Han shot first (Friends)

43:02The MCP integration is the clever bit. Your agents can actually talk directly to the database. They can query data, introspect schemas, execute SQL. without you writing fragile glue code, the database essentially becomes a tool your agent can wield safely. Then there's hybrid search. TigerData merges vector similarity search with good old keyword search into a SQL query. No separate vector database, no elastic search cluster, semantic and keyword search in one transaction. One engine. Okay, my favorite feature, the forks. Agents can spawn sub-second zero copy database clones for isolated testing.

43:42This is not a database they can destroy. It's a fork. It's a copy off of your main production database if you so choose. We're talking a one terabyte database, fort in under one second. Your agent can run destructive experiments in a sandbox without touching production and you only pay for the data that actually changes. That's how copy on write works. All your agent data, vectors, relational tables, time series metrics, conversational history lives in one queryable engine. It's the elegant simplification that makes you wonder why we've been doing it the hard way for so long. So if you're building with AI agents and you're tired of managing a zoo of data systems, check out our friends at TigerData at TigerData.com.

44:28They've got a free trial and a CLI with an MCP server you can download to start experimenting right now. Again, tigerdata.com.

45:05gray-haired, we'll make it more gender-neutral, sorry. People running stuff. You want the new people to come in with some old people around who have inherent knowledge of how things ended up and why they're there, and to kind of help keep things going without it being a dramatic shift. But you got to let people come in and change the project as they want. You just don't, yeah, you don't want stacked in a way that doesn't let fresh people come in and contribute. So no, I would not want an AI version of me to be on forever. No, I know you wouldn't. I was just trying to have some sort of way of making a segue.

45:38So you're in rare company, you know, somebody who's like had the means of power, like you've had it in your hands, you've had the power in your hands, like George Washington, you know, he let it go. And, you know, you're in, you're in rarefied air here, Brett, you had power. And you just said, Nah, I don't want this power anymore. I'm gonna, I'm gonna let it go. It's been a couple years now. Yeah, it's been two. Two years since you've retired from the steering council. Any regrets? You ever went back in? How's it working out? Regrets. Regrets. No, I mean, so I was on the steering council from its inauguration through the first five years of its existence.

46:17That was great. It was great seeing friends of mine every week and having an excuse to hang out and to help out and all that stuff. but when we were choosing our governance model one of the things I suggested is that we had specifically so we had guaranteed turnover. I didn't win that argument and there was enough going on when we were trying to make those decisions, I didn't press it. It was more of an idea. But I decided to stick to it plus potential get on the way and it's like, yeah, this is just a good time to just step back and let someone else come on and have that guaranteed slot to be open.

46:50So no, I don't have any regrets. would I ever do it again? Possibly. Given enough time. Once again, I think the kid would have to be a little older and I'd be able to focus a little bit more on it. Because it's one of those things that's kind of always I don't know about YouTube, but my brain never shuts off. So stuff's always going back in my head. And so right now it's a lot of kid stuff and I want to make sure that it's not pushing kid stuff out of my head. So I want the kid to be a little older before I do that. But I do miss hanging out with the people in the training council because I luckily know most core devs personally at some extent.

47:26So it was just a great excuse to hang out and just talk every week. Yeah, no, I don't, no regrets. Yeah, it was just a thing. I just stopped doing it. It was no weird anything. It's not like I felt like, oh my goodness, I don't have the power anymore. It's just, I mean, even when I was on it, I still had to go through the same process as everybody else. So it's not like I was in any way privileged for the stuff I wanted. The closest I had a privilege is I could ask my other steering council members, I have this idea for something. What do you all think? And they could shoot it down maybe a little earlier than somebody else.

47:56But everyone else can also still email the steering council and ask the same question. So it's still not that privilege. I might just get response time that's faster. So, yeah, no, it was fine. It was over too stressful to begin with. It just took time. So I just have that time now to focus on other stuff. It's interesting to be in that position now. So what did you, if it didn't have a lot of power, what kind of lack of power did you, or I guess, what kind of power did you share with everyone else in the community? Well, I mean. And what did you have uniquely that was, you know, part of the steering council?

48:30I'm just saying from my perspective and not being on it anymore, my power wasn't unique. And the steering council does have power as a whole, right? As a council, right? Because the steering council ultimately decides what Python enhancement proposals get accepted for Python, and that drives which way language goes. So that's where its power comes from, is its ability to make those kinds of steering decisions. But the council had enough workload when I was on it, and I think it's continued since, that there's not a lot of, we think we should steer things this way. When we first started it, we thought, we'll have these grand plans, and we'll try to spec out where we want things to go long term and where we think people should focus.

49:14No, don't have the time. Enough stuff's coming in from the community and other core devs asking for stuff that there's not enough space to come up with how to steer things. We never even had to worry about that. There's no power of, I want to make Python go this way. As a senior council member, you have a power in terms of having an opinion that people know matters to an extent, but it's not a, I'm going to drive this huge project. Um, and only steering council can make it happen. It's just the steering council may list can say yes or no. And that's really it. It's just a body that says yes or no for stuff that they get asked.

49:50What is the driving direction of the steering council? Like when you, is everyone voted in? Do you keep the seat in perpetuity until you want to let go of it? You know, what are some of those mechanics around, you know, being a part of it? What does it do? Yeah. The way I've always raised it, the steering council's job is to effectively be the final decider of things related to the Python project. If there's any disagreements about how to handle something, deciding peps, anything that we've just decided needs a final arbiter, the steering council's it. Basically what Guido would have done previously before he retired.

50:29Now, how the mechanics work is we hold elections. There are five seats. Everyone goes up for election every year. We call for candidates. You can self-nominate for candidacy or you can nominate someone else. You do not have to be a core developer. You just have to be nominated by a core developer. You can actually have someone outside be nominated, but it's never happened. At least no one's ever been elected. I think maybe someone outside has actually been proposed and didn't get on. But no one's ever actually been elected that's not been a core dev. So that all happens in October, November. And then the first half of December, we do the voting.

51:11We just switched voting systems to star voting. We can go into that if you want. And then that gets all resolved December 14th, 15th. And then I think it's 14th, typically. and those people start January 1st and they serve for the year and that's basically it. So we kind of have a lame duck steering council in December maybe a little bit in November but some things still get decided where the sitting steering council doesn't feel like it'd be a burden on the next steering council or in any way controversial if they had made the decision so they're not totally stuck. Basically December is taking off which kind of makes sense based on holidays around the world.

51:52Yeah. But yeah, the only abbreviated one was the very first hearing council because that all happened February. Just because it was the first one. It's just timelines of choosing the actual governance model and all that. It happened to start in February and then decided, you know, let's have it end and start on January 1st. So that was a little abbreviated, but not that much. So yes, if we want to talk voting systems, we can totally talk voting systems. Do you have opinions about voting systems? Is it worth talking about voting systems? Well, so, I mean, maybe depends on who you ask. So, so if you actually watch the Python documentary, disclaimer, I'm in it.

52:33I mentioned how I found the whole thing of Guido rage quitting being BDFL, stressful, really stressful to the point that I actually started to choke on my own throat at a restaurant once due to the stress of this or that. It's past tense now. And a lot of it was around voting systems. Because the problem is, is when Guido quit, Guido said, I'm leaving it up to you to decide how you want to run yourselves. Yeah. Period. Like he literally just said it and then Mike dropped and backed off. Right. Now he was still around and he was listening and watching, but he was not driving anything, which meant we actually had to decide how we were going to decide how to run ourselves.

53:17Right. Which is tricky. Right. So how do you decide? How are you going to decide? We've never had to worry about this. The bear is sticky with honey. Yeah, exactly. When you've never had the system in place, because you always just propagated things up to Guido to make a call, we didn't have anything in place. There was no concept of voting on the Python core team because we never had to vote for anything. Guido might ask for our hands, or people would say plus one, plus zero, minus zero, minus one, but there was never a formal way to do it. and voting systems are one of these things you can really geek out on if you want because there's a whole bunch of literature about how certain voting systems lead to certain gamification and stuff versus not like first past the post which is how governments like North America are typically voted on versus whoever gets the most votes but that can lead to games with stuff and then there's like approval voting where it's more like you just say I like this person or I don't there's a whole bunch of stuff so that took a long time just to choose how we were going to choose how to govern ourselves now eventually how long nine months a baby level no it was july till november i think october maybe okay but the deal was is there was a lot of strong opinions were like really strong opinions and it was one of these things where there were times where it didn't look like we were going to reach agreement and because there was no way to even choose that system, it had to be kind of just general feeling consensus.

54:49Because there was no way to really know. There was no way to know, yeah, we've all agreed. How do you know that? You just had to have a feel on a mailing list. Or maybe at a core dev sprint. Just kind of a room feel that we all kind of basically agree and we think this is going to go and no one's going to throw a fit and cause a big stink over this. So even that was hard. The way I always said it is there was a lot of soft power being thrown around by some of us core devs who had been around for a long time and had respect of enough other people to go like, okay, I think we all have this going.

55:18What do you all think? And kind of like push and force people to make it kind of vocally say yes, no, or stay silent as an abstain. And just kind of get a feel of the room like, okay, okay, everyone good? Everyone good? And just kind of like declare, okay, bring vocal grade and move on. And hope no one throws a fit. It was really weird. But it was stressful too because you never knew. You never knew if someone was going to come forward and go, I don't like it. Because it's not like we're in an office where we're all chatting at the water cooler all day long. It's just emails. So it's like sometimes you're talking to someone, sometimes you're not.

55:49And someone can just disappear and not say anything. And so they just drop in and say something. Where if you were in an office, you could totally see it in their eyes. Like, oh yeah, they're not happy. Something's going on. So it was hard. And it was really stressful. And a lot of it was around voting systems and choosing how we were going to choose that voting system. Once we chose the voicing system, it was pretty smooth sailing. Because that was when we had all the PEPs to propose governance models, and we just used the voicing system to choose which one. Which PEP was this? The one that won in the end?

56:20It became PEP 13. But originally it was PEP 8 ,000-something, or 6 ,000, maybe? I think it was 6 ,000. I can't remember off the top of my head. But there's a section of PEPs that had all the proposed governance models. And we just voted on all of them, and the one that won was the council won. Gotcha. There was even one that was basically anarchy. It really ran the gamut. There was another BDFL option that kind of didn't land because the PEP didn't name who the BDFL would be, and some people liked that, some people didn't like that. Some people were honestly afraid people they didn't like would be named BDFL, so they wanted to hedge their bets against that.

57:02It kind of ran the gamut. yeah um but yeah so voting system a big topic oh my gosh yeah it seems like but then you recently switched to stars you said yeah so we originally started with uh approval voting where basically you just said here are all the people i'd be happy if they ended up on the steering council so if you had 10 candidates for five seats you didn't weren't limited to five votes you could vote for 10 people or nine or eight or seven and basically it kind of makes you go like okay You're just basically saying, I approve, I don't care if anyone I approve gets on, it's just, these are the people I want.

57:36But it kind of waters down your vote in a bit, because you can't really say, I really hope this person gets on, but I'd be okay if this person gets on. So I might really want Adam to be on, but if JR gets on, I'm okay. There's no kind of expression of preference with approval of voting. So STAR voting, and it's all caps for STAR, so I think it's an acronym. star voting actually lets you rate so here's my favorite here's my least favorite but you can rank them anywhere you want from a range of uh zero to five where zero obviously means like no no thank you there's a meaning behind this you want the meaning behind star voting sure would it add context i don't know i don't know wikipedia is accurate or trustworthy but let me just let me just say that's the source okay star voting is an electoral system for seat or sorry single seat elections the name and allusion two-star ratings stands for score then automatic runoff referring to the fact that this system is a combination of score voting to pick two finalists with the highest total scores followed by an in quotes automatic runoff in which the finalist who is preferred on more than on more ballots wins it is a type of cardinal voting electoral system yeah Did you guys do the automatic runoff part?

58:51Yeah, we did. So what happens is basically you rank all your candidates from five to one. Or zero, right? If you don't want them at all. But the interesting thing with this is you really want to rank the people you really want on as five and the people you're just okay with but totally don't care about being on as one and then spread everyone else across like peanut butter, right? So you really do want to show a preference. You don't want to put everyone at five because it's like no difference. You really want to show a preference, which is the big shift from approval. But yeah, it just does a runoff.

59:18And then it goes like, okay, who's above in all these ballots at this point? And then, okay, well, it seems these are the top two. If we just have these two run, what does that look like? Okay, this person wins, they're on, take them out of the ballots. Now we do it all over again and just keep slowly pulling it down to see where people's preferences land in terms of, well, this person got all the fives. Because there's a balancing of each other like, oh yeah, this person got fives but only got like 10 % fives versus someone else who got like 30 % fours and just how all the machinations of the numbers and the math and all that work out.

59:51But it was just interesting that we did that shift to really kind of let people express that preference. It's not just a binary on or off on the council. It's a, no, I want you on versus, yeah, I really want you on too. I'd be okay. I won't be upset. But if you don't get on, I'm not going to cry either. And that was the big shift. And that all just happened in this last election was the first time we ran it. And it went just normal, probably. Yeah, it was interesting. It was smooth. Everyone seemed to be good with it because it didn't change it such that it was like only vote for your five people.

1:00:27You could still express yourself, but it wasn't really complicated to express yourself. The only trick was you had to understand you wanted to run the full five-point gamut. You didn't want to just pool F5. And then Tim Peters, core developers, more or less is retired from participation actively but still comes around on occasion as a big voting nerd and ran a whole bunch of things on it and showed that the outcome actually matched up with what would have been expected. And that, yeah, the system worked and it was, people got in the way you wanted, even if you ran different kind of voting systems and all that.

1:01:00So people's preferences were expressed appropriately and the top five people who were preferred going on. The more interesting thing was that on this council run, for five slots, only six people ran. so there was a good chance you were going to get on um actually one person who didn't get on actually was previously on the council so they got basically voted off but there there was new new person who got voted on so there was still good turnover but yeah we're talking about like we have talked about as a team like okay why was there so few people and i mean it's the usual thing right as i said it's there's a time commitment you've got weekly meetings you have office hours, which I went to.

1:01:43Once again, I'm just a normal core dev now, so I still have to book office hours with the steering council, talk to them about ideas and stuff. As I said, you talk about it, you read peps on your spare time, you respond to things, answer questions carefully, because as I said, your opinion now has a bit more weight to it. There's that aspect. We also didn't know if people were worried about turnover or not. It just seems like people are honestly just busy. there was some not community stuff, there was some code of conduct issue stuff that happened about a year ago and that complicated some things that probably made some people go like yeah, you know what, I don't know if I want to deal with code of conduct stuff but the steering council kind of fell on their lap as I said, final design around stuff you gotta steer things, you gotta do things the bug has to stop somewhere yeah, and I think that may have scared some people off like I do not want to be that person who has to publicly say that person did something bad and they need to take a step back.

1:02:42Well, it's hard enough to volunteer your time and thoughts and efforts and then to also have to like do stuff that nobody wants to deal with. Like, well, volunteering to do that is much harder to get people to volunteer to do all the things, but someone's got to do it. Well, and you hear, you find out stuff you did want to know, right? Like I was on, I mean, I was on the steering council, I was also on the Code of Conduct Working Group for the PSF, for the Python Souser Foundation. I was like, you get emails for reports about people that you know. I was like, I really wish I didn't know this. And so it's also got that aspect to it.

1:03:16So it's not even what publicly you have to say, oh yeah, this calls someone out for doing something bad. It's like you also get told stuff that you just didn't want to know, even if it doesn't lead to something. I was like, yeah, personally I'm not happy with what you did, but not professionally. But at least from an open source perspective, it doesn't affect the project. and it's none of our business so it's right we don't know not necessarily i don't know why you told us but it's like we hear you you've been heard but now i have this weird little secret in the back of my head that the public doesn't know and i i have to live with it you never wanted to know it in the first place yeah you can't unlearn what's well it's what we tell our children especially with like movies and certain things you want to like my friend saw this or whatever why can't i not watch this it's like well you can't unsee things you know especially like horror movies or certain things you just yeah right don't want them to be exposed to early on it's like you can't unsee certain things you should really protect your innocence as a kid and it's not exactly the same thing but it's it's a version of that like you want a version of i would say bliss right you want some version of bliss to operate from just because bliss is kind of nice you know it's kind of nice and when you can't have it because it was taken from you because of your responsibility or role you might regret or resent your role our our little our little ones really sweet kid.

1:04:30No stranger danger at all. Totally will say hello, goodbye to strangers and wave and all that wonderful stuff at this age. The day someone is a jerk to them, oh man, and that innocence is a bit lost versus just like, because right now at this age, they're just going, what do you mean mine? I thought we all just shared everything. What's going on? It's still kind of like, I don't get ownership. No big deal. Why won't you share? Why'd you take it from me? But the day where that starts to click is like, oh, people can be pushy and not be nice and all that stuff. It's going to be a sad day. You want to make sure that as a parent that it's coming.

1:05:11And hopefully you give them the tools to understand why the world works that way sometimes, but it's not always. And how to cope with and deal with it. Try not to let that not hope, but that innocence carry on. That's the age of innocence. Yeah, let them carry on being good people and understanding. It's okay. It's going to happen, but here's how you deal with it. And here's how you continue to be a good person. Don't let it totally darken your view of the world and get bitter. I mean, it's one of the reasons I actually stopped being in the Kodakon of Working Group. I'm going to have to be done with this as my kid.

1:05:43I don't want to be done with this open source community as well. I need to kind of, not compartmentalize, but kind of step aside from that aspect of dealing with that stuff. and I'll just worry about with my kid and let that be that. So as a regular old core dev, did you have to go to the steering council for this whole lock file standard you've been working on? How did that come to be? So actually, no, because the steering council has devolved power for packaging to select people. So this is another kind of side effect from when Guido did things. So Guido never cared about packaging. It was not his thing.

1:06:23He's just like, no, not my thing. Let the community handle it. So the community handled it. And what ended up developing was kind of, I mean, not camps, but basically just groups of people who were working on packaging outside of the core dev team. Now there were people like me who were core devs and working on packaging stuff, but it's always been a separate group of people. And so when the steering council got set up, there's a pep that says, for packaging decisions, we perpetually defer to the select people to make calls for anything related to packages. So technically the steering council still gets final say, but they've just said, yeah, but these people are the experts in the area, so we're going to let them do it.

1:07:03And this happens sometimes. They're called PEP delegates, where the steering council goes, we don't have the expertise in this area, you do, we're going to let you make the call. Now it doesn't happen too much with the steering council, it's happened more under Guido, just because the steering council is broad enough, usually across the pipe people that council can get enough information to do it and they'll just ask expertise, opinion, and help. But with packaging, they just said like, yeah, forget it. Over there. You all can handle it. You've been doing it this long. You keep doing it. So I had to make Paul Moore, who's one of the perpetual pep delegates for packaging, happy.

1:07:37So that's actually who I had to make happy in the end. Okay. And that took you four years. yeah six if you measure from when I first mentioned it on Twitter okay yeah I mentioned that back when it was called Twitter probably it was yeah six years ago yeah we're actually getting close to we're coming up on seven now actually based on these timelines okay so yeah well as my as the guy who built our house always used to say it's just stuck with us it's a long and windy road Brett it's a long and winding road yeah it sounds like this has gone a long way Now we're talking lock files. So, so, um, couldn't you just look at the, what MPM did and just be like, lock files, they got to figure it out.

1:08:22Just do that thing. I don't know. If you ask the yarn people, I don't know if they had a degree back way back. Yeah. Okay. So maybe the cargo cult isn't the move. What was the move? How did you figure it out? And what'd you figure out? Well, so when I, so after the Python two to three transition, right, I had a goal during that time to try to make it so that anything that was in language that changed would either be an error or be a warning or there would be tooling to help you just make the transition for you and you wouldn't have to worry about it. So there would always be some way where you didn't have to magically look at things.

1:08:56Something would warn you or your code would just make it happen. When I got that done, I looked around for other projects to work on. You know what? Packaging is a thing. People always complained about packaging and Python being hard, blah, blah, blah. Now, granted, this is well over a decade ago, obviously. So I'm like, okay, I'm going to go help up packaging people. So I went over and started to participate over there. I viewed it as kind of trying to make things a choice and not a requirement when it came to tooling. So setup tools, for those of you who know Python, packaging and all was the long-time tool everyone used to package up stuff for Python.

1:09:34But it was the only thing people used to package up for Python because it just kind of was there. It just continuously was there in the thing everyone used. And so what happened was everyone just through convention just assumed set of tools was there and that's what you had to use. So it had a lock on everything, not necessarily by design or choice, it was just there, which kind of sucks for the set of tools people because we can't change anything. We're going to break the internet. It's one of those, we do something and everything's going to go awry because we're just everything everyone uses for this stuff.

1:10:07And so we worked towards making it so that setup tools was a choice and a requirement. So if you know Python packaging, there's a thing called pyproject.toml that I helped create with a bunch of other people. And inside of that, you can specify your build system. So you can say, I want to use this tool from this version of requirement to build my package or build my project. So that kind of separated out setup tools. And so that put me down the road of realizing that the artifacts are really the key thing you really care about. It's the things that get produced, not what tool you use to produce them.

1:10:43That's the real commonality and the compatibility layer that you really care about. And so I sort of look around at the next kind of thing of what was kind of locking people into PIP. What were some of the things that PIP was doing as an installer that wasn't shared by the community? That was very PIP-specific. Not because the PIP developers were trying to trap people in PIP, but it was just convention and what they had done, and it just kind of got pulled forward, and everyone would use it. Requirements.txt files, if you're familiar with those, was one of those things where you just kind of write down line by line what your requirements are, and that was it.

1:11:19But it had no structure, it was PIP-specific, and it just had some quirks to it and all that stuff. So I was like, yeah, you know what, I'm going to try to tackle lock files. Because also at that time, tools were starting to come out, that were managing workflows, and they all started to do their own lock file format. I was like, okay, so if you want to deploy to your favorite cloud host, and they want to do the install for you, how do you make sure they support the tool you want to use? Well, you can't. You have to list. I am going to support this tool, this tool, this tool, because your lock file that says what to install is only made for that project.

1:11:54That kind of sucks. So I said, okay, you know what, I'm going to try to solve this one. And so I tried. And it failed initially, because I tried to lean in really hard on security, which cut out building from source. Because for anyone who tries to build from source, and you don't know what compiler you're going to use, it might compile on Tuesday, but good luck having it compile on Wednesday. So I said, I don't want that headache. So I may only use wheels, which are the binary artifacts for packaging that Python has, versus source distributions, which is the source thing. And I did it and the community said, nope, don't want it.

1:12:30If you can't do source distributions, not going to work for me. And Paul went like, yeah, enough people are saying that they don't want this without this, that I'm not going to take this. So it killed that first version. So I went back to the drawing board and I'm like, okay, I had some co-authors on that pep. I felt bad that we did all this work and it got canned. I used their time up to do this. So I'm going to try again on my own. and as part of that, I'm going to try to develop my own tool, almost a pip competitor as a proof of concept, to show how this would work. So I started to have to re-implement pip from the ground up, which is partially why it took so long.

1:13:08I had to write my own resolver, I had to be able to read metadata from packages to figure out what they listed as to install, literally everything ground up redo. It's built out of some libraries, but not all of it, So not all of it was available to me. So I was like, sorry to tease stuff out and do all that. And then I finally got far enough along with that. And I just came back and said like, look, okay, here's the format. It's readable. I can do it myself. So I know it's doable. Here's all your other log file formats. I've looked at all of them. Here's the commonality across all of them. So there's no, I'm not doing anything revolutionary here.

1:13:47It's more just taste decisions of like how to format it and what information to keep and all that. But it's all there. And it was just a lot of talking and a lot of figuring out what everyone wants and their lock files and all that stuff. And it's just, once again, a lot of players where it's basically working for me right now. So I don't have, I don't feel the need to have to change. And so I'm going like, I love this because suddenly their tool, which isn't maybe not the most widely used, would pretend to use more because now they don't have to do it. Or I don't have a lock file format. I'd love to have one.

1:14:18I'd use this. so it was kind of all over the map but it just yeah it's a lot of conversations uh and i started to butt up against the due date for our kid so i actually had to start putting down deadlines uh which actually worked out really well um saying like if you have a week to give me opinions on this it's not listening to you past this and people are totally cool and understanding because it'd been going on for so long i've totally listened to the deadlines and it all all got done March 31st, 2024. But yeah, the reason we couldn't just use package lock dot JSON and all that is Python's packaging story is complicated because we allow you to be to use Python as the glue of the internet.

1:15:05Right? Like the fact that you can ship pre-built binaries of like C code that you pull into Python, right? Node doesn't have this. If you look at any Node package that has like pre-built binaries, they have their own little hack in their install script to like, I'm going to go download this custom thing based on what I think you're running on, right? But Python packaging solve that for the community overall. And so there's a lot of complexity around that. And it's how do you handle that, right? Like, how do you make sure you grab the right thing on Mac versus thing on Windows or things are only on Mac that you care about, but not on Windows, right?

1:15:36Because that happens too. And then there's got version compatibility and all this stuff. And it's a lot of stuff to have to keep track of. To the point that we're starting to see people try to shove stuff on PyPI that are not Python related just because we've solved these problems already and their community hasn't. So they're trying to piggyback on PyPI. Now, I would suggest you don't do that because the discussion is suggesting legally it might not be a good idea. At least for PyPI, right? Because it's legally specified only for the Python community. Don't take this as an idea. Please don't do this.

1:16:09But if it's Python related, go ahead. Anyway, it's being discussed, but that's why it's complicated, right? And we can't just steal someone else's approach, right? Like Rust has their nice thing with Cargo, but they just compile everything from scratch. You can do it by nice. Same with Go, right? Node's a little different because you can have multiple versions of different things. As we know, Node modules can get really big with a bazillion copies of the left pad with different versions. We don't have that. We have a flat namespace. So you've got to have one version of anything ever installed at once by choice.

1:16:42And so it just leads to different decisions, different complexities. And it just took a lot of work to get everyone on board and get it all lined up and make sure everyone who had a log file was mostly happy. UV was kind of one feature shy of being happy, but the thing that they weren't happy with is not standardized. And I was not about to be the one that standardized it. so I'm kind of waiting to see if they ever propose workspaces basically, kind of from monorepos like workspaces from Rust for Python, because they said that was the one big missing thing so I'm hoping someday they'll do that, get that into the spec and then they'll actually even use the standard as their primary, instead of their uv.lock but everyone else is on board and doing it the other holdback is pip will write out a lock file, but they don't read and install from them yet just because it's all volunteer run and people just have to have time.

1:17:35It's interesting to think about how the package manager shapes or could shape the language, like workspaces, for example. Can you hypothesize how that may translate to language change given that you've steered before? Well, it wouldn't be language change as much as just how do you specify potential stacking of requirements, right? Because if you're thinking of a microservice, that's all microservice run app that's all in one monorepo right so you're going to have sub projects that are designed to run independently and that's the whole point of microservices but you might want to make sure everything runs smoothly so you might want some of those requirements to propagate up to a bigger lock file and have that all i suppose you're right in the case of cargo cargo probably hands a lot of that for rust yeah well and because it's a language thing it's like the way the language what exactly is cargo to rust i don't know what is it is the compiler is it the many things you know there's a lot for it so how do you like describe what it is yeah so cargo is a so rest c is the compiler and cargo is kind of the workflow tool for rest so cargo is the thing that will download stuff from crazy.io for you tooling yeah it'll run rust c for you so i'd call it the workflow tool and because rest made the decision that we're going to compile you list all your dependencies and we'll compile that into a single binary, they don't have some of the complexity that we necessarily, but the whole workspace thing is you can say like update everything out once with one command and it goes into all your sub projects.

1:19:01Or you can say, I, I require this, this crate that I keep in my monorepo as a dependency and it treats as like a local dependency and has it all worked out. So it all kind of plugs together. Right. And it's all kind of just works. And it'd be UV's job to manage that, not the language Python. I was incorrect on saying the language change. I get that now. Yeah, and because Python is that flat namespace, it kind of becomes a question of like, okay, like, all right, your microservice over here says I've recorded this version of spam. This other one here says they need that version. How do you match it all together so that if you have a single deployment on a single Python version and you all installed in the same place, even though you're all developed independently, how do you have that all kind of propagate up?

1:19:41Or how do you do it independently and then have it still kind of pulled together and all that stuff? So there's a bit of complexity to it. But hence why Ashroll said, yeah, this is something we really need before we can use this as our de facto lock file version. Now they support it. To be very clear here, UV will export to a pylock.tomle file, which is the file format that I developed, and they will install from it. So they're totally supportive, and they were participants in the discussion. That's just not what they will use by default. They still have their uv.lock because that feature was missing.

1:20:11As I said, I was not in a position to want to try to standardize a new thing. I just wanted to have a way to write down stuff we already knew or had and did in a way that everyone could kind of just rally around. What do you think about the rise of Astral and UV in the Python community? It seems like they've gone from relatively obscure to like almost everyone's default suggestion. Just use UV. What do you think? Okay. I'm going to preface everything I'm about to say here with I know Charlie. I know multiple people at Astral. They're lovely people. I want to see them succeed, et cetera, et cetera, et cetera.

1:20:50I have nothing against Astral, okay? To be very clear, I want people there. Because I've been accused of trying to kill the company before. This is Rigby all over again, man. That's two Silicon Valley jokes I've dropped in this show, and nobody's gotten them, okay? Just so you're all clear. I don't acknowledge on purpose now. It's part of my game. It's two. I honestly, when I booked to come back on after you all called me out and said a lot to come back, I almost tried to figure out if I could somehow squeeze in time to rewatch Silicon Valley before I came on, but I couldn't make it work. So unfortunately, that's a lot of squeezing considering your time constraints.

1:21:23That's why it didn't happen at all. Right. That's the thought that counts, though. We appreciate that. It's okay. Well, Adam appreciates it. I don't care. Continue. I'm sorry. No, it's all good. All those things, but we're at the but. The cliffhanger's been set. We're going to cut it right here. So this is the only part that's going out. I'm just kidding. Yeah. UV came out February of 2023.

1:21:44I know this because that was when they, February? Yeah, I think they announced February. Because I think my version of my PEP for LockerFiles came out July. And I was like, and it was totally taken by surprise. I was like, I've been working on my own resolver and this stuff and some of the features you have for like a year by myself here. And suddenly it just drops on me like, oh, where are you doing that? Like, okay. So hence why I know the date. One of the things that UV popularized but actually predates them is the idea of workflow tools whose focus is to get your Python code running, not getting your Python interpreter running.

1:22:27It's a very succinct distinction, but it's a key one.

1:22:33PDM, Hatch are the other two big ones next to UV. what happened was there was a project called Python Build Standalone that does pre-built Python binaries that are relocatable so they could drop the target I don't know what if it's GZ or XZ or whatever on your desk anywhere and it just works so what happened was projects realized oh you know what if I can auto install Python because that's one of the big complaints that some people have installing Python is a pain because we don't come in box and all OSs anymore like we used to. And the installs are different per OS. So people are like, there's no universal explanation on how to do this.

1:23:17And there's a lot of little details on how to get things working. We have virtual environments, which is our isolation mechanism. So when you install stuff, you don't stomp on other things in your machine. And so it's like, all right, you got to get Python somehow, get that installed. Okay, now I got to create a virtual environment. Okay, now I got to get everything installed into my virtual environment. Now I can run my project. Lots of little steps. And what UV popularized in Hatchapdm, the decent job that they did prior, was go like, you know what, we're going to give you a unified tool that does all that for you.

1:23:49So like UV run, Hatch run, PDM run, they all do the same thing. They go like, all right, you need Python version something based on the metadata of your project. Do you have it? Now, okay, I'm going to go download it from Python build standalone, which is an astral project at this point because the maintainer didn't want to run it anymore and took it on for the community. Download Python, install it on your machine. Now that you got it, okay, going to create a virtual environment in your project. So that's done. I'm going to figure out what needs to get installed based on what you said, which could be nothing, just the default install, whatever, or you maybe added extra fields for extra stuff.

1:24:27Install that, and then you said run. Now I can get the Python I just installed through the virtual environment that has old things. Run your stuff, get your code running. and just one command, all done. And that's a breath of fresh air for a lot of people. Because as you heard, that's a lot of little steps that are easy to miss for people and AI to throw shade at everybody. And UV really popularized it, right? Because UV got such a big splash when they came out with UV due to the performance numbers that they were getting. That everyone went, oh my god, this is amazing. And then they did, and it's like, oh my goodness, and it does all this stuff.

1:25:02and so total credit to them they're great at marketing and they land out a good time their performance is exquisite and they are really good about launching new features when they land in the packaging ecosystem and so they get out really fast with them and people see it and so that's it but the concern the reason we brought this up is just about having concerns and this is the usual it's great but QV is a VC backed company they have to make money not that I think there's any ulterior motive here but that might implode. I hope they don't I know the company, I know people who work there I hope they also have jobs along in the future and I hope they're very successful but there's always that concern we've all seen tech just kind of have problems I know some people do not love how certain things in the node community happen behind companies because of how they manage things it's a concern and once again you never know what the future is going to hold you never know if it's going to be somehow money pressures are going to cause them to have to make a decision as I said I fully expect that that will never happen but there's always that risk and when everyone's instructions are use UV it's a slight concern it's not a huge concern I don't think anyone's panic but I do know people for instance who flat out refuse to use UV specifically because if it goes away on them it's going to suck to have to transition everyone else to other stuff.

1:26:32So they'd rather just use their own thing. So Brett's doing his thing of, all right, how do we help push things in a way that there's kind of agreement among the workflow tools of how things should work, right? Standardized so that other things can do the same things and that we don't have epic lock-in in case of a massive failure. Now UV is open source, right? MIT licensed even. Yeah, MIT licensed, so very permissive. there's probably company things that are required I imagine does it touch server side stuff that's like Astral run server side things or I don't know they've announced that I think they have an enterprise thing that's right that they will run on your behalf as an index server for surfing on Python files so that's what they have and it's designed to tie directly into UV so this kind of plays into it right like UV has stuff specifically for their enterprise products.

1:27:28I mean, it's typical freemium, right? Which I don't think anyone's going to balken to be upset with. No, but props, by the way, to Astral for picking up that, what'd you say it's called? PDB or something? This project that was no longer maintained? Oh, the acronym's PBS, Python Build Standalone. Python Build Standalone. So that's cool. You know, like that guy didn't want to do it anymore. It couldn't. And so here comes somebody else. And so hypothetically, and probably even practically, UV as a project could be the exact same thing in the case that Astral fails. Somebody else could pick up and run with it, right?

1:28:00Yeah, exactly. And this is why some people don't worry at all. You have some people going like, whatever, I don't care. If they misstep, we'll just fork immediately because it's MIT and we'll just maintain it ourselves. Although granted, it is in Rust, not in Python. So the talent pool for taking it off from the community is not as broad as if it was in Python. But other people are like, yeah, but I don't want to have to worry about that step to begin with. And the longer they go as the leader, the more ingratiated they get and the harder it's going to be to switch. So me being me, I'm trying to make sure that the stuff UV is doing from a very baseline experience is kind of less special, not special, because there's things they all do.

1:28:49It's just trying to get them to all kind of go like, yeah, this works, and if you do this, everyone kind of agrees in some way. So I'm currently working to try to get Python.org to be the producer of these pre-built binaries, not just Python build standalone. There's nothing magical they're doing. They've got some patches that I'm working with them to get upstreamed, which they've always said they want to do. It's just a lack of time. This is one of the things, right? Startup, they only have time crunch all the time. So they're trying to slowly get them up, and I'm helping them as best I can. So props to them for trying to help that.

1:29:22And they honestly want us to own it because it's way easier for a whole group of people to own how these builds work versus just them. So they're working with us to get that so that we can start distributing baseline pre-built binaries from Python.org. And then they'll probably have some extra optimized or more esoteric platforms that enterprise customers might care about that we don't. Which is a perfect fit. Letting them own enterprise is great. That's really where I want to see them shine. And for the general community, let them be an option, not a requirement. The other thing is I have another pep that I are talking with some people behind the scenes, including them, about standardizing virtual environments and where they go.

1:30:03Because weird problem here, because everyone kind of sticks virtual environments wherever they want, tools like VS Code or your editor, whether if you're a sublime text user like Jared, I assume you still are. Maybe you're on Zed now. I can't keep track. I'm on Zed. You're on set now. I finally made the switch and it stuck for the first time. Congratulations. So whether you're VS Code or Zed or whatever, your tool doesn't know where to find this virtual environment that's so critical to your project because every tool does it their own way. So I've written a pep to try to standardize where it goes or at least have a marker file to tell you where to go to get it so that all the tools can kind of just know where to go to look to find the thing.

1:30:42Because it's such a fundamental thing for discovery. But it's a little stuff like that that I'm just trying to clear up so that if I recommend using UV and for some reason it doesn't work out, you just have to say, go use Hatch. It's going to flow through and work nicely together. That's where I'm taking it from. It's not because I want to take away from enterprise or anything from a company perspective. It's just the baseline everyone in the community wants experience should be universal enough that swapping tools is small. and then when you want the enterprise special stuff, totally get attached to UV, totally grown with it.

1:31:17You can totally just attach to the hip to it and go for it. That makes total sense to me, working at an enterprise company. The way it is, enterprises have such special needs that that's not something we're ever going to standardize. But for the stuff for kids, educators, hobbyists, when your models write you a little Python script and you're just like, okay, I just need to run this real fast. UB is totally a good, reasonable thing to say and use right now. And as I've tried to be very clear, I have no problems with that being the thing. I want to make sure that that never becomes a problem and just hedge our bets.

1:31:52It's more the thing. So that's kind of what my current big project is. Because I want my little kid, when they go to school, and if they learn Python in school, I don't want them coming home and saying, Daddy, why is it so hard to get Python going? I want to make sure I can say Amen So I want to make sure that they have and everyone on the planet subsequently has a good experience getting Python running and I think that was one of the big things that happened in 2025 was the community going you know what I like this workflow I like this tool that I just say run my code Python code and it does all the other little details to get my code running that was I think the big thing and that's what I want to work towards making the norm and doing what I can to help UV and these other tools out to make it so it's acceptable to more and more people to do because it doesn't lead to weird issues with your cloud provider or your tooling.

1:32:54It's not going to plug in with Zed. It's like, well, no, no, Zed just plugs into the standards. You could totally use PDM, you could totally use Hash, you could totally use UV and it's just going to work and no one gets left behind either. right because i always like to help the help projects thrive because they all have little innovations that help with their projects or anything else and that's kind of helps builds up and then people try workspaces right if uv helps and then they try it out and that we get a good experience because everyone's trying to learn the thing and now we can standardize that right so take away the the drudgery lower bit and make everyone be able to innovate at a higher level well if they're a if they're a business right when you go to astral.sh and you hover over products you see tools and platform all their tools are liberally open sourced right uv is dual licensed it's actually uh apache 2 and mit and you can choose whereas the platform pyx uh what'd you call that picks i don't know pyx pyx that's what i say in my head i i think all of theirs are not or acronyms so you actually say all the letters like ty is not ty it's ty so i from for all their For all actual stuff, I say all the letters.

1:34:02Would you say UV then? OOV. This is OOV, man. You've been saying it wrong the whole time. I'm just kidding. But okay, so PYX. Yes. If I'm leading in on business model, they are trying to build massively adopted tooling in the Python world. And they're venture backed. It seems like PYX is the way they jettison into products. Correct. Although I just, you know, caveated the fact that I hovered over products on the website. Yeah. Sure. Tooling is products, but they're not products that you generate income from. Yeah. They're products that you extract income from by way of more like your future child, not being upset about getting Python running more easy because maybe UV solved that problem or X, Y, or Z.

1:34:51Maybe it is truly X, Y, or Z with them considering PYX. Yeah. What are your feels on that? What do you feel? How do you feel about? I mean, I suppose we get back into NPM world a little bit, right? They were a separate venture-backed company acquired by GitHub. We know the story there to some degree. There's some challenge there with NPM inside of GitHub, security concerns, massive usage. I mean, it's success. Yep. A major success, but just has some security challenges. And, you know, not a lot of resources behind it. And we've speculated on that in recent episodes. We'll link them up. But if PYX, the next step in Python packaging is it, how do you think this will fare for them to turn this into a revenue generating packaging manager?

1:35:36I mean, so without having talked to Charlie about the business model, other than what he said publicly, yeah, I think basically what happened was is Charlie made rough while he was in between jobs and really knocked it out of the park and got really popular. And he's, oh, this is really cool. And people came forward like, hey, I want to give you VC money to try to fix Python tooling. And he took it and he built an amazing team. There's a lot of Python developers and really top people in the community who went, yeah, you know, I'd love to help you do that because this is a problem I see and I care and I want to see that happen.

1:36:13And it's why I'm not really concerned about actually stepping wrong. This is all just hedging bets just because we're all paranoid. And so he built that team and they were able to build UV and get it really popular and get really broad usage, right? which I think has won them a lot of name recognition and a lot of just getting into companies. Like, yeah, just use UV, just use UV, just use UV. Which makes it way easier for them now as a business. Like, hey, if you need an internal package index for Python that's highly performant, will give you whatever enterprise stuff you need, like backups and just making sure you only give packages that you've vetted or only have certain life senses.

1:36:54It's like all the enterprise stuff you typically need at a level. Come use UVX. Buy UVX from us and we'll work with you. Because I know Charlie's also done a lot of work talking to enterprises to find out what they need. Which is why UV has certain knobs on it that are very good for enterprises that most of us will never care about. And I think that's what they've done. They've gotten enough name recognition and have already just gotten in enough people's workflows that it just becomes the second nature thing to just pick up. It's like when you're on any of the cloud providers, whether it's Azure or AWS or whatever, it's very easy to just say, oh, I need a solution for this.

1:37:31Oh, you already got on the cloud platform run? Fine, click a button, you're done, and you're in. I think it's a similar play here, which totally makes total sense. And once again, PYX is totally an enterprise thing, and Charles made it very clear. That's why it's not an MIT thing. It's an enterprise thing where we expect to make money, and it's just going to do whatever enterprises need to happen, and that's cool. And he's used that to help fund UV. I mean, private registries make sense, obviously, right? Enterprises, private registries. You got, you know, maybe even internal tooling, lock and key.

1:38:03You got a lot of opportunity there. I think. Oh, yeah. There's a lot. You've said Pi PI, which I thought was did you catch that, Jared? Pi PI? Instead of Pi Pi? Okay. I'm going to check you on that because that is a little strange. I've never heard it Pi PI, but that's cool. It's because I haven't done an audio book for you, Adam. That's why. I was going to take Brett's pronunciation as gospel. Astrophage. I actually did a whole lot of new talk on this at PyCon one year. So it is PyPI, and in PyPI, which predates PyPI, is the Python JIT interpreter written in Python. So P-Y-P-Y, that's PyPI. Py P-Y-P-Y is the Python package index.

1:38:48and that is why there's the distinct pronunciation difference so people don't that makes sense pi makes sense package index okay i stand corrected i thought it was a private investigator well that's that too uh funny backstory to and to let you know that python has a sense of humor originally it was learning the cheese shop after the cheese shop sketch from monty python because python's named after monty python not the snake but people were worried pointy-haired, to also date me, pointy-haired managers wouldn't take Python seriously if our index was named the cheese shop. So we went with PyPI.

1:39:24Now, if you get a 404, by the way, on PyPI and scroll down, you will hit the cheese shop sketch as hosts on YouTube by Monty Python. So we've kept the name around a bit. That is nice trivia. Let me see if I can get this done here.

1:39:52this guy's not a liar people try it works i do have a dad brain though so it means i do forget and maybe get a little facts wrong but it's not on purpose yeah i've had machete order out of order but you got the uh cheese shop so like i think it's great that ashley's gotten this money and they're putting it back into the community and stuff. As I said, I view this as a hedging our bets as a community to kind of just help alleviate the load on them and anyone else who does this workflow stuff from my perspective and just trying to do that. But it's tricky, right? Admittedly, I went to office hours for the steering council last week and I talked to them about this, would you be up for leading this kind of stuff?

1:40:32And they were like, pre-built binaries we get, that's totally something that's in our wheelhouse as it were but running workflow tooling is outside our purview we're in charge of the language we don't do this kind of high level stuff and we also don't want to step on people's toes Python very much tries to not tell people what to do this is why we have multiple type checkers that will try to type check to a single spec but there's not one shipped inbox with Python the community drives a lot of this This is one of the drawbacks of not being driven by a company. But it's also one of the great things about not being run by a company.

1:41:08People are allowed to flourish and create a thing, and it's allowed to make things happen. People, you learn, you kind of rally around stuff, and it slowly propagates through the community. Then you can standardize that, and then everyone gets it automatically. So they just said, yeah, we don't want to cut people off by blessing a workflow. So I'm just talking with people behind the scenes right now, going like, okay, how do we all kind of coordinate as among workflow tools? And I kind of include me because I wrote the Python mantra for Unix, which kind of does very, very lightweight workflow stuff to just like, how do we communicate?

1:41:43How do we do that? And how do we just, yeah, once again, just try to keep stuff moving so that, yeah, the community just keeps working towards things so that we all just work together and everyone benefits, right? As I said, UV included, right? So that's what I'm working towards right now is trying to, you know, the, the world is led by the daring man. and I think Charlie's one of the daring, you know, to undertake Rust for Python, bold, and that's daring, too, just to like... I did it first. No? You did it first. Okay. Well, you're bold too, Brett. So here's the backstory. So there's a Python launcher for Windows.

1:42:18So the py command, if you're on it, I know Enem is no longer a Windows user. That little tryst is over. Back, Windows, get back. back in your proverbial box. So on Windows, there's a PY command because back in the day, the way when you ran the installer, Python wasn't put on your path. It was installed in a place and you'd install the PY command, which would know where to look on your machine to get it just because the way Windows historically has done things. Okay. Backwards. No comment. No comment. On Unix, though, you didn't have this basically because Python ended up on your path, but there was an interesting benefit to the PY command on Windows was when you ran PY, it always used the newest version of Python that you had installed.

1:43:04On Unix though, if you type just Python or Python 3, it would oftentimes be the latest Python you installed. If you install Python 3.14, I think 3 is the latest release, and if you install Python 3.13.11, I think is the latest one, if you did it in that order, whose symlink would win in your user bin. It was usually the last one you installed. Sounds like the older one. Yeah, right, which sucks. Why do I always have to try to remember having to type Python 3.14 or always having to remember what the latest one is? Or if you have scripts, right, where it's like I want to shebang it to the newest version of Python, you can't really because it might not choose the right one.

1:43:47So I wrote the Python launcher for Unix in Rust as my Rust starter project long ago. I remember this, a long time ago. Because I wanted this on Unix to have this command that automatically search your path and find the newest Python 3.whatever version and use that. And then on top of that, I added a little nicety where if it finds a virtual environment in a.venv directory from your current directory up to your parent, it would automatically select that instead of trying to find Python on your path, which has a nice benefit where when I start a project right now, I just go pi-mvemv.vnv and that's the command to create a Python environment using the newest version of Python I have installed.

1:44:30And then from that point forward, I just type pi and it just always picks it up and I don't want to think about it. I don't even have to activate my virtual environment in my shell. It always is just a thing in that project and I'm done and I don't even have to think about it. It was kind of my little prehistoric precursor to workflows, tools, doing this way more involved. But this is why I'm jokingly saying to Adam, I did it first because I wrote that in Rust before UV. You were bold, Brett. You were bold. Charlie stole both your ideas. He stole Rust and Python. He stole this. Oh, dang you, Charlie.

1:45:04Dang you. I mean, he did. Already solving my problems. He did a great job. I mean, Ruff is really, really fast and really great. UV is fast, although funny enough, a guest of yours on Change Log and Friends, Andrew Nesbitt, I think. Yes. Yeah. So Andrew wrote a blog post in December of 2025 talking about why UV is so fast. If you read it, he points out a lot of it's not to do with Rust. It actually has to do with some of the algorithmic choices in terms of what they use for a resolver, because they use PubGrab, and also not having by-course compatibility concerns. So they get to wash away all the cobwebs that PIP has to continue to support and avoid all that overhead.

1:45:45So that's not to say Rust isn't faster. I think it's my last blog post, actually on my blog. I wrote, should I rewrite the Python launcher for Unix in Python? Because it's in Rust, partially because it was my Rust starter project. But if I ever get pre-built binaries done, hopefully I will, theoretically you could do what uvx.sh does. If you don't know that one, that's a new one from Astral. If you go to uvx.sh, it's a website where they will auto-gen a shell script to download UV. I think it's UV. Maybe they'll do a direct download, but they basically will download, install Python privately on your machine, then download and install the project from PyPI, and then make sure it all ends up on your path.

1:46:29So you don't even have to make sure you have Python. So it basically makes it a one-stop shop for a shell script so that the old curl, pipe, essays, run thing works for Python projects now, thanks to that website. So I've gone like, well, maybe I should do something like that. I mean, I had the idea. They implemented it nicely. So now I don't have to think about it. But keep beating you to your punches. Hey, I'm a one person dad versus a startup. This is why Brett's trying to kill Astral. It all comes out. I do not need more comments from people online saying I'm trying to kill a company. I'm not trying to.

1:47:05I want Charlie to be successful. And I hope Charlie gets everyone at Astral gets very rich. So you say. So you say. Yeah. You said one thing there that stuck out to me was pre-built binaries. What are you doing there? All right. This is getting really into the weeds of binaries on Linux. Go in the weeds, man. Get your hatchet out. So think about your historical configure make install steps to get some C code on your machine, right? From source. Yeah. Right? When you do that, paths to where to look for.so files for dynamic linking get hard-coded into the binary. Technically, it's in the RPath setting, or RPath field in your elf or dwarf file or whatever file format your OS uses, Unix OS.

1:47:54But basically, it hardcodes with absolute paths where to look on your machine to load these.so files. That means it also is very much tied to your machine and where it ends up. so like when you do like dash dash prefix it kind of hard codes some stuff that this is where things are going to be and i expect them to be there and if you move me i'm going to break right so dash prefix does a bit more with compiler instead anyway there's stuff and it gets hard coded in the thing and it makes it non-portable it's tied to your machine and the reason linux operating systems can get away with it is because they control the layout for everything you install so like when you do a DNF install whatever or apt install or whatever.

1:48:38They've already done all these builds with knowing where everything's going to end up. So it just works. People like Homebrew, though, that you actually install in different locations and stuff and you don't know what they're going to pull in, they can do that too, right? To an extent, but not really. They have to make it portable. Conda does this as well because a thing people did for a long time was they used Conda to install Python and then actually would use pip to then do installs for the packages. Because condo was the easy way to get Python on your machine. But UV is starting to eat into that, I think, personally.

1:49:11But the deal is, this stuff gets hard-coded and it's not portable across machines. So what Python Build Standalone did was they patched some stuff in CPython that made some assumptions about things being static about where they ended up. And then also have some scripts to patch or to set certain settings on the resulting binaries so that they also use relative paths of where to look and what to do. And let you just drop the CPython binary anywhere in your machine and it would just basically work. And that's the deal. It's just Python never was structured for that and we didn't worry about it because people just, it's a thing, but it was never a critical thing.

1:49:53And it slowly has become more of a critical thing because people have found that there's a really good workflow that's nice, right? Because this is one way to kind of like package Python in with your pre-existing application as like a scripting backend if you don't want to compile it directly into your binary itself. So that's what I like about when I say pre-built binaries is basically right now for python.org, right? You can get a Windows installer, although you should honestly install the Python manager from the Microsoft store if you're on Windows and then let that do the downloads for you.

1:50:22Adam, are you on Windows? No, dude. You can get the Mac installer, which is a graphical installer, and that will also install. And then on Linux, there's nothing. It's just source. You can get the tarpaul. So what I want to do is I want to get it so that for Windows, for Mac, for Linux, and for any other tier one, well, that's all the tier one platforms, but even hopefully tier two platforms that Python supports, because I created a tier system for our platform support for Wazzy. Tier two? No. We can talk about tiers after this. so to explain to Adam too far in the weeds come back a little bit I said go in the weeds but I didn't really mean it this is machete order he's machete in these weeds man I'm sure this has something to do with some home lab where Adam's running FreeBSD to run a ZFS server so that's why I'm assuming he's asking no I've moved on from that no Adam moves on very quickly that's true I'm in some good spaces man I've been playing with okay Incas latest and greatest man it's so much fun I mean the best.

1:51:27Zincas plus ZFS. The world, man. The world. So cool. Okay. I will look forward to the show notes to find out the link to the thing you just mentioned. I live vicariously through you, Adam. I don't have the time right now with near Tom. I'll lead you, Brett. I will get my machete out, and I will lead you through the weeds. Thank you. Anyway, so that's what I want to get it so that you can get basically as zip files, Harbaugh, whatever. For Windows, Mac, Linux hopefully Wazzy as well maybe one or two other platforms you can just download and just unpack and just have Python ready to go because you just don't have that right now and that's what I want beautiful world right so that would bypass the need for them to they would have to compile it right it's the same in that exactly it's just a download one how to do it 17 steps to learn how to do it because just to play with Python you have to learn how to compile Python and install Python and version Python and all the things Python like can we skip that step please it's more honestly it's more about making the install step universal.

1:52:26Right now, if you're on Linux, you can get it, but you have to know what command to use to get it. And to be frank, people like Debian mess with the Python install in ways that we don't, on the core team, fully agree with. But you also have to know what to get. Like on Debian, they don't install VEMV, which is one of the key steps I listed. They consider it an application, so they rip it out based on Debian policy. So you have to know to install Python 3-full to get pip and vmv installed and not just have Python without vmv and without pip, which are kind of critical. Yeah, see, this is too many steps.

1:52:58That's some thick weeds, man. That's too many steps. This is why Jared's been upset all these years and you're here to clear the air. I've got Python PTSD. Yeah. He's like, I resist it like the plague, Adam. I do. I'm just like, gosh. But once again, instructions on Mac are different. Instructions on Windows are different, right? Everyone's got a different way. There's no universal kind of just like, run this command regardless of OS or just get this one tool installed and then it'll automatically just get you Python. Right. And that's why people love you. Then you got to have a build pipeline, right?

1:53:25You got your GitHub actions going. You got your containers going. You got, right. Are you in that world? Are you in the CI crazy GitHub actions? Oh my, oh my gosh world. Well, I helped set it up for Python. So yeah. Well, then you're going to want to play with Incas then. Okay. We should talk. Well, I don't run it anymore. That is one thing I've handed off to the youngsters. Oh, okay. Well, you can guide. I'm the guy who got CPython's development onto GitHub. And thus, we actually used all, we predate GitHub Actions. But once I got us over there, I helped get us onto GitHub Actions and not rely on Travis and stuff.

1:53:59And CircleCI. TravisCI. That's a name I haven't heard in a long time. Just shut up right now. This show's over, y 'all. He said Travis. So you see how old this old stuff goes, right? And all the stuff we've had to move on because things don't last. Nice to meet us hard to talk about Jenkins. But now you get it, right? But this is why there's any hedging of concerns around companies. We outlasted people caring about CVS. We outlasted SourceForge. Python came out in February of 91. Resilient. Before Linux came out that August. We're an old project. We outlast a lot of stuff. And we're kind of not going anywhere.

1:54:37We're enterprise level. We're not COBOL level yet because we're just not that old. but effectively you can't get rid of Python at this point. We're everywhere. So it's one of those, we kind of have to plan for the future in terms of making sure we are to an extent self-sustaining. So you've explained it. You took us in the weeds. What is the state? Give me a brief version of the state of prebuilt binaries. Is this close? Is it shipped? Who's running it? Where can you go? So I'm running it. Okay. So if you go, so I have a pro, so there's a GitHub project, kind of coordinates. Go to github.com slash python slash, I believe it's pre-built dash cpython.

1:55:15But it's on the Python org, right? Yeah, it's on the Python org. Basically, it's all discussions really there of just like, here's the goal, and then talk with the stakeholders who care and just trying to work stuff up. I'm trying to work with Python Builds standalone to get their key patches upstreamed that they apply when they do builds. I said they're busy, so it's a bit slow, but they're working with me slowly through them. Once those are all up, then we're going to figure out exactly what needs to happen for a build to patch or tweak or whatever file they have to script to make the actual build happen.

1:55:51So it's no longer just tweaks to source, it's actually the steps to do the build. And then at that point, the release managers will just do it as part of their releases. It'll just happen. So I'm going to say it's being worked on. Whether or not it is a 2026 thing or a 2027 thing, I don't know. because I'm kind of depending on other people on being the experts and I'm just being acting as product manager in a way for this stuff to kind of make it happen. So I'm driving it, but I'm not, I have to ask other people for help so I can't just make it happen all on my own. So it's at the speed of me and other people being able to work together, get stuff done.

1:56:25But if you need some help with the actions part of it and the builds pipeline part of it, I got some wisdom. I can, if not all of it, some of it, because I've been tinkering on this front here and I'm like knee deep in solving a problem right now. I mean, we can talk when it comes time if you want to contribute to Python somehow. I would love to. But right now I'm at the stage of just trying to make it so that if you did a build on your own machine, it would work. Then we can talk about what it's going to look like on GitHub Actions because that's when we talk about like, okay, like what Linux distro do we use to make sure we get an old enough version of GLibC that it runs everywhere we want to run on and that kind of thing.

1:56:59It's probably going to be all in the Linux, I think, because that's already what's... What if you can build it yourself? container up man container up brother let's talk later we're going too far in the weeds I love it though my gosh it's the best getup.com slash python slash prebuilt dash c python that's the url right if you pulled it up that's it that's what I got I went to the python org I searched for pre found prebuilt dash c python that's it so it's in the getting patches upstreamed And then after that is getting the scripts so that it can be done as a release. And then it's getting it in as a release.

1:57:39I don't think it's going to happen for 3.15, but hopefully for 3.16. This needs to happen. This needs to happen. It's going to increase the use of Python. It's going to make onboarding newcomers, onboarding agents even so much easier. Yeah. Especially when an agent is taking a brand new machine and helping a new developer get to that next level. Like this is paramount. Like you need to put the gas, the pedal down, gas all the way, burn the fire, hatch it with the weeds, all the things. Okay, get this done tomorrow. If you need some security, I also got you a Wazzy build too. So you can totally have it use Wazzy so that it can't run amok.

1:58:13Right on. that's a whole nother topic that is i might have to bring you back for that one snarky dot ca y 'all okay this guy snarky.ca get it while you while it's hot it's a free usb on the it's a gift it's just sitting there it's a free usb stick on the uh on the curb good throwback pick it up pick it up and give it a little love hey give a little love this guy did it before you did it he's gonna keep doing it well you're long gone that's brett cannon brett is the best man python will last forever and brett cannon for life you're the best man anything left in closing here no just love to see you too thanks for having me back on it's always good catching up and shooting shooting stuff about movies and tv shows and all that good stuff movies and software man doesn't get much better tv shows movies software movies and software y 'all that's how that's how we do it around here brett you're a friend thank you so much we love you so much man stay cool stay awesome bye friends bye friends bye alright that is the changelog for this week thanks for listening so which Star Wars order is the best Star Wars order would you listen to AI David Attenborough do you have strong opinions on open source voting systems let us know in the comments links in the show notes we love hearing from you thanks again to fly.io for partnering with us and to Breakmaster Cylinder for producing a steady stream of the best beats in the biz next week on the changelog.

1:59:44News on Monday, Steve Ruiz from TL Draw on Wednesday, and changelogging friends on Friday. Have yourself a great weekend, set up a blog if you haven't already, and let's talk again real soon.

2:00:13Game on!