

What is the definition right? Exactly Because, to me, the example that you gave you are fixing something, a data quality issue, that comes from the source system and you're fixing it in ETL. You're basically creating a band-aid for a problem in a source system. Right, and you can say we're short cutting here because easier to fix in ETL than fixing the source. And I think there's an argument to say this is technical debt, and maybe interesting indeed to have this food for thought, because Navator really thought about it as technical debt. At the same time, like to say, any data transformation, like, of course, sure you create it. Like because you do a transformation, you increase complexity. You also, you can argue, you create a dependency, like certainly there is a certain link between systems, that where something happens in between and you're dependent on what happens in between. But for a lot of other use cases, like it's a very conscious choice to do a transformation and which is not really cutting corners, which is not yeah, I think it's like to me.

Speaker 2:

what I kind of disagree is that it feels like they're equating complexity with technical debt and I kind of like that. I don't agree with that thing. You can have a system that is very complex, but if you say that you didn't cut any corners, this is the best. Maybe you can say, yeah, we could put the problem into two and say different things, but increasing complexity does an equate to increasing technical debt. I think that's kind of where I disagree, but I do agree like some things that I thought it was interesting is like maybe we should spend more time decommissioning old features instead of trying to maintain everything. Like, maybe that's because indeed we decrease complexity which will make you more robust and sustainable in the future. And according to them, it is technical debt because it's something that will come and bite you later.

Speaker 1:

Like so many things in life.

Speaker 2:

Amen. So actually he makes the point to go in the data quality, so actually double down on the data quality. So there is data contracts and all these things to kind of fix these things at the source. So again, I think it's a different approach. I think it shines a different light into data quality, which I think is interesting, even though I don't fully agree with all the definitions there. What do you think? What do you think of this conclusion? Like data quality is the answer to your problems, I think at the very least.

Speaker 1:

it's interesting to think about the concept that it can be technical debt, and I don't think from the moment we create a band-aid by saying, oh, this time stamp is in a string field in the database, I'm quickly going to make an actual time stamp in the ETL. I think very little people are conscious that you're very actively creating technical debt there. I think it's a bit of an overstatement to say everything is because then maybe, if that is the case, then maybe we shouldn't write any code.

Speaker 2:

Actually that's the article. I think they say something like this is like yeah, if you just only wait to not do it, you shouldn't do any code. I see what you're saying. I think it's also interesting the scope, like a lot of times it's like oh yeah, if you add more features it's better, and I think by thinking like this, you're just like no, actually adding more features is not better. I think that's also an interesting takeaway. So, all right, all right. So I think today we have a short episode.

Speaker 1:

Maybe. If you have any questions, please leave them in the comments. Which topics did you want to discuss? We have Blue Sky left.

Speaker 2:

Ah, you want to see the Blue Sky.

Speaker 1:

Maybe very shortly, man, I'm kind of excited about it, so Blue.

Speaker 2:

Sky. Do you know what Blue Sky is? It's like a social media kind of thing, or that's awkward. Sorry, I should just say no.

Speaker 1:

Put your in awkward position. So it's an alternative to Twitter. It's pretty like that, Like threats.

Speaker 2:

It wasn't that bad. I said social media thing. It was a very blank statement. Yeah, that's true, I played it safe. That's true. It made me feel like I was completely off.

Speaker 1:

I personally. Well, I have an account. I follow a lot of few people, but I don't actively use it. I have the feeling it's growing. Let's see. Let's see when. They released and I think it's quite cool. They released RSS feeds today or yesterday.

Speaker 2:

So if someone doesn't know what an RSS feed is and why you're excited.

Speaker 1:

It's because RSS feeds tie the internet together.

Speaker 2:

Oh, that's not the hot take.

Speaker 1:

I should probably know what the abbreviation stands for, right? Oh, okay, I think the abbreviation stands for really simple syndication, and it's an XML format that basically typically allows you to subscribe. If you follow the XML, you subscribe to the RSS feed, then you get popups to new events published in this RSS feed. So what does typically have RSS feeds? A someone's blog. So if you subscribe to the RSS feed of Merillo, you actually have a blog. Right, I'm going to no comment. But if you subscribe to the RSS feed and maybe in a few months you write a new article, people get an event in this RSS feed. So the other is a new blog post. Maybe a better example the Data Topics podcast has an RSS feed. You can subscribe to it. New episodes get published to it. So how can you do this? You can actually use a tool to subscribe to these things, but you also see across the World Wide Web a lot of automation due to these things. So you have a personal blog, maybe because of this RSS feed, it automatically gets published to other things, triggers, other things, et cetera, et cetera. And now Blue Sky has RSS feeds for users. So if you're very active on Blue Sky, you now have your own RSS feeds as a user, where you can basically subscribe to everything. They just say.

Speaker 2:

Which I think is very cool, nice. So then you still follow people on the Blue Sky, kind of thing.

Speaker 1:

Yeah, yeah, but, like I said, don't use Blue Sky actively.

Speaker 2:

But then it's very similar. So it's like you would follow people and then some of them you can actually subscribe to their RSS feed and then for those people you get a notification. You don't have to do that.

Speaker 1:

To me the RSS feeds. You need not really core to what they do. They're really like a Twitter clone basically, but this really comes on top. So if you're already using an RSS reader to subscribe to some blog sites, to podcasts, you can also do this.

Speaker 2:

So it's more like if you want to plug in RSS feeds from Blue Sky to something else, that's the big enabler.

Speaker 1:

Yeah, so for readers, they can easily add you to the reading list. But also, if you were the author, you can now use it RSS feed, for example with Zapier or something, to trigger a lot of other things. So I'm going to post them to Blue Sky with a tool like Zapier. I want to also take this content automatically posted to Shader, to Trats, to whatever. So these type of automations are not possible and the RSS format is extremely universal and that makes it very easy to integrate and serve, because services also often offer this via their own API, but their own API is always very custom.

Speaker 2:

Yeah, yeah, yeah, that's why RSS feeds are very nice, but maybe so this is my personal curiosity here yeah, so the RSS feeds like, if you subscribe and there's a new event, do you just get a notification for that event or I just keep? Basically, I keep requesting because I use RSS more in the. I make a get request, I got the RSS, the XML basically, and then I will parse that and then from that I can get the information on, like meetup groups. So meetup has an RSS actually, but then the way it works is that systems like the actual program, it would just keep pinging the RSS feed periodically and then you will always parse that, or is there's me very in that question?

Speaker 1:

So how I think this works and I'm relatively sure that the so if you look at the RSS feed of something, it has a lot of metadata about what RSS feeds and it also has an array of events. Yes, so each event has a GUID, a globally unique identifier, and it will, if applications that read this RSS feed will pull the RSS feed, read the full RSS feed typically and see if there are new global identifiers from the time that they read it last. Okay, I see, and that's how you know that something is new. Okay, so then they, but they read this stuff. I'm hoping not bullshitting here, but I think that's more or less how it works. That's fine, I believe you.

Speaker 2:

Okay, cool, no, but I was just wondering if, like, there's something like they just give you the last one, or whenever something they send like a web book kind of thing, it's like you need to ask it, or do they tell you? You know from what you say, you need to ask it you need to pull it.

Speaker 1:

Cool, that's cool. Typically it's also really like like an address on the on the internet. So, for example, for the date routes website, it's date routesio slash feexml. Typically it's really an address that you can pull.

Speaker 2:

I think our podcast is with bus sprout.

Speaker 1:

You have to go to review the bus sprout in domain.

Speaker 2:

That's cool. So we have a few some other things in there we don't have time to cover today, maybe next time. But do we have time for a game? Yeah, let's do it again. Let's do it. Hit it, All right. So last time, Mariam, she won. She was the big winner.

Speaker 1:

So Mariam was one of our previous guests, right, yes, another previous one, the one before that, back in 2023.

Speaker 2:

So long ago. Yes, so she won. So I kindly asked her would you like to come up with three more quotes? And I was like she's like yes, for sure this time. She did not tell me the answer, so she sent me an email with it. I haven't opened it yet, I promise Okay.

Speaker 1:

Okay.

Speaker 2:

So I'll play the game with you today. What is the game so? What is the game? Good question. So basically, there are three quotes. One is AI generated Jena, ai generated. I guess it's a bit.

Speaker 1:

Yeah, exactly.

Speaker 2:

Basically, one is fake to a real, or sometimes we mix it up to a real one or fake and it's our job to guess the real one.

Speaker 1:

Okay, let's go for it.

Speaker 2:

Let's do it. So the person, the character for today, is Samwise Gamke, I don't know, from Laura the Rings. So yeah, I don't know. I don't remember how to say last name. No Google in Bart.

Speaker 1:

No, no, maybe need to put a character on.

Speaker 2:

You're gonna put a character on Good idea.

Speaker 1:

Actually, you have a screenshot. You do it. That's a good question. How you are, are you pronounces in English Gamke? I don't know.

Speaker 2:

I feel like I should I mean definitely watched it right.

Speaker 1:

But so you're gonna give us three quotes Exactly Two are fake, two are fake, one is real. We need to guess the real one. Yes, exactly, exactly.

Speaker 2:

Okay, so have it here, little Sam.

Speaker 1:

Nice, nice.

Speaker 2:

Okay, all right, you ready.

Speaker 1:

Are we both gonna do it, or are you not the answer there? I do not know the answer Okay.

Speaker 2:

So we're both gonna do it. We're both gonna do it. Okay, Are they together?

Speaker 1:

I need to trust you on this, that you don't know the answer no, yeah, yeah.

Speaker 2:

Okay, I'm a trustworthy guy, bart.

Speaker 1:

Can you trust me? Like I'm not trustworthy, I'm starting to find it suspicious that I never win this game. But okay, let's go.

Speaker 2:

Let's do it so. First quote is even the smallest person can change the course of the future, just like a pebble creates ripples in a pond Deep.

Speaker 1:

What do you think? One real to fake? Right, I think, yeah, one real to fake Could be, could be.

Speaker 2:

You think?

Speaker 1:

Maybe it's a little loud.

Speaker 2:

Yeah, it's way too much on the nose. Yeah, it's way too much part, I think. To me it's a clear fake. Why, would he say, the smallest person Like it.

Speaker 1:

He's a small guy.

Speaker 2:

Yeah, but like Jej Pt clearly saying that like, think of a character their hobbits, right.

Speaker 1:

Why would he say like, oh, the smallest person, Because they're like they're also bigger things in Lois Strait.

Speaker 2:

But I feel like it's too. It's too like if you ask Jej Pt, they're gonna pick what the derriest mole.

Speaker 1:

It's very easy, that is true. Yeah, like what is the what is a clear characteristic, Exactly?

Speaker 2:

Right, If you say something, if you were a hobbit from Lord of the Rings, first thing to say is like either the weight of the ring or whatever, or they're gonna say hairy feet or something and then just say small, so that's, that's that's, that's my thinking. That's my thinking. I mean, or I'm just trying to fake you, huh yeah, that's why I'm not still, I'm still loud. You never know, I'm a master of disguise. Second one. The grandest tales await those who embark for in the realm of accomplishment. It's the journey that casts the longest shadows.

Speaker 1:

This is really something to write on a tile and hang up somewhere, right yeah?

Speaker 2:

I was at a hard time reading it.

Speaker 1:

I have to feel like this is, or like an impressive background and like as a quote, and you put it on the background. You have this picture. You need to have it on the as your background, as your wallpaper, when you're like 16 years old. Yes, yes, every day, like you wake up and then a few years later you think Megan was a bit weird. Yeah, this.

Speaker 2:

Yeah, are you saying for personal experience part, it feels like no, no, no, no, okay, okay, what's your take on this?

Speaker 1:

I'm going to listen to the third one.

Speaker 2:

Okay, I'll already tell you that I don't sounds funny. Grandest Tales to me is like grandest. This is like I don't know. Yeah, he was small, huh, yeah, but that's again. It's the same thing. The only thing I talk about is small and also it's like the journey shadows is because, yeah, like it's a dark thing Like the. What's the name of the? The, the devil guy in Lord of the Rings, south owner, or whatever. In any case, last one is the job that's never started as takes longest to finish Is with the, the, the Megan. It's the job that's never started as takes longest to finish. I'm reading word by word but I'm not sure I understand what he means, but I think that's the correct one.

Speaker 1:

As stakes, that takes longest to finish.

Speaker 2:

No, it's, literally it's. It's the job that's never started as takes longest to finish.

Speaker 1:

That takes longest to finish. Maybe I think it's simple as to the point. There's a typo. It's probably not gender, cgpc gender, so you're. So, yeah, my name is Karim based in the city to the movie. This is the quote that Mariam had on wallpaper when she was 16 years old.

Speaker 2:

Maybe we got her. I'm going to go for this one.

Speaker 1:

This is the real one.

Speaker 2:

This is the real one too. But I'm wondering, like, if we should play safe and play the spread.

Speaker 1:

You know, I'm sure you're sure this time I'm going to have it correct.

Speaker 2:

I think so too, because that's the one I would pick. But I want to play the spread because I read the one Mariam to in twice in a row.

Speaker 1:

You can have a way of explaining it.

Speaker 2:

I'm going to guess it's the second one. Okay, the second one is the real one. Okay, and drum roll, I need to find that email. Okay, I need to find that email I had just here.

Speaker 1:

Yeah, I know right, you press delete, okay, okay.

Speaker 2:

The correct answer is the third quote. There we go.

Speaker 1:

But only I guess it. That's true, it's just me.

Speaker 2:

Really, that went for it, so that means that next time which is funny, right, like you were the very first person to be the puppeteer to create the quotes you started- this, that's right. And now, years later, you won the game for the first time, and you and me, you lost. I thought you were going to go the croix, croix, croix. Is that what you wanted? Yeah, that's true. I did lose, so that means next time. I need to generate quotes. Next time is generate quotes. It's the first time you win, right, so it's. I'll explain how it goes.

Speaker 1:

Yes.

Speaker 2:

No, no, but indeed. So that means you're the big winner. That means the next time you get to come up with the quotes and try to fool all of us. But I just want to say that I also thought.

Speaker 1:

Do we have a certain team to work with?

Speaker 2:

Any, I think we did Give me one.

Speaker 1:

Give me a team.

Speaker 2:

I think we did the first one. I think the first one was rappers. Maybe no, but then let's do something else. Let's think of fictional characters, anything.

Speaker 1:

Very broad Okay.

Speaker 2:

Let's do something like a fictional character from something you watched recently with your kids.

Speaker 1:

What am I, kids? Peppa Pig.

Speaker 2:

There we go.

Speaker 1:

That's hard to do. I'm rewatching a Star Trek series, okay.

Speaker 2:

Okay.

Speaker 1:

Okay, cool, okay, that's it.

Speaker 2:

I think that's anything else we want to share.

Speaker 1:

Before we call it a pod, no, no, hey, enjoy the vacation.

Speaker 2:

Oh yeah, thank you. I'll see you after that, hopefully more 10. Thanks everybody for listening.

Speaker 1:

Thank, you See, ya, ciao, aiyeee börné Games. I did it. Tchaoooo, יleigh.

Using iPhone as Webcam
Web Browsers and OpenAI's Approach
Impact of Fake News & Technical Debt
Technical Debt and Blue Sky RSS
Guessing Quotes From Fictional Characters