The Changelog: Software Development, Open Source - The Wu-Tang way (Friends)

Episode Date: May 3, 2024

Our friend Ron Evans is a technologist for hire, an open source developer, an author, a speaker, an iconoclast, and one of our favorite people in tech. This conversation with Ron goes everywhere: from high-altitude weather balloons, to life on Mars, to Zeno's paradox applied to ML, to what open source devs should learn from the Wu-Tang Clan & more.

Welcome to Changelog and Friends, a weekly talk show about high-altitude balloons. Thanks to our partners at Fly.io. Launch your app near your users all around the world. Fly makes it easy. Learn more at Fly.io. Okay, let's talk. Today we're joined by the incomparable Ron Evans. Ron is a technologist for hire with the Hybrid Group, an open source developer working on TinyGo, GoBot, GoCV, and more, an author, a speaker, an iconoclast, and one of our favorite people in tech. Adam and I have both loved talking with Ron since the first day we met him many years ago, and we think you will too. Okay, here he is. I am talking to you through outer space, which we can discuss as well.

How so? Well, I'm on Starlink. Oh, so all of your words are going up into the air and then coming back down. Yes. Before they get to us. It's amazing. Do you know exactly how high those things are? Because they've solved the latency problem of other satellite internet providers, right?

Because they're not that high. That is absolutely true. They're in a lower orbit than the Ka band that the original direct PC, Hughes aerospace satellites. Those are actually a lot further away. The Starlink constellations are relatively close. That's why you can see them with the naked eye, much to the chagrin of astronomers. And anybody who likes to look up in the sky at night, you know, now you're seeing all this stuff flying around out there. Well, I look up in the sky and I like to. I also like to see human works flying around in space just because it inspires me.

Okay. It makes me think, hey, the future is actually here. You know, it's not. By that, I mean, I find New York City or Madrid or Hong Kong, many places, you know, Los Angeles at night. Yeah. If you're flying to Los Angeles at night, that's the actual LA you want to see. Because it's, you know, you're flying for an hour over just endless lights. And you're like, an hour over just endless lights and you're like where's

pretty amazing where does the city begin so i mean i find it of our human works there's a beauty to it you know i'm not oh tear down human civilization no i'm like build up human civilization now i understand the problem of littering space with a bunch of junk that's a different yeah eventually like la you know the smog and then eventually we have so many low earth, low orbit satellites up there that I don't know, maybe we never have darkness again or something. They don't exactly, they reflect light, but they don't put off light. Do they? You could just send up a high altitude balloon and just hook on the one and

just, Whoa, there we go. That would be interesting. Well, actually one of my next balloon projects i'm hoping to connect to some satellites actually how many balloon projects do you have ron oh my gosh what am i next i know he just assumes that we know this guy has balloon projects please tell us oh well um so last year was my third year in the row of speaking at FOSDEM, the awesome free and open source conference in Belgium, in Brussels. And it was my third edition of the Go Without Wires saga.

Ah. Right. First year, it was Go Without Wires about Bluetooth. Second year, Go Further Without Wires. OK. About local area networking with Wi-Fi. Last year, it was Go Even Further Without Wires. I think you're running out of titles, though.

Oh, no. No, not even close. Okay. So it was about long-distance radio networking using LoRaWAN, which is a standard for using unlicensed spectrum, you know, that's free to use for anyone in the public. Oh, I haven't heard of this. LoRaWAN.

Yeah. LoRa, like long distance radio. Okay. And then on top of that is a routable protocol called LoRaWAN, like wide area networking. And the real pioneers in this space are the Things Network, which is a open source slash commercial open source organization that has a whole network of these routers. Basically, you can connect to these local long range radio networks, and then it has a backhaul to the internet. That way, you don't necessarily have to have your own private network. It can just get routed through the Things network to whatever cloud servers that you want to.

So this is, of course, really useful for applications in industrial or agriculture or smart cities. They have quite a few of them um there's a bunch of telecos that are also participating in this so the finale of my talk was a high altitude balloon program with tiny go using a raspberry pi rp 2040 pico board since these tiny balloons are called pico balloons so i'm very literal i thought oh i should build one with a raspberry pi picot i mean like obviously yeah like it was in the instructions you know as far as i was concerned right it's in it's in the title it's right there in the name yeah it's in the title wrong so i had a long distance radio i had a gps i had an

accelerometer it was all hand wired using a proto board and wire wrap, just because I got into this retro computing thing. And it happened to be the weekend that everybody was talking about the balloons of Chinese origin. Oh, right, that were floating over the US? Yeah. Now, to be fair, OK, all parts are of Chinese origin. So it's just sort of like a-

Hard not to find Chinese origin. Yeah. I mean, certainly anything that's in the toy category. So less than 250 grams, you don't have to file a flight plan or anything. It's basically toy balloons, which literally I had ordered from Amazon and other such places, got at toy stores. I mean, I only work with toys now generally. I mean, it's the only thing you let me have anymore.

It's a policy. Just toys only. Like if you just give him toys, it'll keep him busy and he won't get in too much trouble. So how high up would this balloon go? So this particular balloon, it was on Hackaday. You can get some of the stats. It was called Tiny Globo, the first one.

We released it on Sunday there. And we did a countdown. We released it. It reached an altitude of about, I think, somewhere between 6,000 and 8,000 meters. And it traveled approximately 400 kilometers from Brussels, south of Orleansleans france before we lost contact so i don't know if it was the french air force shot it down or if the battery died or you know something i haven't gone back to france since then you're not going to collect this hardware you're just well no it didn't land it just kept floating that was just where we lost contact how do you know it didn't

land ron well i mean i imagine it landed somewhere sometime okay well you just said it didn't land it just kept floating that was just where we lost contact how do you know it didn't land ron well i mean i imagine it landed somewhere sometime okay well you just said it didn't land so i was like what well at some point you don't know where it landed no no but it was very successful first flight got telemetry the whole way um every 30 seconds thanks to the things network um it proved that the long distance radio networking and support in TinyGo actually worked just because, yeah, I mean, it was way out of my hands. So that was very cool. Then we did a reprise of it at the GopherCon EU in Berlin in the summer.

That launch was a little trickier. It traveled just, it was just a little ways north of Dresden when we lost contact. I'm like, oh no, not an American balloon going over Dresden. Like, you know, I won't be able to go back anywhere. Right. That was, the wind conditions were a lot more difficult. So both of those were battery powered and my next one will be solar powered. That way I can actually circumnavigate the globe and ideally connect to some of the new satellite networks that are being built that will be supporting this same LoRa protocol.

Wow. So yeah, I'd like to go around the world in 80 days with a toy balloon programmed entirely with TinyGo that's sending telemetry data. Just as far as citizen science and reducing the cost of entry for experimenters and explorers and people to find out more of the things that they can really do, this is a great.

So these are some of my hobbies and also a great way for proving out some of the technologies like if it yeah if it works here it'll probably work in a more conventional you know farm vehicle or something that's you know of terrestrial origin it's really pushing into the extreme being that high i'm looking at photos on hackaday and it looks like it's got not just one balloon but multiple. Seems like maybe three, maybe four balloons. Is that right? Like they're clear balloons. What's inside the balloons? How did you design this? Give us the details on the exact innovator's dilemma of making this real and going in the air like that. Well, like any wise person, I go onto the internets and find out what other people have done before I do anything.

Okay. So you watch YouTube. Yeah. There's a whole community of Pico Balloon enthusiasts who are doing. Really? Oh, yeah. Of course there is. Of course there is.

Now they're doing amazing things. So I borrowed some of their design ideas. Specifically, the balloons are toy balloons made of vinyl. And you have to stretch them out. They're filled with helium that you could just get in any, you know, balloon type party store. In fact, that's where I get my helium is from party stores pay with cash or the party

store. No, no. Yeah. So you, these, uh, they're transparent ideally so that the, you know, sun can get through to charge whatever solar cell. This one, I didn't have the solar. Right, which is beneath it. I didn't have the solar thing set up. I actually had the equipment, but I didn't try to use it because I had some software problems, actually, with the low power mode. So I had it working to charge, but it would discharge in about five minutes.

So I couldn't, you know, the idea with these things is it, you know, recharges when it hits the sun, wakes up, sends some telemetry data, goes back to sleep, and so on, you know, as long as it's in sunlight, you know, generally done with a thin film solar panel and some super capacitors. So I had the design. I built it, you know, again, from specific parts that I happened to have on hand and I was able to do with the wire wrap. But these balloons, you stretch them out, you know, since gas expands when you hit higher altitudes, that's the reason why there's four of them, is that each one is only partially

filled. You'll fill them partway. That way when they reach, you know, what I thought was going to happen was that it was going to go a lot higher and then pop. What instead happened was it reached apparently an altitude where it was just stretched out enough to not lose any gas or have any bursting. Perfect. So yeah, beginner's luck on that one.

The second, it was a little harder just because there were a bunch of big trees right there in the venue. So that's always very exciting. So you want to circumnavigate the globe. Do you want to be able to then control these things, like drive them? Or are you just saying you send it in a direction and hope it ends up going around the world? Oh, yeah, there's no controlling.

They're just released, no catch. You know, they just go. Generally, this type of high-altitude ballooning is done either by people doing this sort of toy balloon thing or people who use latex balloons with a payload with a parachute. With those, you actually file a flight plan to be responsible to make sure that you're not going to interfere with any air traffic. Yeah, absolutely.

That's something, I mean, just because there's no actual law preventing you doesn't mean you should just go and do something stupid. Yeah, there's wisdom there. He's earned that wisdom. A tiny modicum of common sense will perhaps, like if you think it, I mean, you can't think of every possibility. Sure. But it's certainly a good thing to you know to not you know release like balloons

right next to for example the you know u.s naval air station in san diego which is why i did not do that at gopher con us i'm like you know this is both stupid and it's also very rude like impolite like you know you should know better and also you should just know better yeah definitely exciting definitely exciting jared i think you'll also get on the wrong people's radar, potentially. Yeah, no, you're definitely on the wrong people's radar right away. Because you're already not filing flight plans, which is you're okay with the grams and the weight, but at the same time, if you did this with, let's just say a paper trail,

not paying cash at the party store, for example, you could be called upon for some i'm a known white hat which is why they would be so angry oh okay oh because you'd be darkening your hat up doing that no it'd be just a matter of like i actually do know better right that's true yeah now i mean you know this is also a matter of like, what is your purpose in doing this? My purpose is to help show young scientists the opportunities they have to do some, you know, exploration in a way that costs very little and that gives them a chance to, you know, to learn quite a lot.

But these are things that are quite important, you know, not just in the future, but really in the present. Yeah. You know, a great example of citizen science being a lot better was a cool project a few years ago when the Fukushima incident occurred called SafeCast. And it was basically being able to turn your mobile phone

and connect it to turn it into a homemade Geiger counter. And that way they could get a lot more accurate readings of the type of data. Just because the government would go in, they would take their readings at a certain meter height, at a certain frequency, and it was a relatively small data set. So this allowed a bunch of citizens to start collecting data. And one of the things they found, for example, was that there was a lot higher concentration of radiation lower to the ground than at the two or three meter height of these

poles. So that was one reason why it tended to affect younger people, pets, and animals, and people, you know, who were in wheelchairs just because they were lower to the ground. And so this higher concentration affected them more than adults. So this is data that they wouldn't have had if they didn't have, you know, just people who were interested, who cared about this. So, I mean, it's a great example of open science,

open source technology, working for social good. You know, to me, that's the sweet spot of, you know, why we do this. Well, CloudFlare's Developer Week is over, but there is so much to cover from that week. And I'm here to give you a roundup. So here we go. Their fully distributed serverless database, D1, went GA.

It now supports 10 gigabytes of data, and they added new exporting solutions and insight tools. HyperDrive, which accelerates your Postgres and MySQL databases, also went GA. You know that monthly workers' pay plan they offer? How much does that cost? $5? Yeah, $5. And how much does it cost to completely speed up your database operations using HyperDrive? Zero. Yeah, $0. With Qs you can now send and acknowledge messages from any HTTP client. They also

added the ability to add delays. Yes, they added delays. Workers Analytics Engine, which provides analytics at scale. Flip the GA switch too. They launched a brand new AI playground that lets you explore all the hosted models on workers AI, which, by the way, also went GA. That's right. Production grade global AI inference that you don't need to deploy.

All available seamlessly in workers or directly from a REST API call. They also announced a partnership with Hugging Face, so you can now quickly deploy an app using these models. And FineTunes are here, y'all. That's right. They offer LoRa support. Upload your FineTunes from Wrangler and apply them to their most popular LLMs. There are so many ways for you to build with AI using Cloudflare. It's awesome. And Cloudflare doesn't support Python? Wrong. They do now.

Python Workers is here. From the same command using Wrangler, you can now launch a worker that can fast API, langchain, numpy, and more. R2 got event notifications. You can get notified now when an object is created, changed, or deleted and handle that event in your worker. And who says you can't spell SDK without SDK? Craig did. And that means they have new SDKs for you to use. TypeScript, Go, and Python. And that is just a few announcements from Cloudflare's Developer Week. Check it all out for yourself at cloudflare.com slash developer week. Once again, cloudflare.com slash developer week.

Once again, cloudflare.com slash developer week. Can we talk about that Geiger counter for a second? Because I'm curious about this because like if you, are they readily available? Can you buy, can you just go buy one? Oh, sure. Sure. just can you just go buy one oh sure sure you know and this was um using i believe some photographic parts that you'll have to safe cast i don't know all the details i did not make one of these myself you know i got enough problems without ordering parts that could be used to

make a geiger counter you know well did you explain what the use of a geiger counter is so that the folks who don't know because Because I'm like, I'm barely initiated. So for the barely initiated or the fully uninitiated, what is a Geiger counter? And why would you want to include one to gather data given this kind of mission? So a Geiger counter is a device whose purpose is to be able to read the ambient level of radiation in a local area. And they're generally used when you're working around sources of radiation, whether they're natural ones like uranium, you know, ores, or artificial ones, you know, the x-ray machine, for example. Now, there's a reason why when you go in and get x-rays that the technician, first of all, they leave the room.

That way they're standing on the other side of a lead shield. Lead is a very dense element. And so the x-rays are unable to go through the lead. That's why they give you a lead apron to put over your sensitive parts when you go in for one of those x-rays is to protect you from this radiation just because repeated exposure this is actually one of the reasons why it's going to be quite difficult for us to leave the earth and go to for example colonize mars is the colonists by the time they got there would be barely alive if alive at all just because of the radiation in space okay so this so a gyro counter is one of the original, it's a very old device.

It's a very simple device, generally speaking, that is able to measure this ambient radiation. And so the SafeCast team, they were basically using an Arduino connected to some salvage electronics parts to be able to then connect this to your mobile phone and take readings that you could then upload to their you know web-based data store yeah i'm glad you

explained that because you did a much better job than i would have done explaining a geyer counter although i do understand the radiation and measuring it i was surprised i've never looked into this to know it's readily available where you can just maybe go maybe not readily literally available like in literally any radio shack or go buy it at Target or Walmart or whatever. But they're available to maybe hobbyists and purists who want to like look into this. But the reason why I think it might be odd for any given civilian to buy this thing is because you can learn things about the earth and its atmosphere. And let's just say scientific beliefs that may not actually be factual in our world like you can learn for yourself as a scientist would for example even

specific to the radiation belt that's around the the earth that you say prevents or could prevent human travel to mars or let's just say the moon you know just even staying in a high orbit right you know i um the van what is it the van what's van allen is that right van allen belt yeah that's that's one of several belts right ionosphere which protect us from and it's a very thin i mean if you look at the actual atmosphere relative to the planet like it is a thin little tiny piece of you know transparent it's not even aluminum nothing basically it's like it's like jello it's like this tiny little skin you know right like an eyeball almost like a membrane around an eyeball yeah that's just protecting us from yeah you from literally getting cooked.

When you said that about Mars, do you lean the way of exiting Earth as humanity may or may not have? Are you a conspiracy theorist? Which way do you lean when it comes to humanity's departure from Earth, and have we or have we not? Oh, I think we're a long way. There's a great, I can't remember the name of the couple that wrote this book it's about you know how to colonize mars basically okay that's not the actual title sure we'll have to look that'd be a terrible title but phrase version so they're they started out super optimistic they're like yeah we're going

to see this in our lifetime and by the time they got to the point where they could actually write the book, you know, the enormous reality check of the distance between the technology we have today, you know, and the technologies that we actually would need is quite vast just to keep our, just to keep people alive between here to there, let alone- Then once you get there. Like when you get there, now what, right?

Yeah, whole different problems. I mean, if you live in a modern city, you're already drinking the refined, but water system in your city is already purifying people's urine and you're drinking it today, okay? Don't tell me that, Ron. Come on now.

This is the municipal water supply in most major cities okay but it's the but it's also got other sources right on mars it's like this is it you know it's the only water we have right this is the only water you know we've got did he answer my question though jared or do you think he he's talking mars he's talking mars circumnavigated so do i i think that it's a great thing for the human exploration of space but the reality of where we'd actually need to be to get there is a long way from this you know spaceship earth is the spaceship this is the place we have, preserve, and improve this one because it's going to take quite a few more generations

before there's any reality of being able to send humans to survive even one generation. It's going to take an incredible collective effort just to do that little tiny bit. And spaceship Earth is the place so right you think we could practice on the arctic you know like let's survive cold and like colonize the arctic and the antarctic we have there's people there but very few right like we could let's build cities there and thrive there where it's very harsh terrain and and environment but hey they have oxygen which which is nice, you know?

And if we could do that, we could, you know, if we can do that, maybe we could do something like Mars. But until then. We had one, it was called biosphere. Oh, is that a Pauly Shore movie? That was a movie, right? Biodome. The Biodome, a pure self-contained environment where five scientists are about to be sealed off from every conceivable form of contamination except one. Yeah, Biodome.

Biodome was probably one of his better movies, which is really frightening. Which isn't saying much. But that was actually based on the true story. Hard to believe. So they built it in, I think, the Arizona desert. The idea was to have a totally self-contained system, where I forget how many people, like seven people entered it or something.

How many walked out the other side? No, they all left. But they really hated each other by then. They didn't actually speak to each other for several months. Wow. You know, one group and another. You know, it was kind of an amazing experiment.

It's kind of like the real world. Remember that old MTV show, The Real World, only this is inside of a biodome. Oh yeah, it's even more real. That would have made good television.

Yeah, they'd still be doing it if that existed at the time. Yeah, really. You know, maybe somebody should pitch that. That's actually kind of a cool idea.

But I mean, they did learn quite a lot. One thing they learned was having so many different environments was very impractical. That having a smaller number of environments was more practical for being able to try to maintain continuous ecosystems. Because they had to do a lot of maintenance. I mean, it was full-time maintenance work.

Did they try to have a desert and a forest? Did they have all these different? Yeah, it was right it was full-time maintenance work did they try to have a desert and a forest and they have like all these different yeah it was like it was some 70s movie you know where it was you know kumbaya like hey man hippies in space you know right and then the reality is like yeah but i mean man we got to do weeding in space like all day all day we're like all the plants we're trying to grow are dying and the ones we don't want are taking over and we're in space like what do we do now you know so but the the worst thing about it though is they didn't keep going like they stopped they did one run they're like wow that was really crazy and they shut it down you know that they needed to keep doing that for 20 30 more years right to really learn something as far as you know science you know there's a

big difference between r&d for a scientist and r&d for businesses you know scientists they go and they do a bunch of research they spend a bunch of time they come back and they say after spending my entire life i've discovered that the answer to that question is no right and everyone's like oh noble prize amazing okay now you try to do that when you're working for some company forget to your whole life's work like two months later your boss's boss comes in oh how's it going like oh right i'm starting to think like we can't do this like what program's canceled you're fired everyone right you know like you know the answer was yes like we were just asked we were

paying you to figure out how like this is not science that's one of the hard things about science and about research is that there are certain incentives right the where the capital comes from there are certain tests and certain things that are never going to be funded in order to be run and so you're not going to have you you don't get, you know, credentials by failing, but not a failure. Like proving a hypothesis false is not a failure, but when it comes to getting more money, it seems like it is a failure. You know, nobody wants to say no. This is, segues perfectly to the conversation about machine learning, I think. Let's do it.

Let's do it. Take us there. Just because, yes, if your discovery is that something isn't going to work, this is an enormous contribution to the community knowledge of humankind. Because you figured out, oh, let's try this other direction. But this is not something that's commercially exploitable, generally speaking, and we need quick results and this is one of the big problems with fundamental research into things like machine learning and commercialization of things which must be productized now whether they work or not we just need you know monthly recurring revenues like you got to do whatever it's going to take to get me there and if you don't you know again you're quite fired you know yeah But do not look at the man behind

the curtain. And it's a classic case of Roy Amara. Roy Amara was the guy who said, we have a tendency to overestimate the short-term benefits of technology and underestimate the long-term ones. I'm going to make an unpopular prediction. 90% of the companies that are trying to do machine learning will fail. And the reason they'll fail is because they won't actually

produce anything of value whatsoever, perhaps of negative value, right? The other 10% will survive, not because they're actually doing anything worthwhile, but because they'll figure out how to do something. You know, and this is just... That's all part of the process, right because they'll figure out how to do something. And this is just...

That's all part of the process, right? Right, that's part of the process. It just gets distorted a bit by large sums of money. There was a great video. Every entrepreneur is putting machine learning in their investor deck. Just like, oh, put it on. For sure.

Whether or not it makes sense. Every company is an AI company now. You know, and this is kind of throwing away the baby with the bath water as far as, okay, a bunch of this is sort of nonsense, but that doesn't mean we shouldn't be trying to learn something from it. But with so much noise and so little signal, it makes it quite difficult to sort it out. I did a talk in spanish called that machine always lies i haven't never done it in english hopefully i'll get a chance to do it at some point but it was a takeoff on you know my that my brother always lies

you know the logic puzzle so it's one of these um like game things? It's more of an example of exclusive war. So you are a traveler, right? Some versions of the story, you're a soldier trying to go home from the wars. Others, you know, you're a traveling merchant. You come to a crossroads. And standing at the crossroads are two identical twins. So one of them, there's a sign. It says, one of these brothers only tells the truth, and the only one always lies.

OK, this is ringing a bell. You can only ask one question of one of them, and you have to figure out which way to go. So it's one of these classic logic puzzles that I think it's been around since maybe the 1600s or maybe longer. It probably was around longer. It's just they never got around to writing it down. What's the move? What are you supposed to do? How do you figure it out? So think about it this way, right? One brother will always tell

the truth. The other one will always lie. You're only allowed one question. The question is quite simple. Which way would your brother tell me to go? You ask one of them. You ask the one that tells the truth, and then he... It doesn't matter which one, right? If you ask... Both of them will give you... Exactly. So this way, the one that always tells the truth will tell you about the lie of the brother who always lies. The one who always lies will simply lie. So you go the opposite way of the way that either one of them tell you,

and that's the correct way. Couldn't you just ask the one that tells the truth, which way should I go? It doesn't, you don't know which one is which. Oh, you didn't say that. That's right. Well, because- You said it far and I missed it.

That makes sense. Well, no one thinks, I think, that logically the one who does tell the truth can overhear you asking the one who may or may not be the one who does tell the truth the questionar you asking the one who may or may not be the one who does tell the truth the question so you can say which like he had said because you can ask one of them it's a 50 50 chance this person tells the truth the other one is a 50 50 chance of being a liar and they overhear it that's 100 chance it's if you ask that question well no

it's perfectly consistent in this example. You know, something like exclusive ore. Maybe a better example would be Zeno's paradox. Zeno's paradox is the oldest paradox I know of. So Zeno is from Crete, okay? And Zeno says, all Cretans, people from the island Crete, are liars. So now here's the paradox. If he is telling the truth, and he is from Crete, then it is impossible that all of the people from Crete are liars because he simply told you the truth.

On the other hand, if he did not tell you the truth, so you see this is known as Zeno's paradox, and we're just touching upon the paradoxes which exist in human thought. It's not perfectly rational. So how to encapsulate that and reduce it down to any form of quantitative calculation is something which we actually don't really know how to do. So calling it artificial intelligence is out of the gate. My brother always lies. That does not mean these things are not useful, potentially, and that we should not learn from them.

But pretty much everything you've been told is not actually true. And a great example of this, I mean, let's say you have a company. Your company does some database sorting. Don't you want people to think like this database sorting is so powerful that could literally destroy human civilization as we know it. If you buy this database index sorting, this query engine, it literally could destroy society as we know it. You as a CEO would be like, how do I get my hands on that thing? I want that baby. You wouldn't be like, oh my God, we could do a database query that could destroy our entire planet. You think like, man, I got to get that before the other guys.

If you're selling this database engine, you literally want people to think it's as if you have nuclear fuel inside the database engine. Now just substitute machine learning for database engine and I believe we have a pretty good view of what you know the the distance between hype and reality you know and how that is a great marketing like everybody wants to get this thing on their side get get it away from their, you know, their competitor, you know, and that's just before I, we go into some other area. I think it's great to read the paper that really started the whole controversy from Timnit Gebru, the researcher at Google, the paper is named On the Danger of Stochastic Parrots. And I don't read a lot of academic papers.

I'm not an academic. I'm on the technology side. I'm the person who's looking for other people's research to turn into my development. Because I have to go to the money people and say, yes, this is going to work. These people proved it. Not, well, we don't know. So I'm looking for research that I can turn into something. Yeah. Because I'm a technologist. That's my thing. So I don't read a lot of academic papers, but I've read a few. This one's quite short. And as academic papers go, it's quite concise and quite clear.

And by the way, it's the paper that got her fired from Google with much controversy. And this is before the whole open AI thing came along and sucked all the oxygen out of the air as far as the conversation about machine learning. But I read this paper about, I don't know, maybe a year ago, a year and a half ago. And the first thing I thought was, yeah, okay, cool. But the next I'm like, wait, they got fired for this? Like, and there's nothing in there which could possibly be objectionable. You know, the ostensible reason was for lack of academic rigor. Well, summarize its statement.

What is the abstract? Essentially, it was saying that these large language models, because that was the generative AI that the researchers were concerned with, that they need to be looked at as far as what is the source of the training data, what are the implications for its use as far as on society as a whole and on specific working groups and people affected by it, and that we also need to look at the economic and carbon cost of the computation necessary to to train and use them that was it okay it just said we need to look at these things so nothing really controversial there is a bit

more than that based on using a large language model to discover what this paper might be about it said environment impact bias and fairness misinformation and harm lack of transparency and economic and social impact. So it's a bit deeper than just simply environmental. Okay, I thought I summarized a few of those, but that's a good summary. Yeah, I didn't summarize the summary. It just like it was more than, it seemed you mentioned just one part. That's why I was adding two.

My bad. But it was not saying we should not do this because of these things. It was saying we need to analyze on these axes. Right. Here's a framework for analyzing the impact of this particular technology set. Yeah, which I agree with. I think those are all fair. There's nothing to disagree with.

I read this paper twice. Right. First, just because I knew of it. And then the second time, just because I couldn't put my head around why this was such a big deal. But then it made me think, it's very hard to get a person to understand a thing when their job depends on not understanding that thing. Oh, yeah. What are you talking about here? Right.

I don't know. No, nothing to see. Nothing to see here. Move along. So, you know, and this is where the open source machine learning, you know, to me is such an important, you know, genuine open source. You know, when Lama CPP came out. So when Meta AI's Lama first leaked, I found it absolutely fascinating. Because first of all, these things don't leak.

Have you ever signed a nondisclosure with one of the really big companies? I mean, let's keep in mind that one time, one guy had an iPhone that he had at a bar. And literally, the San Jose police came with guns drawn face down on the ground, tie wraps, dragged off to the jail, like this guy was some type of bioterrorist, right? That was a big controversy. I remember that. Right. Okay.

Moto, I think, right? Right. And this is not, like, that's very extreme, but just dial it back only the tiniest little bit, okay? And this is like, there's more lawyers than they have people in your whole social network, right? Like like a thousand lawyers work there right exactly so there was something leaked and somebody was i didn't hear about this story there's somebody something was leaked a lot of meta and they were arrested no no no no no i was using this as an example okay guys

gizmodo deal back in the day where the person who worked at apple took their iphone to a bar and the gizmodo people got their hands on it. He was saying that that person was heavily handled, heavy handed. The big hyperscalers have not gotten more friendly and nice since then. So when Lama, the model, first leaked, ostensibly it was given to some researchers under specific agreements to not publish anything from it. Within, I don't know, 72 hours or something, it was on torrents and people were downloading it. And all of a sudden, it was everywhere, right? Now, I don't know anything, but it just seems to me like there's no human being brave enough to say, it was everywhere, right? Now, I don't know anything, but it just seems to me like there's no human being brave enough to say,

oh, yeah, no problem, they won't catch me, like, you know, with these watermarked things. So I have a slightly different theory, which is a bit more, you know, when you're losing a game of chess, or you think you are, like some people just, or poker, whatever, you just flip over the board, just like

ah you know we gotta start over yeah right well why would they leak at them so so you're theorizing perhaps that long that meta let it leak out on purpose couldn't they just release it though because why leak it when you could just release it you get more goodwill by saying hey this is you know oh that's what they're doing now then they were adam you asked me about conspiracy theories i'm trying to give you one here man come on i'm trying to give it i'm giving you what you Because that's what they're doing now. After a leak, then they really. Adam, you asked me about conspiracy theories. I'm trying to give you one here, man. Come on. I'm trying to give you.

I'm giving you what you asked for. No, plus it's a PR thing. I mean, you're losing the mind share. I mean, these are a bunch of really serious researchers who've done a lot of good work over many, many years. These are not just randos like myself. These are actual people who know things. And all of a sudden, everyone's like, oh, open AI, open AI.

And if you're on that team, you're probably not too happy about this. I mean, research is very competitive. Whether it should be or should not be is, I'm not an actual researcher. Sure. So Lama leaks out, potentially on purpose. whether it should be or should not be is, you know, I'm not an actual researcher, you know. Sure. So Lama leaks out, potentially on purpose. So Lama leaks out, and it really sets off the Cambrian explosion of suddenly new things popping up day after day after day. People take this model and quantize it, and, you know,

then Alpaca is a smaller, lighter version of Lama, you know, and then, I mean, every two days, something new is coming out. So one guy just on the internet says, I wonder if I could run this on my MacBook M1. Like any of us, he decided to devote his weekend to it. Sure enough, kind of got something to work. Because, you know, usually you'd be like, oh, I'm going to stand up some type of cluster of machines with a bunch of GPUs. And, you know, this is kind of the big, it's a big person's game with lots of hardware and money to throw at it. So this person's like, no, actually, I could run this on my laptop, gets it to work. All of a sudden, it changes the game yet again. So I made a few early contributions to that project

just because this is very near and dear to my heart. Take the fire and give it to all the humans and damn the consequences because if we don't, it's going to be so much worse than if we do. Now, I'm no Prometheus. I'm going to be running like hell. Plus, I wasn't the one to actually stall the fire initially i just passed it out to a couple people along the way you're just don't blame me you know you're just the distributor i'm just in the crew man you know you want the big boss over there right but i mean it really is setting off was the beginning of a whole you know know, now there's oh so many and there's so many

things happening in so many areas. And the vast majority of the interesting work is happening in the open. And I think this is a very good thing because it means that the means of computation, you know, sees the means of computation as Cory Doctorow says. You know. I don't think it's good to, on the one hand, limit this to only big companies that have the money to do something with it, or on the other, to have governments that are regulating it in a way that favor those big companies to the exclusion of independent and interesting

and innovative things that are happening out here at the ground level. What's up, friends? I'm here in the breaks with Sama Alam Nailor from Sentry Senior Developer Advocate. So we've been working with Sentry for a while now, and I love Sentry. We use Sentry here at ChangeLog. It's so helpful for us. We don't write many bugs, though, so that's just how things work for us because we're amazing. But I get to see often how many folks use Sentry, and that number has grown over the years. It was 40,000, then it was 70,000, and now it's 90,000-plus teams.

Salma, can you believe that? What are your thoughts on the size of Sentry's impact to software development? Do you know what? I'm not surprised. It's a quality product. And I'm not just talking about that because I work for Sentry, but because I've used Sentry. And I think its success is also due to the fact that it supports over 100 SDKs and frameworks. Like any programming language you want to use, unless it's ridiculously obscure, Sentry's got an SDK for that.

Whether it's an official maintained SDK or whether it's a community SDK, there's a way that you can implement Sentry in your projects with a few lines of code. You don't need to really do much to get its benefit. And I think that's really powerful also in showing that people want to make Sentry work for their frameworks or their languages of choice, because it works. And the fact that you can self-host Sentry as well, it shows how valuable

it is and shows how valuable Sentry knows it is to people. The fact that it's open and out there and you can use it and configure it to your specifications at the code level if you want. And if you want to not bother about that and pay for it, then you can do that too. I'm not surprised and I'm not surprised that it's growing. I sound biased, obviously, but it's the best error monitoring solution I have used in my dev career of many years. And as a front end dev, it feels intuitive. I think a lot of these error monitoring solutions are very backend focused. They're very stack Tracy and not really geared up with a good developer experience. Like here are some logs, here are some things to spit

out. You can read them if you care. But with Sentry, it seems to appeal to more developers because of the way it's been engineered. The amount of SDKs that are available makes it appeal to more developers. And you can get started in Sentry in so many different frameworks in less than a minute. And all the instructions are in the app and they point you to documentation if you need it. I actually just recently created a set of videos called Sentry in 60, where I show you how to set up Sentry for the seven top SDKs in less than 60 seconds. And it's, you know, a joy to use. And so I'm not surprised that that many people use it. Well, we use it and we love it. So get Sentry, go fix it. Too easy. Check them out at Sentry.io. That's S-E-N-T-R-Y.io. And make sure you use our code CHANGELOG and you'll get $100 off the team plan, which is super awesome. Again, use the code CHANGELOG, get $100 off the team plan, Sentry.io. What do you think about Meta's newfound open-ish worldview that Zuckerberg talks about a lot?

You have Llama 3 now also open. They're calling it open source. I'm not necessarily saying comment on the open source definition and all that, but then you also have their embracement of the Fediverse with threads. You have the new Horizon OS, which runs their MetaQuest VR. They said that they're going to open that up. I don't think they're going to actually open source it, but they're allowing other people to build stuff with it.

So there's kind of this, seems like newfound, open-ish strategy coming out of Meta. Do you have thoughts on that? Or do you think that's good, bad, side-eye? What do you think about it? Well, I mean, I can't really comment specifically on Metas. that or do you or do you think that's good bad side eye what do you think about it well i mean i can't really comment specifically on meta's particular take just because i don't know anyone there but i think it's part of a larger trend saying oh the end of corporate open source

i think that's totally not there we're nowhere near the end of corporate open source, right? This is just, we're, but I mean, we need to compare open source as an asset category. I mean, meta throwing money at open source is because they're looking for relevance in the world that they're losing relevance in. And I think this is very similar to what Google's doing in a lot of ways, struggling to try to figure out, oh, how do we keep our revenues going? What's the next thing? But I don't think they know what that is.

I think that open source is a strategy now because it's a way to try to win the hearts and minds of people and to have them invest their time and their energy. But I also think that the rug pull is kind of the number one strategy in open source, corporate open source now, private equity open source. Which is not a good strategy, right? I mean, that's... No, I think it's a very bad strategy.

Damaging. I think if you say it up front, look, we're a business, we're in here to make money. This is the line. This is the free part. This is the paid part. Join us if you like that. If not, don't.

Okay. You know the rules getting in. You can decide, do I want to play or not? These are the table stakes. Changing the rules part way, you know, it's very much, you know, Darth Vader telling Lando Calrissian. I am altering the deal. Pray I don't alter it any further.

You know, this is like, you should have your plan to get out of Cloud City a while ago. All right? Like, when they first arrived, you should have been like, yeah, yeah, the meeting is when? Oh, yeah, no, I'll be there.

I'll be there. And you're like, okay, we're out. You know, don't even take the meeting. You know, run for your lives. You know, and I'm not judging anybody, right? Because the money changes them. You you know if i had a bunch of money and i'd be changed too you'd be like oh man that guy ron used to be so cool like wow you know i feel like the the grateful dead you know they said when they were asked you know

are you aren't you selling out they're like we've been trying to sell out for years it was just no one was buying you know so but i mean i think it's the changing the rules of the game midway or not even midway late in the day but also you know let's define some you know new categories right corporate open source actually now is either directly like a project that's a pet project of one of the hyperscalers, or it's one of the big Linux Foundation-type groups where it's basically not something you get to join as an individual. Like, hey, can I get in the club? Like, oh, yeah, sure.

You just got to pay our annual dues of this many bazillion credits, and you're in. Right. No problem. Right. Right. So, I mean, that's a way of making sure that it's sort of a cartel,

like they control the narrative in a way that doesn't generally threaten anything just because something generally new in an open source that destabilizes a bunch of existing companies is quite dangerous to them. And they don't want that to happen without some way to control it. Yeah. I mean, it's just a way to perform your fiduciary responsibility to your shareholders.

Sure. Buy that company and make them go away. How does hybrid group do it? Because you guys fund a bunch of stuff. I mean, tiny go a mechanoid, I mean, go bot down through,

I mean, hybrid group is your company, right? And you, these are legit open source. These aren't rug pulls, are they? Are you still waiting to pull the rug out? Oh, man.

Yeah, I know. No, I have a slightly different attitude. So there's, but I'll get to that in one second because I want to talk about the second kind, right? So the corporate open source, you know. Then we have what I would call private equity open source, right? Private equity open source is what happens when a company gets bought by a big company, and then their whole purpose is to basically extract all the value out of it with whatever's left.

There are some big companies that do this. You probably have seen them. They go after, you know, mid, some of them, their strategy is like go after mid cap companies, buy them. This has happened to a number of companies. Travis, you know, was a particularly sad example just because a lot of us, you know, depended on Travis CI. Yeah. Code ship was that too.

They got bought by private equity and then, you know, squeeze all the juice out i would say ibm's strategy is you know with what happened with red hat you know was very much an example of that same sort of let's just get all that we can out of whatever we've got in this company left and again i don't i don't judge these people, even in Red Hat's case, could they continue as a private company that builds on top of their open source? What is the next big move? What is the exit for that kind of company? It really is either

remain independent, IPO if that's possible for them, which it totally is, and they did or get purchased by the next bigger behemoth which obviously would be like a google and ip ibm or a microsoft or an apple and then even hashicorp where their purchase from ibm was similar it was like there's infrastructure linux infrastructure there's you know terraforming instruction infrastructure which was actually called terraform basically that whole purchase was for what happened with terraform the relicensing we talked we talked about that a bit with open tofu and we actually missed the ball we talked with adam jacob the morning of the afternoon's announcement so we talked at like what 10

o'clock jared am our time central And then like an hour or two later, after the conversation, we've laid it all down, we put it on tape, and the announcement of the, well, hints of the acquisition was talked about, and then the solidification of it was, I believe, the very next day. We missed that ball.

But yeah, I mean, that's what's happening. It's like you've got core infrastructure like Red Hat, which was essentially the definition of open source enterprise Linux, and everything to open source enterprise Linux built upon. Yeah. Extracted. It sort of indicates the general pattern for these companies

when it's closed source and it's mid-cap companies acquired by PE, it generally means the game is over. Now it's just like cut it up for parts and get as much as they can. And that just means it's kind of over. The dream has ended. And I think, is that the case with HashiCorp? I don't know the answer to that. But it is interesting if the cycle of time, the time it takes from founding of a major important piece of open source that is relied upon as, you know, key infrastructure, the time of creation to the time where it's basically like,

it's not leaving, nor is it going anywhere. Like it's not going to expand, nor is it going to contract easily. So now it's just like squeeze the value out. You know, it's just, we see this with a lot of other private equity investments into other categories. And looking at open source as an asset category is somewhat interesting. It's hard to evaluate exactly. There's a lot more value created by open source than captured. And that's a good thing. That makes it a public good. It doesn't cost me anything for you to use my open source.

It costs me something to create it so this goes kind of to the third model right so we got the corporate open source model we got the pe open source which is like okay you know rug pull extract the value damn the torpedoes you know obviously there's little tiny individuals but then there's this other category, which I like to think of as the volunteer fire department model of open source or the federation of open source. So what is this? If you are doing something with open source that benefits you, and I am doing something with that same open source that benefits me and our interests are aligned, then we'll work on it together. Like we're going on an adventure, we're in the same dungeon party because we're trying to do the same basic thing. And TinyGo is a great example of, you know, TinyGo is a Go compiler written in Go, licensed with the same BSD license that Go is on purpose

because the idea is that nobody really owns TinyGo and nobody really can. The idea is that you use TinyGo to make something of value. Use it to build your business. Take TinyGo, use it, make money, be successful, and contribute back. Put in time, put in money. We haven't had as much of that, but people are very welcome to hire Hybrid Group to help them, or we're probably

going to introduce some paid support offerings, not as an exclusive thing. I'd like to see a dozen consultancies that have consulting practices based specifically around tiny go implementation not just oh come to hybrid group for these things right because it's about letting a thousand flowers bloom and it's also about like if you look at the people who actually work on tiny go as their full-time activity for example there's some people at fastly damian grisky and dan kegel in particular who why are they doing that? Because Fastly's compute platform, which is WebAssembly system interface, you know, running WebAssembly in the cloud, you know, it's really a cornerstone of their technology platform that they charge customers for. So it's in their interest to put people to work on TinyGo because it benefits them.

We don't need a foundation. It's more of a federation of like-minded interests. And it's a model that is a bit more sustainable only because, let's say that tomorrow I decide I don't want to do TinyGo anymore, which I want to keep doing TinyGo, by the way. This is just hypothetical. Or I just decide, that's it, delete all the repos.

It'd be the same as OpenTofu. The internet is designed to route around blockages. No human institution lasts forever since they're made of people. No one's perfectly uncorruptible. Some people get tired, they have to retire,

you know, do something else, do a paid job, whatever, right? The purpose is to make it so that these things are not, you know, anti-fragile as possible so that they can withstand whatever changes occur and the core idea can keep moving on if it's of real value. So it's a different, but the only way to make this work is to be like the Wu-Tang Clan, where each one of the members of the clan can go do their own recording,

sign to a different record label, get whatever deal they can to get paid. They don't have to go through this centralized hierarchy. Oh, everyone's got to, generally speaking, when you and I had a band and we recorded and we would go to some record label. So we would sign our record deal. And then if we had a side band,

we would have to sign with the same record label with that other band. We couldn't just go and do our own thing. Oh, no, no. No, no. Like if we just did the side project of just recording some music on a film, they would want a piece of that as well. You know, the Wu-Tang Clan's big innovation to the music industry in part was anybody who's in the crew can go and sign with whatever record label they want. You can bring in whatever guest artist you want.

You can do whatever deal you want, do any deal you can. And this is sort of the same thinking that we have with the TinyGo crew. Anybody's free to go and take TinyGo and use it to make something of value and to make money. And that's how it's able to be sustained is because then those same companies and people put time back into it. So it's a little different than this shoot for the swing for the fences VC mentality. That's a lot. You know why we haven't taken investors. You know, we're happy to take grants where you give money and we don't promise you anything and we don't have to give it back.

Right. Yeah. promise you anything and we don't have to give it back. But we're not looking, TinyGo itself is not a thing that you can only invest in it by actually rolling up your sleeves and doing the hard work or paying the people to do it. It's like the opposite of death by a thousand paper cuts, it's success by a thousand innovations. I guess the question, though, in that scenario is, is there bdfl how is

stewardship operated how are contributions accepted when you have like corporate partners or lack thereof just participants let's say from fastly as the example you mentioned you know they have an interest to advance tiny go i would say probably advance it in their best interests as well. How are contributions and innovations and features added, or how are they approved of? What's the process to govern contribution or to veto a contribution? What's the mechanisms there? It's highly informal. Part of that is because I've seen a lot of obsession with governance that has taken a lot of time and energy and not really yielded substantial benefits. They still had massive drama, and they forked the project and went off on their own, and

everyone hated each other and was all mad. And they had all kinds of governance documents. We can't get around, nor do we wish to get around the fact that open source collaboration is a first a human activity. And the only second does it have to do with code and technology. And it really has to do with the attitude about how you wish to approach it. I think of it, you know, because the big dictator implies a certain centralized command and control, which has worked really well for certain projects.

Sometimes despite that, not because of, I think. Yeah. Right? I look at it more that, like, I'm the park ranger, the game warden, you know know the curator of the museum not the owner of the park not the owner of the museum right because it belongs to everyone you know i'm just the steward of it to try to like me personally that's my attitude. And there's always the opportunities if somebody genuinely is like, we want to take it in this way. There's not a

consensus for doing that. We're taking it in our own way. They should do that. That's not a bad thing. We don't have to all if we reach a fork in the road, it's not two brothers. This time it's actually a business decision about something or another. And I mean, TinyGo has grown into a very big thing in WebAssembly. It wasn't originally created to do that. That was an emergent property of the ecosystem, one I'm very glad for and that I'm very involved with now myself.

But that was not, the original vision with TinyGo was very specifically for small embedded devices and bringing Go, you know, to the microcontroller. And then it just so happened, oh, you know, we could actually compile this to WebAssembly. There was some, you know, very nascent support in BigGo for WebAssembly at the time. And that's something that the LLVM tool chain that TinyGo is based on, not BigGo, but that TinyGo is based on. So it's like, oh, this would be a cool feature to add. We should do that. No real specific, more just like, oh, it's like a low-hanging fruit.

Some people seem really interested in this. We should do it. Next thing you know, oh, wow, TinyGo is so much better for WebAssembly than BigGo. We're like, wow, it is? How interesting. Why is it better? What does it do? We didn't even know ourselves some of it.

We were just genuinely interested. It was an emergent property of the community, of a problem that they wanted to solve that BigGo was not solving for them. So a bunch of people started devoting time and energy to improving that WebAssembly support in TinyGo. And then the WebAssembly system interface, which is WebAssembly on servers and serverless. The WebAssembly system interface, which is WebAssembly on servers, you know, and serverless, the WebAssembly not in the browser, you know, part of the fact that we were interested in it

ourselves, that also we move a lot faster than BigGo. You know, we can, you know, we can innovate a lot more quickly. We haven't given a 1.0 guarantee yet, so we can change things. You know, we're not, even though largely I would say the road to 1.0 for TinyGo is more about the hardware interfaces than it is about compatibility with Go itself, because we've had pretty good compatibility for a while.

But it was very much an emergent property of the community, and as more and more of us said, oh, wow, WebAssembly, that's really a very interesting thing. We have some problems we want to solve with that. And Go being a really good language for whatever, that this has been an aspect of TinyGo, which was not something that was part of any master plan. It's something that the community wanted, that the community is and does, and is not at all, you know, there's no need for it to be a zero-sum game. Oh, it can either be for embedded, or it can be for WebAssembly, or it could be for Linux

or for Windows or for macOS. You know, that's a false dichotomy. It could be for whatever we collectively want it to be for. But again, with that sort of curation in mind, you can share the park, but you can't just go in and start digging it up to build your own little thing when that's interrupting the, disturbing the flora and fauna of the ecosystem.

And so that's where in a compassionate, you know, kind way, but also looking at like the why. Why is this person asking for this thing? Usually it's because they either have some need or they can't figure out how to get the thing you have right now to do the thing that they want. Either way, that's kind of on you. If it's not a need that they can, if it's a need they could solve some other way, you could recommend that.

If it's a thing that your software doesn't do yet, but that it could, you could mention that and say, oh, maybe you could help. Sometimes they say, oh, I don't know enough programming. Yeah, but they already took the project in a better direction just by saying, hmm, here's this thing that somebody might need. Like WebAssembly, again, a really great example. There's a lot of people who are very involved in blockchain applications. I'm not really one of them, right? I don't judge people for what they choose to do

with their computational power. You know, mine currencies, render graphics, play games, churn through large language models. This is on you, not on me. I'm just trying to create technologies that are useful. So there were a bunch of people who are using TinyGo specifically because they wanted to use WebAssembly as part of their engine for doing their processing. And so they were looking for basically what we call WebAssembly unknown, Wasm unknown, which is sort of a naked WebAssembly. It's WebAssembly with no expectations of what the running environment would be. So for the listener,

if you haven't checked out WebAssembly, first of all, do. There's lots of interesting things happening. You can run it in your browser. You can run it on serverless applications like Fermion Spin. You could use it to build plugins for your current software like Xtism. Or you could use it to actually run on microcontrollers and embedded devices with TinyGo and Mechanoid. So there's all sorts of different interesting areas to do this in. So this group of blockchain enthusiasts said,

we really want to run TinyGo with Wasm with no external dependencies. And the people who were doing Wasi are like, oh, that doesn't sound very useful, just because it doesn't have any ability to call any specific thing. And these folks were like, oh, that's OK. We don't care. We have our own APIs. So it took a while before it actually turned into something which landed in TinyGo, just because it had to build a little bit of a critical mass. Like just one person wants it, just one or two,

they're not willing to do any programming work on it, but they'd like to have this feature. Okay, that's a signal of intent. There's maybe other people. Is it useful for other things as well? Well, yes, as it turns out, that same pattern of Wasm unknown is exactly the pattern that's being used for Mechanoid

for running WebAssembly on embedded devices like microcontrollers. So if it was not for the blockchain community sort of priming the pump of starting to think about this, then those of us

who are actually interested in using that same pattern for something completely different would not have maybe worked on it and adding it to TinyGo. So again, the community is defining what the thing is, because the purpose of TinyGo is to serve the community's needs. It's a means to their ends, whatever those happen to be. I really like this Wu-Tang Clan analogy. And I'm over here just reminiscing on Wu-Tang. You can still record a disc that only one person has that's like $6 million or something.

Yeah. You could use TinyGo to do it. That would be cool. Have you written down this Wu-Tang Clan form of open source and promoted it as an idea? I'm sure I read it somewhere. Okay, so somebody else is doing this. Are there certain projects, obviously TinyGo being one of them,

that lend themselves well to this form? Or do you think this is something that almost every open source project could adopt as a way that they go about doing things? It's like, well, we're going to do the Wu-Tang Clan thing. We're going to be a loosely affiliated group of people who all have similar ideals and want to collaborate and do awesome stuff and maybe make some money like wu-tang well first of all you know you gotta have some flow i mean most likely you don't okay i mean you know like not every one of my songs was actually any good luckily i didn't play for too many people.

They were like, oh, wow. I mean, that's a really bad idea. I'm like, oh, yeah, OK, yeah. You know, cool. You know, so. Sure. But it comes to like the thing that it is. Like, you're not going to make money off your programming

language, man. Like, I'm sorry, OK? You're probably not. It's just not going to happen. You're killing my dream right now ron you're killing my dream i know it's like all these people are like i was i'm gonna make my own language and like i'm gonna be retired like a mansion and a yacht it's like no you're probably gonna make not only you're gonna not make any money it's gonna cost you a great deal of money just to even do this thing right now lots of money is made off of programming languages okay but it's not like the language not selling the language yeah like

nobody's getting like oh yeah we sold python no it's like yeah we made this thing with python and we sold it for a bunch of money and wow python's really cool i mean pick a language, go, Rust, any language. If the money is not there, the value creation is. It's creating enormous value. I mean, I will take a giant leap and say the singularly most valuable asset on earth, which is consistently undervalued is the collective all of open source. Like, how much is that worth? It is worth a lot. Like, I know there was a study, I don't remember what, some university they did trying to figure out, okay, if there was no open source, how much would it cost to write it all? It was like a bunch of trillions, you know, not to mention like, okay, we better get started

because, you know, like we got, okay, we got $3 started because you know like we got okay we got three trillion dollar budget but like how are we going to spend it like it took years you know it's not like we're going to like okay we know exactly what to do now right so it's an emergent phenomenon so that does not mean that there's not ways to do it but i think it's not like a project which is clearly part of core infrastructure for a bunch of companies is that, that they're not planning on selling the cloud enabled version of themselves in order to make money. Right. Cause that's a lot of, Oh yeah, it's this cool server that, you know, does some thing and like you can work on it and you could stand it up

yourself or you could use our cloud service. And then the next week somebody's oh yeah we could stand up a cloud service and then the original people like wait no not like that yeah exactly no i have i think about it more like hey if they could do the same thing we're doing but cheaper we should just buy it from them and resell it ourselves if we're trying to make money you have to think about do you care about it being from you? Do you care about it existing? Or are you trying to use it to make a living? There may be some overlap in those things, but this is not a strategy that's going to work for every open source project, especially if it's a thing where the cloud-enabled version

is the thing you're going to sell, unless you're perfectly cool with the fact that your own direct competitors are your collaborators in the project. Now, if you've got some secret sauce that you can add that you think is going to make it better, but that can't be copied, like, oh, our UI is so much better.

Well, yeah, they can look at your UI like, oh, we should do that. Like, okay, that's not enough. You have to have some differentiator. Either that or it needs to be very verticalized, like it oh, we should do that. Like, okay, that's not enough. Like, you have to have some differentiator. Either that or it needs to be very verticalized, like it's for a particular industry. You know, go for industrial computing. Go for automotive.

Go for entertainment. You know, go for whatever, like verticalized as opposed to horizontalized. I'm going to let him keep going. Like, how many can you come up with here? Go for what? I can keep going. I've got a pitch deck, too.

I'm trying to think of what go for entertainment looks like. Uh-oh, I may have said too much. Well, we did not sign a nondisclosure agreement, Ron. Yeah, no, I don't sign. I mean, I'll sign them if pressed, but I've never asked anyone else to sign them. I respect the customs of others, So I'll be like, yeah, I'll sign your NDA, but I wouldn't ask you to sign mine. I'm like,

are you kidding? Like if all you did, if you, if hearing my idea is enough to just go do it, you know, please hire me. You know? Right. I think it'd be funny to have an NDA and then come on a podcast to talk about stuff, you know? Well, yeah, that's also quite hilarious. You know, I'm like, wait, you know you know come on please don't share this with anybody it's literally the point exactly like they're missing but again they're it's more like opportunistic using like oh open source yeah we're

going to totally like wrap ourselves in the open source flag it's like there is none like yeah we just want to talk to like the open source community. There's no single open source community. There's many. In fact, one can splinter into two or three or 10 at any point. I mean, you actually can't. It's an algorithm. It's not a formula. You don't just substitute the numbers and now comes open source at the other end.

You turn the crank and you don't really know what's going to happen. Anything could happen at any point it could be magic or it could just be like you know we spent a bunch of time to we redid the wheel but our wheel is square and made of wood and all of a sudden one day we saw round rubber wheels and we're like hey you know we just our project is now archived because exactly no longer maintained like that's it we're done you know thank you thank you good night you know sometimes that's a relief you know sometimes that's a relief you're like oh a better way cool we can just stop this maintenance process you know well again it's like do you want the ego gratification of like all my names on it or do you want to solve the actual problem and you know do

something else you know actually dr nick william, he was a person who really instructed me. He gave a talk about this, about retiring open. This was many years ago, back when we were all involved in Ruby. I think he's still in Ruby. I still love Ruby, but I haven't actually used it in years. But hugs and kisses to everyone in Ruby, because I think it's a really cool language and a cool world. He did a talk, and he was mentioning about letting go of projects. And his metaphor was, my kids, he has a bunch of kids,

I don't just have them living at home forever. I mean, eventually they have to leave. And either they go off and they do their thing or whatever. I was like, whatever? But maybe the kid's metaphor was, that was too extreme, but there was more that, you know, you don't have to maintain the project forever. Either other people will rise to the occasion or the thing will be replaced by something else, but you, it's not on you. Like there's nobody, you know,

and if there's something genuinely better that you could actually use yourself to solve the problem you started out to use, well, you learned something. And it was cool. And now, yeah, move on to the next thing. Jump on their project and help them. Yeah. Some of the contributions I'm most proud of are a single pull request I made on someone else's project where I was able to solve a problem. And I'm really proud of a project that I maintained, GoCV, which is the Go wrappers around OpenCV, for the same reason that there's a lot of people who've made one pull request. It was missing one

thing, and they were able to figure out how to find that thing was missing, what they had to do to add it, and it was successfully added. And their problem was solved, and they moved on because they were trying to do a thing, not be an open source contributor. They were just trying to, like, to me, that's a massive success as a collective. Like, we made it possible for you to solve your problem

and to help the next person along the way. And yeah, it took a lot of work to do that. It's much easier to just do things directly. Yeah, right. But you have to be in the long game if you want to create this sort of sustainable, maybe I'll call it sustainable open source. It's what VCs would call a lifestyle open source.

And they're using it as a, for them it's like an insult. Like, oh, that's just like. Yeah, that's an insult. You know, and like for me, that's like, okay, cool. That means, you know, maybe I want to be in. Cause you know, like, I don't want to shoot for the, shoot for the stars and crash into a mountain.

Like, I don't want to actually crash, you know? Right. Like, plus we can actually get up that mountain so many other ways that don't involve like rocket fuel and, you know, half destroying the village that we other half of the rocket lands on. Like we could just hike up there and with our camera and like get the shot and we're done. Well, VCs are fundamentally about the moonshots. They're trying to ride the ride the rocket ship and then have it exit before it blows up. I mean, they're not about a sustainable lifestyle. So of course, for them, lifestyle, that's an insult, like you say,

but for open source maintainers, it's all that we have because the rocket ship, I mean, you got a rug pull to get your rocket ship, right? I mean, most, I mean, some, some people started open source and, you know, made money in addition to it, but not, I don't think because of it, you know, but for most of us, we're just trying to get money off the thing you make with it, not off of the thing. We're trying to, we're trying to get stuff done. We're trying to run a business. We're trying to help people write software and we are collaborating on the parts that we can. And that's a lifestyle. Like you want to keep doing that, Ron, you've been doing a long time. I guess Adam and I have as well at this point, you know, we don't want to crash into the side of the mountain on purpose and we don't want you know, we don't want to crash into the side of the mountain on purpose.

And we don't want anybody to. I don't want the person who maintains the dependency that I have on that. We run our business off of. We don't want that person to leave the Internet and give up and be gone. We'd love for that person to thrive and sustain and be able to maintain that project of theirs. And we'll help them if we can, you know, or even hang it up like that's OK, too. Like, thank you for your service. Like, you've done a done a lot you got us this far we'll take it from here you know i mean that's yeah i really respect and admire the people who know when to do that

and it's you know usually we don't as a community like we don't have the sensitivity a lot of people don't have the we don't talk about mental well-being i don't want to say mental health because mental health sounds like illness like i'm going to use the term well-being only because it makes it sound more like that sustainable like mental illness like okay you're better now back to the salt mines you know versus like well-being like i'm sorry um to inform you i will i'll be unable to come to work this week because I'm feeling too well. I'm feeling too good this week to come to work. That sounds like a good one, yeah.

Not like what? If we were athletes, we would have personal trainers and team positions, and they'd be like, oh. Dietary. Right. So we're doing Olympic-level cognition. It's like, here's a coffee and a donut. what feeling burnout no you'll be fine we just got to make it through this next sprint you know it's just and there's a lot and it's you don't want to talk

about it six pack a mountain dew stay in the basement right and you don't want to talk about like oh i'm feeling stressed because that's like you as a professional athlete saying, yeah, I'm not sure about my ankle. Like, oh yeah, we're not sure about your contract then. Like, you know, good luck selling cars. You know, I was a hero last week and this week I literally have no job. Like this is, I think in people's minds and I think it's in their minds for a good reason because saying, oh, don't worry, the industry will take care of you. The industry will fire you in a heartbeat just because a bunch of investors said we have to let go of 7%.

Why 7%? Because everyone's doing 7%. Like, because 7%. You know, Bondo, it's what plants crave. And you know, and you're like, no, it can't be a documentary from the future, but I'm afraid it is. Right.

So, yeah, don't tell your boss about you're not feeling good because it's too risky. This is what I think is in a lot of people's minds. And I can't tell them that's not true. Don't worry. Because I think circumstances have shown that maybe isn't necessarily true true so what do we need to do we need to find mutual support not like you you can't tell you have to have peers that you can communicate these things to and we have to build a vocabulary around it in order to first of all not tell the suits what's going on because they're going to be like all

freaked out. Like you're trying to tell me my star player can't play? Like, no, no, don't worry. Don't worry. Like they're worried. Okay. Cause they can't, you know, they don't even know, like, is this like a soccer ball, you know, football? Like, well, actually that's the same thing, sir. You know, in this country, like, wait, you know. So I mean, this is a fundamental dichotomy between the creators, many of us who like, we like to create, we're going to create, we want to be of use. Like, that's our purpose for being. And like, yeah, we need to thrive and survive. But we're not doing it for the sole purpose of exploitation. We're doing it because we get satisfaction out of improving the human conditions and we're human. So we can be improved along the way, right?

We're not just a resource to be exploited, you know, just like the brain machine it's on the blink. I'll get some more brains. You know, how many times have you heard literally something that is the equivalent of that in management meetings? I've heard it a lot. And, you know, over the course of my career, and I don't generally try to single it out because that's not very effective, but it is a thing that, I mean, anyone who's feeling extreme burnout, you know, definitely do seek help outside of your workplace. There are people that you can talk to about mental well-being. And we as a technology profession need to be able to talk about this in a way that

doesn't threaten our employment. So that's my hot take. Yeah, well said know i've been through times in the past where i had all those feelings so i and i don't tell people usually i'm telling everyone in the world on this podcast because i think it's important that we talk about it and it's been a long time since i had that feeling because i'd learned to recognize the signs because i did seek some professional assistance you know to figure out what was it that I, what was wrong with me and what could I do about it to be healthier and happier and better for it. And so, you know, I'm a resource, reach out to me privately. If you know, if you need to fellow human, I care. So that's what keeps, that's what keeps me doing it. Just because I'm an idealist and I think, all right,

in the end, this is going to be better than all the other alternatives, which is closedness, which is how much I can get just for me. And maybe sugarcoat it with some philanthropy, but when the act itself is a form of philanthropy philanthropy then it's a sustainable way to live and that you know ultimately what is the technology for you know i think we were talking about this the last time we were in uh in portland at oscon actually you know yeah 2019 yeah it you know, the purpose of technology is to improve the human condition. So, you know, that that's what keeps you going.

Before I let you go, I would be remiss not to bring up one of my favorite go times of all time. I want the director's commentary. You actually called into go time from the future. I'm sure you remember this. Well, it hasn't happened yet. That episode was called 2053.

I saw that episode and I was like, this is crazy. Who is this guy? Like, this guy looks just like me, but like, you know, but. You were from the future. You were calling Matt Reier and Natalie Pistovonich to tell them all about the future of Go. This is the reason why I'm making this call. I'm using all of the battery energy that I've saved for several years in order to make this transmission

to send you a warning from the future. You see, I am the last Go programmer alive in 2053. What? No, don't say that. And it's terrible. All I do is maintenance programming. I haven't added a new feature in over 20 years. Can you give me the director's commentary? This whole thing was your idea.

I think you pitched it to Matt or somebody, and you actually had this cool setup where you looked like you were calling in from the future and everything else, and I just produced the episode. I had no part of the planning. I would love to hear, like hear like where'd you come up with that and why did you do it all right directors directors commentary yes so matt sends me a message hey do you want to come on go time like you know today or it was tomorrow maybe i don't know it's like you're typical like oh man we're out of guests like great planning who can i think of who's like your typical, like, oh, man, we're out of guests. Yeah, great planning. Who can I think of who's always available?

So the idea originally was, what if we were like the product managers of Go in the future? It's kind of a meh idea, not that great. I mean, it was an idea. OK. Well, so I'm like, all hmm all right i'm gonna go over to the studio i have a collaborative workplace called la pipa which is where the local creative technology community get together it's not a co-working it's more like a collaborative space really amazing spot.

A lot of cool things happening. So we have a studio there that's set up for music recording, for streaming. All my dead program episodes that I did live from La Pipa were that. So I'm driving over, and all of a sudden, it hits me. I'm like, oh, I know what I need to do. It was just, you know, I don't spend a lot of time alone. Okay. When you leave me alone for a few minutes, it's really dangerous.

I start thinking like by the time I get to my destination, who knows what's coming on. So I call him over, my collaborator's over at the studio. I'm like, look, I'm doing this thing. It's a pod, you know, it's a streaming from the future. I need some lasers. And do you have any tinfoil yes bird now you have to understand i know we've collaborated a lot together and alex lawton my co-producer of the dead program uh streaming you know he's a brilliant guy english slash spanish he totally gets it all right right. He's like, all right, no problem.

I'm like, I'm coming in hot. Okay. Cause you know, we got to do the show. Right. So I get there and I, you know, they got the lasers set up, you know, on the smoke machine, you know, just cause lasers and smoke machine featured like you know duh right yeah so i i tell matt look okay new plan i'm gonna call in from the future all right and you're gonna ask me some questions and uh yeah that's like that's it that it. That's the whole idea. Now, Matt is the straight man in the comedy act is always interesting. Right. But, but he gets it. Right.

So it just, it started going and I had some ideas that I thought about on the way over just because it was not, it was very comedic, but I also had some very serious things to say. And if you kind of parse it and listen to it with that eye, it's a cautionary tale and it's a genuine, like heartfelt outreach to a lot of different communities about some things that I think were going on with Goat and the world at the time. But like my idols, you know, Robin Williams or Eddie Murphy, you know, people who were, I mean, I'm no Robin Williams I'd like to be,

you know, because more lucid human being, more present, more quick-witted, you'd be hard to find another human being with that type of rapidity of thought. So these were my, somewhere between that and the old Jewish comedians, you know, take my computer, please, you know. So it just, you know, and a definite, you know, hat tip to Back to the Future and some of these other, you know, tropes that formed us. Okay. These are the things that made us who we are because we were them in part

and they were us. Right. And Matt was amazing. Natalie, they were amazing. Like, it was just, it was a great moment of inspiration and I really appreciated that everybody could you know could go with so little information about what i was going to do and that it came out so fantastically i really am very grateful to you know all of the sources i

stole from the past and the future that i stole from know, and all my collaborators to put it together. And it's, you know, it keeps getting views. That thing is, look, until 2053, that thing is still, like, we're like, let's see, you know, we still got two more years before, you know, Nostradamus is wrong, you know. I wonder when I watch that, if it was like that short film that you see gets produced into a full-on movie or something that's like a precursor to something that is at least annual or repetitive in some way that has not just one occurrence but some sort of episode

not so much like weekly or it needs to be a thing but something where it's a thing you do more than just one single time well i will when is this show coming out this show's coming out sometime in the future this friday friday if i were you out humans of the internets okay i would go to chicago in july to the gopher con us and i would make sure well first of all go there for the community day before the whole thing in July to the GopherCon US. And I would make sure, well, first of all, go there for the community day before the whole thing starts for the hardware hack session. That is always amazing. We'll have drones, hardware, all sorts of cool stuff, right?

That's during the community day. That's before. It's totally free, by the way, to go to the hardware hack session. You don't have to pay extra. You just have to have a ticket. We're not even checking your tickets, actually. Like, you know, maybe somebody is. Don't tell them that.

Well, do get a conference ticket, right? But definitely be there day one of the conference, because there may be a chance that something from the future may appear in the present. Well, that's still in the future. It's the future, but it will be the present. It'll be a present. It'll be the present when the future comes then. Yes. Yes.

Is this a conference conference thing, like a talk? Or is this like a maybe thing? It's some before thing. What are you trying to say that first day? Well, the day before the conference, come to my hack session. I see. Okay. the day before the conference, come to my hack session. I see. The day of the conference. The day of the conference, make sure you're there for the opening talk.

That's all he's going to say, though. The official announcement hasn't come out yet, so I may well have just scooped. That's all right. We like scoops. But yes. A little treat. It's going to be outrageous. I love outrageous.

That's a very awesome word and a very awesome thing going to be outrageous. I love outrageous. That's a very awesome word and a very awesome thing to be outrageous. Like many of the things I do, it will be a spectacle. Whether it's the spectacle I intend or some other spectacle is not important. I promise a spectacle and spectacle I will deliver. There's a girl on the internet her name is Ali Spagnola I don't know if either of you have heard of her but she has a really she's she's like uh maybe uh the same way you run explore and scientifically explore the permutations of all code and Geiger instruments and whatnot to like check out radioactivity in the atmosphere, this thing,

she's like that, but with like human behavior and like just humanity, she's in the fitness and she eats the same thing literally every day. She's like a human science experiment. If you haven't seen her, no, I wanted to check that out though. You'd be an absolute treat to go check her stuff out.

She's just a, just, she makes music. She does all sorts of healthy things fitness related things just like a an literally outrageous type person so she says that word a lot it's part of her kind of core brand and so i just thought i'd mention that because she's pretty pretty wild if you haven't seen that before i haven't i don't usually use that word actually i'm sorry you should i mean i spent too much time in California. I usually say awesome just because, like, everything's awesome.

Yeah. Yeah. But this is one that I'm, it's the 10-year of GopherCon. So something special has to happen. And, well, I will do my best. Well, we appreciate you stopping by and doing your best at making an awesome episode with us this was lots of fun always is Ron

in fact we haven't seen you since OzCon man it's been that long I miss you guys the real world the internets are all well and good but I haven't left for another planet yet I'm still here not as often as I was but I haven't left for another planet yet. I'm still here.

Not as often as I was, but I'll definitely be at GopherCon. I'll probably be at some other stuff this year. I'm going to try to make more of an effort. I do a lot of European conferences. I live in Spain, so it's very quick and easy inside the European Union.

We've got to get to a FOSDEM, Jared. Jared and I haven't been to a FOSDEM before, european union we got to get to a fosdam jared you know we haven't been jared and i haven't been to a fosdam before so i think we need to make that a priority i think that could be like an easy connection with ron because you've been there three years straight uh i'm sure probably even more than that given oh more than that actually this was my seventh year in a row giving a talk at fosdam okay which is kind of amazing just because, you know. Yeah. Is FOSDEM in the January, February time frame? When is FOSDEM usually?

It is February. It takes place at the campus of the University of Bruxelles in Brussels. It's kind of an anarchist conference, by which I mean there's no registration. You just show up. So the organizers, the way they organize it is each room of the campus, you pitch, they take pitches for subject areas. Like there's those sub-organizers is 100% responsible for choosing the speakers, organizing them, making sure they do their thing. It's 100% on them. The campus, they have streaming AV, but they just provide that running throughout the day from volunteers. There's no corporate sponsors. There's, it's, it's like a social anarchist conference. And it's really quite amazing. You know, very much a lot of people

who are maintainers just show up there and hang out. The hallway track is incredible. I've never even made it in the rust room, but I've had the most amazing conversations with people waiting to try to get in. So I highly recommend Fostam. Also, the night before it begins, it takes place over the weekend. It's Saturday and Sunday. The Friday night, traditionally, they have a opening party at the Delirium Brewery, which is a classic Belgian Trappist brewery famous for the Delirium Tremens Ale. So there's people I would only see at that beer event, like Charles Nutter.

Charles Nutter. I've known Charles for many years, yeah. And maintainer of J.Ruby. Yeah. An amazing guy, super smart. Didn't expect that J.Ruby would be so successful that would take over his entire life and career. And since I'm not too involved in Ruby, the only time we would bump into each other is at the beer event the night before FOSDEM and a great chance to catch up on

things just because he's a great guy. And, you know, we just, our paths just don't cross because of, you know, life. So yeah, please come to FOSDEM. I'll try to make that happen. We'll see if we can do it next year. Yeah, all that to say is like when we see you, I feel like life is just a little bit, actually a lot more better after seeing you. Like you're a joy to talk to digitally like we are right now in a podcast form and via video, you know, low earth low earth orbit satellites that you're

streaming from via starlink so thank you elon and those folks making that kind of stuff possible mr musk yeah thank thank you mr musk for my bandwidth please don't cut it off again after those last things i said yeah i mean i meant them but like please don't listen you know don't bite the hand that feeds you absolutely don't bite that yeah i mean. I mean, it's good to see even in this form. But I know that when we met you, Jerry, what was that conference? Well, I think it might have been, it was GopherCon, OSCON. Was it GopherCon? It was, right.

Okay. At the pre-party, right? When at the pre-party? The day before the conference actually kicked off? Maybe. I don't know. You'll notice I love those pre-parties before the conference kicks off. It's almost guaranteed if there's a conference and there's a pre-party you'll find me there 2014 is that right jerry was it 2014 that

sounds about right yeah it sounds so long ago i remember being like who is this guy because you were talking our ear off on gobot i think at the time or what was the ruby one uh r2 r2 yeah and you had like a sphero partner i believe that year you had your hack day the last day the unofficial last day of gopher con that was the first gopher con yeah that that was actually pretty amazing you know it was just i had a bunch of hardware not just because i had just given this talk but because i was actually going to berlin for another conference right from there and to do a workshop. So I had all these flight cases full of robots from Sphero and drones, all this different stuff. And we have a day in between of travel

and there was this room for the community activities which was like nothing planned. There was nothing planned. It was just a bunch of people hanging out. Yeah, it was the best. So it was just like open my flight cases, let people play with my toys. And the next year it was like, yeah, so you're coming for the community day for the hardware hacking? And it's like, oh, is that a thing now?

Because yes. The answer was yes. I mean, you don't even have to. You have to be a hardware hack. Yeah, absolutely. Well, all I have to say is that this is ChangeLogging Friends. And so you are, Jared and I would consider you very much a friend. Absolutely.

And I think you're an awesome human being. I love all the work you do in the community. I love just the heart you put into things. And I think that your outlook is infectious in a positive way. And I just really appreciate you coming on and just sharing like this crazy journey you take in software and just like just dragging us with you happily, of course. But yeah, very cool.

Thanks for having me. I really appreciate all of the great guests that you have, all the interesting things. You know, I learn a lot from listening and it's really fun to, you know, it's fun to chat with you and fun to share. And, you know, I really appreciate the work that you're doing and thanks for having me on.

Yeah, you bet. All right. Bye, friends. Bye, friends. There you have it. High altitude weather balloons. Life on Mars. Zeno's paradox applied to machine learning.

Collaborating on software. The Wu-Tang Way, Lifestyle Open Source, Improving the Human Condition, and zero, yes, zero mentions of Adam's favorite TV show. What else could you ask for from this humble talk show of ours? If you dig it, tell a friend. They might dig it too. Thanks once again to our partners at Fly.io, to our Beat Freakin' residents, the one and only Breakmaster Cylinder, and to our friends at Sentry. Use code CHANGELOG when you sign up for a Sentry team plan and save yourself $100.

Too easy, right? Next week on the Changelog, news on Monday, Paul Orlando's new book, Why Now? How Good Timing Makes Great Products on Wednesday, and Annie Sexton back here on ChangeLog and Friends on Friday. Have a great weekend. Leave us a five-star review if you want some free stickers, and let's talk again real soon.