РАСШИФРОВКА 🔗
WEBVTT
00:00:00.100 --> 00:00:05.180
Hello and welcome to Python Bytes, where we deliver Python news and headlines directly to earbuds.
00:00:05.310 --> 00:00:11.820
This is episode 492, recorded August 18th, 2026. I'm Michael Kennedy.
00:00:12.170 --> 00:00:13.240
And I'm Calvin Hendryx-Parker.
00:00:13.520 --> 00:00:19.320
This episode is brought to you by Logfire. So thank you to Logfire from Pydantic.
00:00:19.820 --> 00:00:25.240
Check them out at pythonbytes.fm/logfire. It's about observability for your AI apps and
00:00:25.350 --> 00:00:29.320
beyond. Tell you more about them later. Connect with us on social, links on the website.
00:00:29.540 --> 00:00:30.700
Be part of the live show.
00:00:31.100 --> 00:00:33.300
Check it out on the website, pythonbytes.fm.
00:00:33.600 --> 00:00:34.040
It's live.
00:00:34.320 --> 00:00:39.000
And we have a really cool newsletter that we send out mostly every week with all sorts of extras.
00:00:39.360 --> 00:00:40.720
So you can sign up there as well.
00:00:40.900 --> 00:00:43.780
Just visit the website, click the newsletter, do the thing.
00:00:44.080 --> 00:00:46.680
We'll keep it private and not abuse it.
00:00:47.040 --> 00:00:48.140
Speaking of abuse.
00:00:48.600 --> 00:00:49.280
We're good people.
00:00:49.440 --> 00:00:49.920
You're good people.
00:00:50.220 --> 00:00:51.680
Let's make sure we keep people safe, Michael.
00:00:52.040 --> 00:00:52.440
Exactly.
00:00:52.800 --> 00:00:54.840
Sometimes you can have nice things on the internet.
00:00:55.180 --> 00:00:55.600
That's true.
00:00:55.820 --> 00:01:01.440
And well, the good thing is that with the solar eclipse in Europe last week, a new release came
00:01:01.460 --> 00:01:10.640
out of some security updates for Python 3.12, 3.11, and 3.10. The key bit here, well, there's a lot of
00:01:10.700 --> 00:01:15.380
key bits, which is there's lots of security updates that have happened. They're pretty important if
00:01:15.440 --> 00:01:21.079
you're doing certain kinds of activities. So for example, if you are using tar files to, if you're
00:01:21.020 --> 00:01:25.720
extracting tar files with any of these versions of Python, you are going to want to update ASAP
00:01:26.420 --> 00:01:33.000
because this basically fixed a path traversal bypass of the data filter. So a symlink escape
00:01:33.210 --> 00:01:38.420
that could bypass bits and pieces here and get people access to places they're not supposed to be.
00:01:39.020 --> 00:01:43.720
Kind of important. Now the important bit here is that this is only a source release,
00:01:44.280 --> 00:01:50.240
is a source release only. If you want to upgrade to these versions of Python, the 3.12.14, 3.11.16,
00:01:50.940 --> 00:01:57.640
and 3.10.21, you are going to have to download the source tarball or patch your existing build
00:01:57.960 --> 00:02:02.120
process to get these released into your production environments. But if you are on these versions,
00:02:02.200 --> 00:02:05.500
you're going to want to get those things patched. I mean, it's nice that these are in security only
00:02:05.560 --> 00:02:10.340
fixed mode. There's not a binary version. I have to use this. There is not a binary version of
00:02:10.360 --> 00:02:15.799
these been released. These are in security fix only mode. I mean, hopefully you're getting into
00:02:16.120 --> 00:02:22.220
Python 3.13 and 3.14 as they are the current releases. It's nice that the community supports
00:02:22.520 --> 00:02:26.980
these versions that far back. I don't know when 3.10 was released, but it was quite a while ago.
00:02:27.800 --> 00:02:32.440
And you've got to be able to stay on top of these. This is in your hands now. This is kind of more
00:02:32.500 --> 00:02:38.180
of a do-it-yourself, but at least the fixes are available. So tar file is a big one. There are
00:02:38.580 --> 00:02:45.019
four fresh CVEs that are being covered by these fixes here. One of them is for the sourceless
00:02:45.040 --> 00:02:52.360
file loader. The other one is the cookies morsel, which there's basically a couple of fixes there
00:02:52.360 --> 00:02:59.400
on that. And then the ftp lib ftp cp. So there's also a bunch of denial of service type fixes,
00:03:00.120 --> 00:03:07.520
some cleanups in HTML parser, config parser, Unicode data normalize, element tree XPath bits.
00:03:08.120 --> 00:03:11.660
There's some header injection fixes. Again, just lots of cleanups around the edges.
00:03:14.100 --> 00:03:18.260
Michael, have you ever used the web browser module from the standard library?
00:03:18.720 --> 00:03:20.900
I'm learning now there's a web browser module.
00:03:21.090 --> 00:03:22.700
There is a web browser module.
00:03:23.460 --> 00:03:24.580
It allows you to pop a web browser.
00:03:25.000 --> 00:03:25.500
That's pretty handy.
00:03:25.780 --> 00:03:27.920
I've used it for a couple like little demo-y type things,
00:03:28.180 --> 00:03:30.920
but there was a allowed control character in there
00:03:30.950 --> 00:03:34.060
that would have caused an action prefix bypass.
00:03:34.680 --> 00:03:37.260
So if folks were sneaky and knew those things existed,
00:03:37.330 --> 00:03:40.260
they could basically jump in and bypass some of these security controls.
00:03:40.860 --> 00:03:49.080
The HTTP client also now caps chunked trailer lines and the 100 responses to 100 each.
00:03:49.300 --> 00:03:53.620
So a hostile server could previously hang a client, even if you had a socket timeout in place.
00:03:54.040 --> 00:03:56.620
So lots of memory safety stuff, lots of other little cleanups.
00:03:56.740 --> 00:03:59.560
Again, it's a pretty sweeping set of security fixes.
00:03:59.840 --> 00:04:02.840
And they chose to release it now because it's as good a time as any.
00:04:03.160 --> 00:04:07.280
There's no set schedule for these specific releases on these older versions of Python.
00:04:08.060 --> 00:04:08.700
Excuse me.
00:04:09.060 --> 00:04:15.060
But it's good we've got people watching our
Load full transcript
backs and making sure we stay secure out there.
00:04:15.660 --> 00:04:17.579
Like I say, you can go look at the release notes.
00:04:17.880 --> 00:04:19.100
There's a ton of fixes in here.
00:04:19.299 --> 00:04:22.800
It's a really long set of releases and fixes that went into us.
00:04:23.480 --> 00:04:28.420
Thanks to all the Python core devs and volunteers and people submitting patches.
00:04:28.780 --> 00:04:34.280
I'm sure some of this stuff probably came out of the new models that are able to detect these kinds of pieces.
00:04:34.640 --> 00:04:37.720
But obviously, humans are putting the pieces into place.
00:04:38.220 --> 00:04:39.420
So go be safe out there, Michael.
00:04:40.880 --> 00:04:41.740
And upgrade your Pythons.
00:04:41.740 --> 00:04:42.320
This is always something I'm worried about.
00:04:42.600 --> 00:04:43.260
Yeah, I mean.
00:04:43.580 --> 00:04:44.560
Upgrade your Python, please.
00:04:45.300 --> 00:04:48.340
Here's also one of those reasons why I kind of want to stay on top.
00:04:48.640 --> 00:04:48.960
Yeah.
00:04:49.180 --> 00:04:51.080
Of having a fairly new version.
00:04:51.370 --> 00:04:52.840
And I'll be like, oh, we're not going to touch it.
00:04:53.160 --> 00:04:54.840
And hey, it's 3.8, but it works.
00:04:55.280 --> 00:04:59.640
It's like, yeah, but when this kind of jump, this stuff comes out of the blue,
00:05:00.040 --> 00:05:03.220
you're all of a sudden scrambling to get your app to run rather than just a point release.
00:05:03.720 --> 00:05:03.860
Yeah.
00:05:04.440 --> 00:05:10.000
It's hard to argue for the old stance of like, well, just because it works, it should just stay in production.
00:05:10.080 --> 00:05:20.860
Or even if it's behind a firewall or behind some kind of level of protections, we've seen that these models can now chain together multiple vulnerabilities to get access to spaces they weren't supposed to get access to.
00:05:21.260 --> 00:05:25.640
All that's required at this point is setting good goals for the agentic AI tools.
00:05:25.740 --> 00:05:33.520
And they can get around a lot of things if they're given unfettered access, even not unfettered access, even if they're sandboxed in, they could still do this.
00:05:33.920 --> 00:05:35.920
So keep up to date.
00:05:37.200 --> 00:05:41.420
Again, this is a reason for staying on the latest released version of Python.
00:05:42.180 --> 00:05:43.980
But the community does support back patches.
00:05:44.460 --> 00:05:45.520
You just got to go build them yourself.
00:05:45.980 --> 00:05:47.700
I'll have more on that later on the building.
00:05:48.380 --> 00:05:48.720
Maybe.
00:05:50.900 --> 00:05:52.520
So yes, but maybe.
00:05:52.960 --> 00:05:53.080
Maybe.
00:05:53.330 --> 00:05:53.700
What do you got?
00:05:53.780 --> 00:05:58.440
My favorite way to install Python is uv Python.
00:05:58.780 --> 00:05:59.120
No kidding.
00:05:59.300 --> 00:05:59.900
Install, right?
00:06:00.960 --> 00:06:03.220
And a quick, I just upgraded it.
00:06:03.300 --> 00:06:04.320
You got to upgrade to get the latest.
00:06:04.500 --> 00:06:05.200
It's just kind of odd.
00:06:05.200 --> 00:06:05.500
It doesn't.
00:06:05.540 --> 00:06:06.940
Oh, they've got someone's got built.
00:06:07.020 --> 00:06:07.300
It does.
00:06:08.240 --> 00:06:09.360
3.11.16.
00:06:10.920 --> 00:06:12.100
And 3.12.14.
00:06:12.510 --> 00:06:15.200
That looks like they've packaged it into Python build standalone.
00:06:15.700 --> 00:06:16.160
That's nice.
00:06:16.500 --> 00:06:19.060
Well, it's good to see the astral folks have got our back.
00:06:19.800 --> 00:06:20.380
Yes, obviously.
00:06:20.640 --> 00:06:21.920
Like that's a theme, right?
00:06:22.200 --> 00:06:22.960
It's definitely a theme.
00:06:23.250 --> 00:06:23.840
Well, there you go.
00:06:24.740 --> 00:06:24.960
Go.
00:06:25.260 --> 00:06:30.140
Well, if you're on this older version of Python, you might need to just go upgrade to uv at
00:06:30.140 --> 00:06:32.280
this point and save yourself some headache.
00:06:32.720 --> 00:06:32.900
Yeah.
00:06:33.180 --> 00:06:33.500
Yeah, absolutely.
00:06:34.230 --> 00:06:38.260
You could technically just use uv to manage your Python and still, I don't know.
00:06:38.280 --> 00:06:39.340
Yeah, you don't need it.
00:06:39.490 --> 00:06:42.460
You don't necessarily, yeah, you can do the rest of the way you're currently doing it.
00:06:42.680 --> 00:06:42.820
Right.
00:06:43.190 --> 00:06:46.320
But if you're going to jump in there, you might as well modernize and fix.
00:06:46.460 --> 00:06:47.120
You're already in there.
00:06:47.320 --> 00:06:48.180
You're already doing it.
00:06:48.600 --> 00:06:49.320
You're already in there.
00:06:50.000 --> 00:06:50.300
Okay.
00:06:51.100 --> 00:06:57.540
Speaking of, you gave a quick shout out to give an AI agent some kind of go find bugs.
00:06:57.800 --> 00:06:59.140
And I 100% agree.
00:06:59.150 --> 00:07:02.700
You and I both likened to this to running a lender for the first time.
00:07:03.040 --> 00:07:04.060
And you're like, there's a thousand errors.
00:07:04.130 --> 00:07:05.100
Oh my God, it's so much.
00:07:05.620 --> 00:07:07.060
Right, so stay on top of these things.
00:07:07.280 --> 00:07:11.520
But these tools are causing some places a ripple,
00:07:12.570 --> 00:07:14.580
other places a tidal wave.
00:07:15.200 --> 00:07:17.400
And Codeberg is one of those places that,
00:07:17.770 --> 00:07:20.520
I don't know, like, I'm not sure how to feel about Codeberg.
00:07:20.750 --> 00:07:23.340
I, in principle, like it,
00:07:23.720 --> 00:07:26.380
but I think people want Codeberg to be more
00:07:26.480 --> 00:07:28.380
than Codeberg wants to be, let's say that.
00:07:28.660 --> 00:07:29.660
What the heck is Codeberg?
00:07:29.780 --> 00:07:30.720
Codeberg isn't
00:07:30.720 --> 00:07:31.280
What is Codeberg?
00:07:32.080 --> 00:07:35.000
It is a GitHub alternative, let's say.
00:07:35.250 --> 00:07:35.380
Okay.
00:07:35.760 --> 00:07:38.560
Evolved out of
previous attempts to create a GitHub alternative.
00:07:38.830 --> 00:07:46.060
It's a nonprofit community-led effort with a democratic process that provides services to free and open source projects,
00:07:46.550 --> 00:07:50.340
such as Git hosting using 4Gio, Pages, CI/CD, and so on.
00:07:50.560 --> 00:07:54.940
So it's already got this restriction like, hey, it's only for free and open source stuff.
00:07:55.160 --> 00:08:00.360
It's not for just general paid software, private repos, and so on, right?
00:08:00.640 --> 00:08:00.820
Fine.
00:08:01.060 --> 00:08:04.100
So that's kind of its mission, and that's what it is.
00:08:04.330 --> 00:08:06.520
But a lot of people are looking at GitHub going,
00:08:06.920 --> 00:08:08.640
I mean, just yesterday GitHub was down.
00:08:08.740 --> 00:08:12.240
I'm like, oh, man, I guess I'll try this PR again in a half hour
00:08:12.920 --> 00:08:14.260
because it couldn't load the PR stuff,
00:08:14.370 --> 00:08:15.600
and then I just got the unicorn,
00:08:16.160 --> 00:08:17.620
which kind of feels like it's supposed to be amazing,
00:08:17.750 --> 00:08:18.640
but it's a bad unicorn.
00:08:19.760 --> 00:08:21.940
Should have like a leather jacket and be smoking or something.
00:08:21.990 --> 00:08:22.260
I don't know.
00:08:22.280 --> 00:08:24.500
I know, kind of the fail whale of GitHub.
00:08:25.180 --> 00:08:25.660
Yes, exactly.
00:08:25.910 --> 00:08:28.380
And there might be people out there that don't know the fail whale,
00:08:28.540 --> 00:08:31.240
but oh my gosh, that Twitter went down bad in the beginning.
00:08:31.800 --> 00:08:32.800
Yeah, so it's like the fail wheel.
00:08:32.900 --> 00:08:33.340
Early days.
00:08:33.479 --> 00:08:35.120
But right now, I just pulled this up
00:08:35.140 --> 00:08:37.500
into the top of the Codeberg page.
00:08:38.200 --> 00:08:38.780
Time recording.
00:08:39.039 --> 00:08:40.580
Yeah, the number two is interesting.
00:08:42.219 --> 00:08:43.680
There's just like, there's stuff we don't want here.
00:08:44.219 --> 00:08:45.140
And we don't approve of it,
00:08:45.260 --> 00:08:46.380
and it needs to get the heck out.
00:08:46.520 --> 00:08:49.160
Even if it is free and open source,
00:08:49.380 --> 00:08:50.960
it still needs to get the heck out, okay?
00:08:51.300 --> 00:08:53.000
So I feel like this was just,
00:08:53.240 --> 00:08:54.480
there's stuff we don't like, let's get it out.
00:08:55.040 --> 00:08:56.859
And the number two you pointed out and laughed at
00:08:56.880 --> 00:08:59.540
is like cryptocurrency projects are no longer allowed.
00:08:59.600 --> 00:09:00.700
And like, you know, I'm fine with that.
00:09:00.860 --> 00:09:01.660
Like, whatever.
00:09:01.920 --> 00:09:02.560
Okay, okay.
00:09:03.340 --> 00:09:05.400
If that's your mission to just not have that kind of stuff,
00:09:05.840 --> 00:09:06.520
I'm here for it.
00:09:06.560 --> 00:09:07.780
Like cryptocurrency is weird.
00:09:08.120 --> 00:09:09.960
I'm not 100% against it, but it's just
00:09:09.960 --> 00:09:11.280
I'm not either.
00:09:11.280 --> 00:09:12.680
So that's why I thought it was kind of weird.
00:09:12.680 --> 00:09:14.520
There's definitely some opinions here
00:09:14.520 --> 00:09:16.720
that had a thought process going on behind them
00:09:16.720 --> 00:09:18.220
that we're not seeing the full story.
00:09:18.220 --> 00:09:20.820
I guess we can go view the full change, but
00:09:20.820 --> 00:09:23.140
Yeah, but this first one is quite weird.
00:09:23.220 --> 00:09:26.180
It says LLM generated content is being restricted.
00:09:26.860 --> 00:09:33.860
More specifically, if you have a project that they somehow decide they believe it is majority LLM, banned.
00:09:34.240 --> 00:09:41.760
Even if it has been here for a long time, it has gotten a lot of attention and a lot of maybe even wrote it by hand to start.
00:09:41.810 --> 00:09:45.000
But then you do some mega refactoring that touches every file.
00:09:45.220 --> 00:09:45.420
Banned.
00:09:46.440 --> 00:09:47.840
So you may like that.
00:09:47.850 --> 00:09:48.480
You may not.
00:09:48.800 --> 00:09:49.320
It's extreme.
00:09:49.340 --> 00:09:55.340
It seems a little extreme given that we just recently covered Linus Torvalds saying like,
00:09:55.500 --> 00:09:58.700
look, using AI to review our stuff is a tool.
00:09:59.100 --> 00:10:01.280
It feels a bit myopic to me.
00:10:01.520 --> 00:10:03.280
And fine, like I said, this is their project.
00:10:03.620 --> 00:10:05.260
They can do it, but people have thoughts.
00:10:06.160 --> 00:10:07.280
People have a lot of thoughts.
00:10:07.860 --> 00:10:08.820
Wait, Armin has a thought?
00:10:09.580 --> 00:10:10.120
Could you imagine?
00:10:10.320 --> 00:10:11.300
Armin Roniker has a thought.
00:10:11.820 --> 00:10:15.420
And he wrote, and the reason a lot of people have thoughts is they were like,
00:10:15.560 --> 00:10:19.160
I would just love a GitHub alternative, like for different reasons.
00:10:19.300 --> 00:10:21.580
Some people are just like, I don't like the way GitHub works.
00:10:22.000 --> 00:10:22.940
It's been down a lot.
00:10:23.180 --> 00:10:25.480
Other people are like, I want a European company,
00:10:25.800 --> 00:10:30.380
not just an American company to kind of be a first-class player in this space.
00:10:30.500 --> 00:10:30.620
Right.
00:10:31.020 --> 00:10:33.740
And whatever, like all of those people are like,
00:10:34.100 --> 00:10:35.320
but Codeberg is now broken.
00:10:35.520 --> 00:10:36.100
So that's out.
00:10:36.840 --> 00:10:37.940
And that was my hope, right?
00:10:38.040 --> 00:10:40.260
Like Obi-Wan, you're my only hope.
00:10:40.580 --> 00:10:42.420
Oh, Obi-Wan doesn't like all of them content.
00:10:42.520 --> 00:10:43.320
Well, I guess he's out.
00:10:44.140 --> 00:10:48.220
Anyway, so Armin wrote an article called Codeberg Divides.
00:10:48.680 --> 00:10:51.140
And so they changed their terms of service, as I said.
00:10:51.600 --> 00:10:53.220
They're entirely
within the rights to do this.
00:10:53.640 --> 00:10:54.780
It's a democratic process.
00:10:55.300 --> 00:10:58.980
But democracies don't inherently make good decisions.
00:10:59.740 --> 00:11:00.560
That's what I'm saying, right?
00:11:00.940 --> 00:11:05.160
So GitHub's never been democratic, and there's plenty not to like.
00:11:05.260 --> 00:11:09.080
But as a core piece of infrastructure, it's been reliable, right?
00:11:09.480 --> 00:11:10.520
So I don't know.
00:11:11.090 --> 00:11:13.020
How do you even enforce this majority?
00:11:13.640 --> 00:11:36.340
Like I said, if I have a project, let's say a web app that uses some CSS front end, and I ask AI to change the CSS front end around, even though I wrote all the code, all the back end, in the diff perspective, chances are it's going to overwhelm the back end amount of code if I have to have it rewrite every bit of HTML because it's grid versus flex box versus whatever.
00:11:36.340 --> 00:11:42.580
This is groundwork stuff that the AIs can be directed very easily to take care of for us as humans.
00:11:43.140 --> 00:11:47.340
Yeah, and so here's sort of the thoughts.
00:11:47.600 --> 00:11:49.820
It's a real shame that open source and free software communities
00:11:49.850 --> 00:11:53.180
are splitting this deeply over LLM and agents.
00:11:55.300 --> 00:12:00.020
We all already acknowledge copyright issues, energy use,
00:12:00.620 --> 00:12:01.240
those kinds of things.
00:12:01.640 --> 00:12:03.840
But it's also just how software is being made, right?
00:12:03.910 --> 00:12:06.560
And people need to, projects are going to need to figure out
00:12:06.560 --> 00:12:07.220
how to coexist.
00:12:08.560 --> 00:12:12.160
And yeah, if you want GitHub to face true competition,
00:12:12.760 --> 00:12:13.420
This is not it.
00:12:13.540 --> 00:12:14.860
Yeah, this is not going to work.
00:12:15.279 --> 00:12:16.340
Michael, I got an idea.
00:12:16.900 --> 00:12:17.100
Okay.
00:12:17.260 --> 00:12:21.200
We're going to do a blockchain project to host a GitHub alternative
00:12:21.760 --> 00:12:24.220
that is a distributed autonomous organization.
00:12:24.760 --> 00:12:27.000
And it's going to have like smart contracts and rules.
00:12:27.300 --> 00:12:28.540
We're going completely the other direction.
00:12:29.720 --> 00:12:32.280
You're making my ears hurt just a little bit with that word.
00:12:34.279 --> 00:12:35.520
So what about this?
00:12:35.540 --> 00:12:39.040
What if we go back a little further and we can like do a sweet arc?
00:12:39.280 --> 00:12:42.060
Like what if we do like a BitTorrent hosted,
00:12:42.920 --> 00:12:44.220
smart contract backed?
00:12:45.000 --> 00:12:46.660
Even more, like the layers go deep.
00:12:47.780 --> 00:12:50.620
So, you know, the decentralization aspect
00:12:50.870 --> 00:12:52.740
of BitTorrent totally makes sense here.
00:12:52.790 --> 00:12:53.880
I think we could make this go.
00:12:54.140 --> 00:12:55.540
No, we could totally make it go.
00:12:56.080 --> 00:12:58.160
All right, so that was Armand's thought.
00:12:58.160 --> 00:12:59.480
I wish this was more forward looking
00:12:59.650 --> 00:13:00.940
in a place that we could depend on
00:13:01.210 --> 00:13:02.260
rather than a place that divides
00:13:02.390 --> 00:13:04.220
and just kicks a bunch of people up.
00:13:04.480 --> 00:13:07.240
Now, one of the reasons I'm sure they did this is,
00:13:07.640 --> 00:13:10.160
well, these AI things use 20 times as much resources.
00:13:10.580 --> 00:13:10.760
Yeah.
00:13:11.000 --> 00:13:11.100
Yeah.
00:13:11.450 --> 00:13:11.560
Okay.
00:13:11.690 --> 00:13:16.600
So I just searched for the terms to see kind of Codeberg terms,
00:13:16.710 --> 00:13:17.840
and I came up with this website.
00:13:19.580 --> 00:13:22.780
The domain is amazing, but it says...
00:13:22.860 --> 00:13:23.140
Oh, wow.
00:13:23.660 --> 00:13:26.300
I asked literally just for the terms of service of Codeberg,
00:13:26.520 --> 00:13:27.940
and the first thing that comes up is,
00:13:28.060 --> 00:13:30.040
I regret migrating to Codeberg.
00:13:30.740 --> 00:13:34.220
And this is actually a really interesting and well-thought-out article.
00:13:34.660 --> 00:13:37.019
So this person just recently left GitHub
00:13:37.040 --> 00:13:39.120
and moved all their projects over here because they're like,
00:13:39.240 --> 00:13:41.380
ah, GitHub is kind of getting clunky,
00:13:41.800 --> 00:13:44.600
and I have just pure open source and so on.
00:13:45.000 --> 00:13:47.160
So GitHub looked like a viable alternative,
00:13:47.310 --> 00:13:50.400
which is like this is the main spine running through this conversation.
00:13:50.700 --> 00:13:56.420
So unfortunately, the latest update to its terms of service
00:13:56.640 --> 00:14:00.200
seemed to mark the first step in changing one part I moved there for,
00:14:00.600 --> 00:14:02.140
namely freedom, is now gone.
00:14:02.330 --> 00:14:03.180
I can't work on it.
00:14:03.540 --> 00:14:05.100
And one of the interesting things about this term,
00:14:05.300 --> 00:14:07.820
I'll move on from this section in just a second.
00:14:08.060 --> 00:14:10.300
But from their terms,
00:14:10.420 --> 00:14:12.880
is the content that harms the reputation of CodeBerg,
00:14:12.980 --> 00:14:14.680
which is like, we don't have a great reason for it.
00:14:15.060 --> 00:14:16.840
And one of the criteria they use
00:14:16.840 --> 00:14:19.040
is it has to have a community behind it
00:14:19.340 --> 00:14:20.980
in order for it to be legit.
00:14:21.120 --> 00:14:24.260
And LLMs and AI don't make community, right?
00:14:24.720 --> 00:14:26.340
It seems like from the blog post,
00:14:26.420 --> 00:14:28.180
it seems like many vibe coders don't realize
00:14:28.480 --> 00:14:31.140
that they don't actually have a community around them.
00:14:31.300 --> 00:14:32.800
And I think that that cuts at two things.
00:14:33.060 --> 00:14:40.020
One, casting all of this stuff as vibe coders and just
junk users who don't really know what they're doing.
00:14:40.680 --> 00:14:42.060
That's already kind of a...
00:14:42.080 --> 00:14:46.120
There's a valid practice of agentic engineering in producing software.
00:14:46.580 --> 00:14:46.740
Yes.
00:14:46.850 --> 00:14:54.100
And this seems to lump it all in with like, oh, it's a bunch of these idiots that don't know what they're doing and we want them off our community.
00:14:55.320 --> 00:15:00.660
Sounds like a bunch of people who were against the stuff two years ago and have never tried it.
00:15:01.360 --> 00:15:03.380
don't realize the effect it has.
00:15:03.400 --> 00:15:04.720
But so just wrapping this up,
00:15:04.840 --> 00:15:05.480
this person says, look,
00:15:05.860 --> 00:15:08.860
I have a bunch of solo developer
00:15:09.640 --> 00:15:10.340
open source projects.
00:15:10.740 --> 00:15:11.800
And my community is people
00:15:11.800 --> 00:15:13.620
who just ask for features
00:15:14.160 --> 00:15:15.440
or ask me to change something.
00:15:16.060 --> 00:15:16.920
They're just people
00:15:16.920 --> 00:15:18.300
who want something from the software.
00:15:18.500 --> 00:15:19.720
They're not really a community
00:15:19.980 --> 00:15:21.700
with rich discussions, right?
00:15:21.940 --> 00:15:22.800
And so
00:15:22.800 --> 00:15:24.560
I agree with this person's take on,
00:15:24.640 --> 00:15:26.740
like, it's a bunch of little tools
00:15:26.740 --> 00:15:27.920
that are my tools
00:15:27.920 --> 00:15:29.280
and I need a good place to host them.
00:15:29.280 --> 00:15:30.319
And I want to share them
00:15:30.540 --> 00:15:31.440
with my little community.
00:15:31.860 --> 00:15:32.960
Yeah, like I've been publishing
00:15:33.580 --> 00:15:34.560
silly little tool for decades.
00:15:35.100 --> 00:15:37.860
And that was from even the SourceForge days, right?
00:15:38.420 --> 00:15:40.380
This is the essence of like open source,
00:15:40.540 --> 00:15:41.600
the cathedral and the bazaar,
00:15:41.740 --> 00:15:43.340
like that whole bit.
00:15:43.660 --> 00:15:44.060
Exactly.
00:15:44.420 --> 00:15:47.460
And so one of the reasons CodeBurg is blocking LLMs
00:15:47.600 --> 00:15:49.700
is because they think there can't be a community
00:15:49.840 --> 00:15:50.920
and therefore it's a fake project.
00:15:51.060 --> 00:15:52.780
This guy's like, I got a bunch of projects
00:15:52.880 --> 00:15:53.520
that don't have community,
00:15:53.580 --> 00:15:54.400
but they're important to me.
00:15:54.940 --> 00:15:56.540
And I think that's actually the majority.
00:15:57.780 --> 00:15:59.220
This results in censorship for me.
00:15:59.260 --> 00:16:08.540
But the same thing is, what if you just charged for the ones or had a very high or very strong rate limit for interactions?
00:16:09.240 --> 00:16:17.560
And if your AI does too much to burden our infrastructure, well, then, you know, maybe you cover that or you move them over to a portion of infrastructure that's even slower.
00:16:17.900
--> 00:16:18.460
Right. I don't know.
00:16:19.160 --> 00:16:22.180
But there's a whole lot of concrete ideas from this.
00:16:22.570 --> 00:16:23.860
I regret moving to Kudberg.
00:16:24.940 --> 00:16:25.140
Anyway.
00:16:25.200 --> 00:16:30.380
It's amazing that was the first post, like first result for a valid search for their terms of service.
00:16:31.460 --> 00:16:35.340
I know, that's a strong own goal right there.
00:16:35.360 --> 00:16:36.480
Speaks volumes right there.
00:16:37.120 --> 00:16:38.660
Well, the others are going to emerge.
00:16:38.940 --> 00:16:48.020
Obviously, the fact that we still live in an open world with a fairly open internet, people can build their distributed blockchain DAO version of GitHub if they want.
00:16:48.240 --> 00:16:50.100
Well, I'm looking forward to setting up.
00:16:51.000 --> 00:16:52.140
We should build on Codeberg.
00:16:54.780 --> 00:16:55.500
Oh my gosh.
00:16:55.890 --> 00:16:56.020
Okay.
00:16:56.520 --> 00:16:56.920
All right, sorry.
00:16:58.020 --> 00:17:00.300
That was a bit of a detour.
00:17:01.800 --> 00:17:02.860
A fun detour.
00:17:03.160 --> 00:17:06.780
But let me tell you about our sponsor, actually.
00:17:07.079 --> 00:17:07.439
You should.
00:17:07.709 --> 00:17:08.480
I think this is important.
00:17:09.300 --> 00:17:11.199
Yeah, because there's some really cool stuff
00:17:11.620 --> 00:17:14.439
that Logfire from Pydantic is doing.
00:17:15.890 --> 00:17:19.240
And this is a bit of an alternative worldview,
00:17:19.819 --> 00:17:22.020
one that I definitely agree with.
00:17:22.220 --> 00:17:33.940
But Logfire is about monitoring and observing and understanding what your AI agents and other web apps like FastAPI and Postgres and so on are up to.
00:17:34.060 --> 00:17:35.640
So it's 2 a.m.
00:17:35.950 --> 00:17:37.600
Your AI agent has failed.
00:17:38.220 --> 00:17:38.480
Why?
00:17:38.800 --> 00:17:40.620
Was it the model, a tool called the database?
00:17:41.340 --> 00:17:46.500
Most observability tools can't tell you because they only see part of your stack, like your web stack or something.
00:17:46.860 --> 00:17:48.500
Pydantic Logfire sees all of it.
00:17:48.760 --> 00:17:51.860
One trace across your agents, LLMs, APIs, and database.
00:17:52.480 --> 00:17:55.260
Down to the infrastructure, services, Kubernetes, and hosts.
00:17:55.720 --> 00:17:59.720
It's built on open telemetry with SDKs for Python, TypeScript, and Rust.
00:18:00.230 --> 00:18:02.880
And it works with any OTEL-compatible language.
00:18:03.340 --> 00:18:07.960
Every prompt, token count, cost right next to your vector searches and API calls.
00:18:08.320 --> 00:18:12.680
You query everything with Postgres-compatible SQL.
00:18:13.320 --> 00:18:18.100
And you give that to your coding agent through the Logfire MCP.
00:18:18.420 --> 00:18:21.280
So stop guessing, read the trace, Identic Logfire.
00:18:21.600 --> 00:18:24.000
AI is still just engineering, as we said.
00:18:24.460 --> 00:18:28.640
So
visit pythonbytes.fm/Logfire today and sign up.
00:18:28.660 --> 00:18:31.780
You'll get 10 million records for free every month, no credit card expired.
00:18:32.700 --> 00:18:34.740
And this is a pretty cool thing as well.
00:18:35.100 --> 00:18:38.820
I need to add Logfire to my app so that I get this cool feature.
00:18:39.000 --> 00:18:39.480
How do you do it?
00:18:39.600 --> 00:18:47.320
They have a button that'll copy some markdown text to your clipboard and you click it and it says onboard your coding agent.
00:18:48.000 --> 00:18:54.760
So it just gives you instructions you can hand to Claude or Codex specifically what it needs to do to integrate your app with Logfire.
00:18:54.940 --> 00:18:55.920
So I think that's super cool.
00:18:56.200 --> 00:19:00.680
So thank you to Pydantic and Logfire for supporting the show.
00:19:00.980 --> 00:19:07.120
Actually, I love the idea of the observability of grabbing the prompts and the costs and the token usage.
00:19:07.540 --> 00:19:10.780
more folks are going to be heading down this route because it's been very hard to
00:19:11.220 --> 00:19:14.180
analyze the return on investment of using these tools.
00:19:14.420 --> 00:19:19.120
And as the costs potentially rise and subscriptions maybe shift and go away,
00:19:19.540 --> 00:19:21.740
you're going to want to know this kind of data to make good decisions.
00:19:22.020 --> 00:19:23.320
We're doing this right now at Six Feet Up.
00:19:23.540 --> 00:19:24.200
Are you? Awesome.
00:19:24.560 --> 00:19:27.000
Yeah, we're having these kinds of discussions, I should say.
00:19:27.020 --> 00:19:28.380
And I could probably use a tool like this.
00:19:29.000 --> 00:19:32.600
Well, I don't buy Steadafim slash Longfire.
00:19:32.760 --> 00:19:34.040
Oh, you know, there's a code there.
00:19:34.640 --> 00:19:35.340
That's a good one for us.
00:19:35.580 --> 00:19:36.140
Yeah, fantastic.
00:19:36.620 --> 00:19:36.820
All right.
00:19:36.940 --> 00:19:37.920
Yeah, it's a super cool platform.
00:19:38.140 --> 00:19:43.240
And like I said, right here, just onboard with your code agent, click that, and then paste it.
00:19:43.640 --> 00:19:49.880
What's really interesting is that the instructions say things like, your user may have just clicked this and not actually read it yet.
00:19:50.080 --> 00:19:52.860
So please work them through this.
00:19:53.500 --> 00:19:54.560
Those are some good people.
00:19:56.000 --> 00:19:57.140
They were doing the good work right there.
00:19:57.440 --> 00:19:59.960
Yeah, same thing, a call in a team knows what they're doing over there.
00:20:00.300 --> 00:20:00.680
I love it.
00:20:01.260 --> 00:20:01.760
So does Brett.
00:20:02.320 --> 00:20:03.460
Brett Cannon knows a few things as well.
00:20:03.460 --> 00:20:04.420
Brett Cannon does know a few things.
00:20:04.560 --> 00:20:05.360
Actually, this is important.
00:20:06.260 --> 00:20:11.860
And again, I feel like the last few episodes, I've been on a bit of a security run with things.
00:20:12.250 --> 00:20:13.320
And this is continuing it.
00:20:13.330 --> 00:20:14.460
And he's got some great ideas.
00:20:14.850 --> 00:20:23.360
When he was writing his nomination post for the 2026 Python Packaging Council nomination or elections,
00:20:23.960 --> 00:20:33.380
part of his pitch why he should be there is we need a secure supply chain for the Python ecosystems around building reproducible builds.
00:20:33.900 --> 00:20:40.580
And so right now, we don't have technically reproducible builds on PyPI, but they're not
00:20:40.760 --> 00:20:41.120
far away.
00:20:41.580 --> 00:20:48.080
There's like he identified three gaps in the current process to get us to a point where
00:20:48.110 --> 00:20:51.820
we have basically the good design goal for him is zero friction.
00:20:52.520 --> 00:20:54.360
Can producers upload something to PyPI?
00:20:54.780 --> 00:20:56.000
They shouldn't have to do anything.
00:20:56.550 --> 00:21:02.960
And all the work lands on the build backends and installers to track and trace and allow
00:21:02.980 --> 00:21:08.260
you to know where your software came from that you're installing. We all are aware of the
00:21:09.700 --> 00:21:15.480
LiteLLM exploit that happened back in March. And part of that was because of a supply chain
00:21:16.160 --> 00:21:22.000
injections into the build process. The other one here is like he mentioned SolarWinds,
00:21:22.240 --> 00:21:27.240
that SolarWinds attack from a few years back. Again, a supply chain injection during the build process.
00:21:27.680 --> 00:21:32.320
So gap number one is nothing right now records the source code, where the source code distribution
00:21:32.340 --> 00:21:37.600
came from so if you add in a json descriptor it can capture you know when you install it where
00:21:37.640 --> 00:21:44.160
it's from what get repo is the canonical source of truth for that code and it's something that'd be
00:21:44.170 --> 00:21:48.200
very important for you to know if you're going to go back through and do an audit of the tools
00:21:48.340 --> 00:21:52.580
or the code and dependencies that you have into your build process interesting so you might
00:21:52.590 --> 00:21:58.720
be able to have a another tier of trust than just it's on pypi but yeah yeah here's how it's built
00:21:58.740 --> 00:22:04.440
Yeah, exactly. And if you combine those with the PyPI two-factor and security measures that they've
00:22:04.520 --> 00:22:08.120
been putting in place, this gives you a little more comfort with where something came from.
00:22:08.860 --> 00:22:13.320
Gap number two is during the build, in the build tools and during the build process,
00:22:14.220 --> 00:22:21.420
having a software bill of materials. So referred to as an SBOM. We can do this with PEP 770
00:22:22.200 --> 00:22:27.120
has a convention for this. Software distributions can't because there are tarball plus like some
00:22:27.140 --> 00:22:32.180
package info stuff. So if you have a wheel,
you can basically enable this already. So the first
00:22:32.180 --> 00:22:37.620
two things, very little friction. A file that describes where your GitHub repository is. The
00:22:37.720 --> 00:22:44.540
second one is using wheels and PEP 770 to have software build materials in there. So either don't
00:22:44.540 --> 00:22:54.100
use SDIS or move to the SDIS V2 support for this. And then basically the mechanisms already exist
00:22:54.120 --> 00:22:57.740
for tying this all together in the pyproject.toml.
00:22:58.000 --> 00:22:59.440
There's a build system section.
00:23:00.140 --> 00:23:01.620
If it has a defined entry point,
00:23:01.830 --> 00:23:03.800
so then backends can record their own environment
00:23:04.200 --> 00:23:06.260
and you can reinstall and rerun the build.
00:23:06.350 --> 00:23:09.080
So the goal being, again, if I build the software today
00:23:09.510 --> 00:23:11.480
and if I build the software six months from now
00:23:11.760 --> 00:23:12.740
or 18 months from now,
00:23:13.010 --> 00:23:15.400
I should get the exact same binary each time
00:23:15.880 --> 00:23:17.100
and there should be no worry
00:23:17.220 --> 00:23:19.620
or at least a way to audit and trace back what happens.
00:23:20.060 --> 00:23:21.900
So now trusted third parties can report
00:23:21.920 --> 00:23:27.600
successful reproductions back to PyPI, which can display like an independently reproduced by
00:23:28.320 --> 00:23:32.820
some authority surfaced in the index. So installers can also prefer reproduced files.
00:23:33.360 --> 00:23:37.820
So PyPI would have to have a couple of changes to support surfacing this data and making
00:23:38.080 --> 00:23:42.920
decisions based on that data. This is not a nice to have. This is definitely, it's not a requirement.
00:23:43.120 --> 00:23:48.480
This is basically like you need to have this as part of your build process or else you're not
00:23:48.500 --> 00:23:54.500
considered a top level or a first class citizen in PyPI because it shouldn't be hard to do.
00:23:55.280 --> 00:23:58.860
One of the things about the reproducibility that you hinched on right there at the end,
00:23:58.860 --> 00:24:04.300
I think is cool. I look at, say, the hash or whatever of the binary that I got. And then
00:24:04.700 --> 00:24:09.480
if I want to rebuild that myself, I should get the same hash if I follow the steps or something
00:24:09.640 --> 00:24:14.120
like that, right? Which is super cool. Yeah. But so many things can be very variable. I mean,
00:24:14.280 --> 00:24:17.200
the GCC compiler on your system,
00:24:17.920 --> 00:24:19.360
the minor versions of Python,
00:24:19.860 --> 00:24:23.140
the architecture, whether it was ARM or x86.
00:24:23.560 --> 00:24:25.600
There's a whole bunch of variables that go on in there.
00:24:26.020 --> 00:24:28.020
It'd be nice to be able to get those from a trusted source.
00:24:28.300 --> 00:24:31.600
And PyPI is obviously doing great work
00:24:31.780 --> 00:24:34.620
at making sure we get safe wheels
00:24:35.060 --> 00:24:36.220
and distributions of software.
00:24:36.560 --> 00:24:39.060
Yeah, definitely some great folks,
00:24:39.660 --> 00:24:40.780
including Seth and Mike,
00:24:41.480 --> 00:24:42.600
over there working on this.
00:24:42.740 --> 00:24:47.100
Well, and thanks to Seth Larson for listening to his idea and helping him check over this blog post.
00:24:47.200 --> 00:24:49.160
So the right folks are involved all around.
00:24:50.010 --> 00:24:53.860
That Python Packaging Council nomination, I think the voting is happening soon.
00:24:53.860 --> 00:25:00.440
I didn't actually note down when it is, but I can't imagine why we wouldn't want Brett to be on that panel of people.
00:25:00.860 --> 00:25:01.240
Exactly.
00:25:01.780 --> 00:25:02.560
He's got the right ideas.
00:25:02.820 --> 00:25:03.360
I'm all for this.
00:25:04.260 --> 00:25:11.600
Just because to keep the world safe, we need to have these kinds of table stick ideas built into our infrastructure of Python
00:25:11.620 --> 00:25:12.840
and the build systems.
00:25:13.310 --> 00:25:14.700
Yeah, and Brett's been working in packaging
00:25:15.800 --> 00:25:18.340
and package adjacent stuff for a long time recently.
00:25:18.550 --> 00:25:19.920
Yeah, so go enable these features.
00:25:20.720 --> 00:25:22.760
You'll start participating in the ecosystem
00:25:23.480 --> 00:25:25.140
and then eventually PyPI will support it
00:25:25.480 --> 00:25:27.640
and you'll be able to know that the software you're getting
00:25:28.160 --> 00:25:29.900
has come from verified sources.
00:25:30.620 --> 00:25:32.460
Maybe that's an extreme, not verified sources,
00:25:32.550 --> 00:25:34.400
but at least the build can be reproduced.
00:25:35.240 --> 00:25:35.680
Yeah.
00:25:36.060 --> 00:25:36.480
Verifiable.
00:25:38.020 --> 00:25:39.440
Verifiable, not verified.
00:25:40.580 --> 00:25:41.100
Subtle difference.
00:25:41.720 --> 00:25:50.840
yeah okay so now moving moving on we have a extra extra extra here all about it a whole
00:25:51.200 --> 00:25:56.540
section of extras well this is when there are so many extras that i'm like this can no longer be
00:25:56.720 --> 00:26:04.560
extra it must be hoisted up into a top level topic all right and i don't know i've got a sidebar full
00:26:04.700 --> 00:26:08.179
of them so let's see i got questions about one of these here for sure all right let's go well
00:26:08.340 --> 00:26:13.960
You talked about the security fixes of like 3.11 and so on, 3.11.6 and beyond.
00:26:14.500 --> 00:26:15.540
We also got a new...
00:26:15.540 --> 00:26:15.800
They are.
00:26:16.180 --> 00:26:18.180
Some of them have landed in 3.14.
00:26:18.620 --> 00:26:18.800
Yeah.
00:26:18.960 --> 00:26:21.220
So 3.14.7 is out.
00:26:21.460 --> 00:26:25.960
And sometimes these are like, we've changed some minor setting or whatever.
00:26:26.220 --> 00:26:28.480
This is a proper release of Python.
00:26:28.960 --> 00:26:34.220
And I know
it's mid-August and in a month and a half, we're going to have 3.15, which is
00:26:34.260 --> 00:26:34.820
very exciting.
00:26:35.200 --> 00:26:37.500
But there's a couple interesting takeaways here.
00:26:37.780 --> 00:26:39.860
Plus, you should just install this.
00:26:39.970 --> 00:26:42.860
So under security, it doesn't fit on my screen,
00:26:43.290 --> 00:26:45.720
the number of single sentence things fixed.
00:26:46.140 --> 00:26:48.360
So that alone tells you maybe we should pay attention.
00:26:48.580 --> 00:26:50.160
So like here, for example, the one you talked about,
00:26:50.340 --> 00:26:53.160
GitHub 15987, the tarball extract.
00:26:53.470 --> 00:26:55.000
And then there's another one somewhere in here,
00:26:55.110 --> 00:26:58.740
like this is the workaround of that fix, also fixed.
00:27:00.920 --> 00:27:04.600
But there's some denial of service-ish things
00:27:04.620 --> 00:27:09.500
like element tree find all from XML parsing
00:27:09.500 --> 00:27:12.580
or even web page parsing has quadratic behavior
00:27:12.760 --> 00:27:14.360
if you give it certain types of inputs.
00:27:14.740 --> 00:27:16.340
And that means you could create
00:27:16.580 --> 00:27:19.020
like a element tree find all bomb type of thing.
00:27:19.300 --> 00:27:22.880
So there's a bunch of this complexity stuff here.
00:27:23.260 --> 00:27:24.540
But this is also a proper release.
00:27:24.820 --> 00:27:27.720
So I'm going to scroll at a high rate
00:27:27.920 --> 00:27:29.660
the entire time that I'm talking.
00:27:30.020 --> 00:27:32.560
And it is way still going on.
00:27:32.660 --> 00:27:36.600
This is all core Python internals, not even the standard library.
00:27:37.120 --> 00:27:37.480
Ooh, okay.
00:27:37.760 --> 00:27:41.820
And then, now maybe that was a standard library as well, but also build and then Windows and Mac.
00:27:41.940 --> 00:27:44.920
I mean, there are pages and pages and pages of changes here.
00:27:45.220 --> 00:27:48.180
And if you go up one, like I said, there's an interesting theme.
00:27:48.320 --> 00:27:51.200
So if you look at free, free threaded.
00:27:51.600 --> 00:27:51.720
Yep.
00:27:52.020 --> 00:27:52.260
Oh, wow.
00:27:52.260 --> 00:27:56.660
This is one of the, I mean, it is lighting up the find section.
00:27:56.980 --> 00:28:01.880
Yeah, the little scroll bar highlight that shows where the find lands.
00:28:02.260 --> 00:28:05.280
There are so many race conditions that are getting fixed.
00:28:05.440 --> 00:28:07.960
This is just something Python has not really had to deal with.
00:28:08.280 --> 00:28:10.200
And I don't know how I feel about this, Calvin.
00:28:10.340 --> 00:28:13.400
I'm very excited about free-threaded, but there's this many fixes now.
00:28:13.880 --> 00:28:19.520
They should have skipped the.7 and made it.15, like 3.14,.15.
00:28:20.120 --> 00:28:20.820
I know.
00:28:21.780 --> 00:28:24.400
It's the lead-in to 3.15 or whatever, right?
00:28:24.660 --> 00:28:24.920
Yeah.
00:28:25.240 --> 00:28:28.800
Yeah, but in the core built-ins, there's just tons of, like,
00:28:29.400 --> 00:28:32.360
Fix the data race condition and free-threaded build of gc.gen count.
00:28:33.060 --> 00:28:37.180
Fix the potential deadlock on intern from string and other functions on the free-threaded build.
00:28:37.800 --> 00:28:43.040
And so if you're doing stuff with free-threaded, you definitely want to go through this.
00:28:43.040 --> 00:28:50.480
I hope some folks who've been posting blog posts around free-threading go redo their benchmarks against this release to see how that has changed.
00:28:50.880 --> 00:28:51.780
Yeah, 100%.
00:28:52.220 --> 00:28:53.440
Okay, well, that was one of them.
00:28:53.640 --> 00:28:54.060
That's a lot.
00:28:54.130 --> 00:28:54.380
I know.
00:28:54.510 --> 00:28:55.500
That was just one of the things.
00:28:56.000 --> 00:28:57.580
Remember you gave me homework last time?
00:28:57.780 --> 00:29:03.600
I'm like, oh, I just, I literally just added the MCP server for Python bytes.
00:29:04.040 --> 00:29:06.800
And you said, oh, Michael, that'll hold that protocol is out of date.
00:29:06.800 --> 00:29:07.100
You got it.
00:29:07.240 --> 00:29:07.860
There's a new one.
00:29:08.080 --> 00:29:08.600
I'm like, no.
00:29:09.960 --> 00:29:11.260
Well, I upgraded it.
00:29:12.050 --> 00:29:16.500
So Python bytes, it's MCP server, which only existed last week.
00:29:16.940 --> 00:29:19.960
Now uses the, let me see if I get this from memory.
00:29:22.360 --> 00:29:26.740
2026.0728 version two protocol, which is more stateless.
00:29:27.160 --> 00:29:29.140
So it's backwards compatible with the old ones,
00:29:29.320 --> 00:29:31.220
and it now adopts the new one.
00:29:31.390 --> 00:29:32.280
So that's pretty cool.
00:29:32.760 --> 00:29:33.960
And also for Talk Python,
00:29:34.440 --> 00:29:36.820
they were basically a similar update there.
00:29:37.160 --> 00:29:38.520
So how long did that take you?
00:29:39.000 --> 00:29:41.120
I mean, were you able to just point the clankers at it
00:29:41.130 --> 00:29:41.760
and say go?
00:29:42.160 --> 00:29:44.720
The clankers grinded for a while to probably an hour.
00:29:45.040 --> 00:29:45.860
Yeah, probably took an hour.
00:29:46.000 --> 00:29:48.140
It was a pretty significant amount of turn.
00:29:48.360 --> 00:29:49.700
But no worries, I got it done.
00:29:50.040 --> 00:29:50.160
Nice.
00:29:50.380 --> 00:29:52.900
You know, if I had run on Fast MCP,
00:29:53.060 --> 00:29:56.320
which is now just MCP server, I believe,
00:29:56.900 --> 00:30:00.700
is the name of the project now, probably would have been better, even though there were breaking
00:30:00.900 --> 00:30:07.460
changes. But this is integral. It's just a part of the Quart Flask app that are these websites.
00:30:08.060 --> 00:30:11.700
And so it had a little more low-level stuff. But at the same time, it wasn't. I mean,
00:30:11.880 --> 00:30:17.760
my active time was probably 15 minutes. But I had to verify and
check. And so, oh, let me go over
00:30:17.760 --> 00:30:21.680
to Claude and ask it if it still knows how to talk to it. Think of the scale you can handle now.
00:30:22.000 --> 00:30:25.300
I mean, people can just hammer your MCP service now.
00:30:25.540 --> 00:30:26.040
I know.
00:30:26.200 --> 00:30:26.760
It's so stateless.
00:30:26.850 --> 00:30:28.620
They can just ask it questions all day long.
00:30:28.970 --> 00:30:29.540
All day long.
00:30:30.440 --> 00:30:32.060
And in principle, I don't really mind.
00:30:32.380 --> 00:30:38.060
I mean, I'm not inviting to not to do this amount of service because I do get that sometimes
00:30:38.150 --> 00:30:38.840
and it's not fun.
00:30:39.260 --> 00:30:42.480
But it is all the database backed indexed queries.
00:30:42.630 --> 00:30:45.680
So it is like sub millisecond usually when you ask questions like this.
00:30:45.820 --> 00:30:47.080
So it's pretty cool.
00:30:47.400 --> 00:30:47.660
All right.
00:30:48.340 --> 00:30:48.420
Yeah.
00:30:48.780 --> 00:30:50.120
Remember we talked about AgentsView.
00:30:50.360 --> 00:30:51.880
I believe you brought this up.
00:30:51.980 --> 00:31:03.900
And I'm like, all right, so crazy that on my M5 Pro MacBook Pro, I see some portion of my work and projects reported.
00:31:04.020 --> 00:31:09.840
And then I go to my mini and I see some other portion, the complement of my work reported.
00:31:10.000 --> 00:31:15.060
And then I'm like, well, I finally set up Postgres Sync, which is really cool.
00:31:15.060 --> 00:31:19.940
So if you set up this Postgres Sync thing, you just get a web page that you can host.
00:31:20.140 --> 00:31:23.080
I'm hosting it on Tailscale, super private behind all my stuff.
00:31:23.480 --> 00:31:23.640
Nice.
00:31:23.910 --> 00:31:26.260
And then that gives you a unified view.
00:31:26.900 --> 00:31:30.520
There's just a little daemon sort of thing that runs on each machine that just, you know,
00:31:30.740 --> 00:31:32.380
normally it's running on SQLite.
00:31:32.470 --> 00:31:34.600
It just looks at SQLite, goes, what's changed?
00:31:34.600 --> 00:31:41.100
And it just pushes those changes and merges them on a Postgres thing that's running in Docker on my mini behind Tailscale.
00:31:41.400 --> 00:31:47.040
I have a question for you on that because we're using agents view a little more to help with that analysis
00:31:47.300 --> 00:31:48.880
and looking at our users and models.
00:31:49.300 --> 00:31:54.000
The Postgres sync, could you have multiple users push into it?
00:31:54.360 --> 00:31:55.640
Because that's exactly what I think I want to do.
00:31:55.970 --> 00:31:59.520
The one thing to be a little careful of, and I kind of would like to just say,
00:31:59.730 --> 00:32:05.900
could we just not put this data here at all, is it syncs the actual chat conversation, the chat log.
00:32:06.140 --> 00:32:06.540
The sessions.
00:32:06.800 --> 00:32:10.340
Because you're getting the full fidelity of the agent's view, but in a centralized spot.
00:32:10.840 --> 00:32:13.780
Yeah, I mean, it is cool that you can go and do things like hit command K
00:32:14.030 --> 00:32:18.300
and then do vector and semantic search across your work.
00:32:18.540 --> 00:32:21.380
But at the same time, the more people send in stuff in,
00:32:22.300 --> 00:32:24.780
it's plausible that you could write another script
00:32:25.010 --> 00:32:26.080
that just deletes that data.
00:32:26.410 --> 00:32:26.500
Yeah.
00:32:26.880 --> 00:32:29.260
After, right, like just goes and just zeroes it out
00:32:29.400 --> 00:32:31.980
because there's actually zero words in this,
00:32:32.110 --> 00:32:34.020
even though the token count was recorded and so on.
00:32:34.280 --> 00:32:35.900
I haven't tried that, but possible.
00:32:36.320 --> 00:32:37.480
I like the fact you can do like,
00:32:37.760 --> 00:32:39.620
I guess it even mentions their team dashboard.
00:32:40.240 --> 00:32:40.520
Yeah, exactly.
00:32:41.040 --> 00:32:42.560
And I am a team of one for this,
00:32:42.800 --> 00:32:45.420
but I'm still working in two places and I need it
00:32:45.480 --> 00:32:46.620
because otherwise it doesn't work.
00:32:46.780 --> 00:32:47.260
Yeah, yep.
00:32:47.520 --> 00:32:47.980
I've got the same.
00:32:48.300 --> 00:32:48.860
I'm going to set that up.
00:32:49.120 --> 00:32:49.240
All right.
00:32:49.480 --> 00:32:50.740
Let's carry on this order.
00:32:50.960 --> 00:32:53.100
Like I said, this is why this had to be promoted.
00:32:54.700 --> 00:32:55.260
Brian Okken.
00:32:55.660 --> 00:32:56.000
Hey, Brian.
00:32:56.420 --> 00:32:56.980
Hey, Brian.
00:32:57.440 --> 00:33:03.820
He had talked over last year on the show about working on Lean TDD, TDD Without the Waste,
00:33:03.860 --> 00:33:04.740
and sort of redoing that.
00:33:05.280 --> 00:33:08.580
And he has just released the Audible version.
00:33:08.820 --> 00:33:09.400
How cool is this?
00:33:09.760 --> 00:33:10.360
I like that.
00:33:10.700 --> 00:33:12.220
And it's him narrating it, too.
00:33:12.500 --> 00:33:17.340
So if you missed Brian's voice from this podcast, you can go over and listen to him read the
00:33:17.360 --> 00:33:17.980
whole book to you.
00:33:18.240 --> 00:33:18.940
It's pretty wild.
00:33:19.060 --> 00:33:20.180
Like, this is audible.
00:33:21.220 --> 00:33:22.220
And it's Brian.
00:33:22.920 --> 00:33:23.500
Yeah, it is Brian.
00:33:23.820 --> 00:33:25.040
And he did a really good job.
00:33:25.040 --> 00:33:27.120
It sounds, you know, professional and great.
00:33:27.220 --> 00:33:28.400
And I very much like it.
00:33:28.500 --> 00:33:30.580
And I know that when he was working on this book,
00:33:30.900 --> 00:33:34.720
one of his goals was to not make it too overburdened with code samples
00:33:35.400 --> 00:33:38.600
such that it could be an audio book that is actually worth listening to.
00:33:38.780 --> 00:33:40.480
So link into that and check that out.
00:33:41.120 --> 00:33:41.680
Two more things.
00:33:41.760 -->
00:33:44.000
These have to do with Talk Python courses.
00:33:44.560 --> 00:33:50.300
So I talked to someone recently who wanted to take a course from Talk Python.
00:33:50.780 --> 00:33:52.980
And they said, well, this is really great.
00:33:53.100 --> 00:33:55.060
And this is exactly what I want to take.
00:33:55.320 --> 00:33:57.880
But it turns out our company, our organization,
00:33:58.460 --> 00:34:02.320
has some subscription to some other junkier Python training thing
00:34:02.420 --> 00:34:03.320
that I don't want to use.
00:34:03.680 --> 00:34:06.360
But they said if they're going to cover it, I have to use their Python thing,
00:34:06.760 --> 00:34:10.300
not some other random internet thing that who knows what it is, right?
00:34:10.600 --> 00:34:12.520
So I thought, well, that's not right.
00:34:13.399 --> 00:34:18.379
Shouldn't people be able to, you know, if you're a company, get a sense of this is a good thing for your company.
00:34:18.460 --> 00:34:22.480
So if you work for especially a large company, this falls into that category.
00:34:22.919 --> 00:34:27.980
I have a program that I set up for you called a free Python training evaluation for teams.
00:34:28.500 --> 00:34:37.139
So what it is, is for up to 15 people, they get complete free access to one of the, pick one, one of the Talk Python 60 courses.
00:34:37.780 --> 00:34:38.879
And then you can go through it.
00:34:39.360 --> 00:34:43.100
And then that team can report back like, hey, was this worth doing?
00:34:43.300 --> 00:34:47.260
We maybe bring Talk Python as a vendor into your company.
00:34:47.770 --> 00:34:52.460
You get a really nice dashboard of how all your students are doing,
00:34:52.810 --> 00:34:53.520
going through the course.
00:34:53.919 --> 00:34:56.419
And yeah, so who's redeemed the seats?
00:34:56.820 --> 00:34:57.540
Are they finished?
00:34:58.020 --> 00:35:00.060
Like all that kind of stuff.
00:35:00.170 --> 00:35:02.000
You can download updates and so on.
00:35:02.140 --> 00:35:03.260
No MCP server though.
00:35:04.320 --> 00:35:04.720
Not for this.
00:35:05.380 --> 00:35:05.820
Not for this.
00:35:05.900 --> 00:35:08.260
But you could download the CSV and then go to town on it, right?
00:35:08.540 --> 00:35:12.460
So everybody gets basically free access to the course, mobile apps, all that.
00:35:12.540 --> 00:35:24.160
So if you work for a team, if I work for a large company and you have a team there and there's like a lot of restrictions on what training options are available, please reach out to me, Michael at talkpython.fm.
00:35:24.800 --> 00:35:25.520
Let's set one of these up.
00:35:25.630 --> 00:35:27.620
It costs you nothing to try it out.
00:35:27.860 --> 00:35:29.600
Get the name brand training.
00:35:30.070 --> 00:35:30.180
Exactly.
00:35:30.640 --> 00:35:32.400
Not the generic lookalike.
00:35:33.080 --> 00:35:33.440
Exactly.
00:35:33.610 --> 00:35:37.460
Not the generic where you've got to press play every single video and navigate around.
00:35:37.600 --> 00:35:37.820
Not good.
00:35:38.420 --> 00:35:44.500
And related to that, I've also set up a government procurement page that gives all...
00:35:44.500 --> 00:35:45.180
Wow, you're fancy.
00:35:45.520 --> 00:35:46.060
That's really fancy.
00:35:46.290 --> 00:35:51.340
It existed before, but it wasn't on a page where people knew that they could get it.
00:35:51.560 --> 00:35:55.320
So it lists our sam.gov unique ID.
00:35:55.510 --> 00:35:59.740
So if you work for a government entity, there's all these rules like, oh, it has to be registered with this thing.
00:35:59.830 --> 00:36:00.740
And it has to be active.
00:36:00.830 --> 00:36:04.400
And you have to have a 508 report on your accessibility, et cetera, et cetera.
00:36:04.960 --> 00:36:05.960
I put all that stuff up here.
00:36:06.100 --> 00:36:11.240
So if you work for a government and you would like a governmental agency, especially the U.S., please check it out.
00:36:11.900 --> 00:36:13.220
This all makes it super easy.
00:36:13.320 --> 00:36:16.060
You probably have to address this kind of stuff some of the time.
00:36:16.280 --> 00:36:18.780
It looks like you put a lot of work into this, and it shows.
00:36:19.100 --> 00:36:19.400
Thank you.
00:36:19.400 --> 00:36:19.660
It's amazing.
00:36:20.020 --> 00:36:20.220
Thank you.
00:36:20.280 --> 00:36:20.560
Thank you.
00:36:20.780 --> 00:36:20.940
All right.
00:36:21.040 --> 00:36:22.900
Well, those were many extras.
00:36:23.320 --> 00:36:24.480
Those were many, many extras.
00:36:24.859 --> 00:36:26.140
We are done with the extras.
00:36:26.320 --> 00:36:31.640
I've got one extra, which is a follow-up to last week's episode around the post-quantum key exchange.
00:36:32.240 --> 00:36:35.380
And the fine folks over at Astral have made a new release of uv.
00:36:36.240 --> 00:36:37.660
That's what it was like five days ago.
00:36:38.160 --> 00:36:42.000
And it now supports and prefers post-quantum key exchange.
00:36:42.540 --> 00:36:44.160
So they've already got your back.
00:36:45.220 --> 00:36:48.320
You can opt in to TLS diagnostics,
00:36:48.920 --> 00:36:51.760
but you're going to get post-quantum key exchange out of the box
00:36:52.110 --> 00:36:53.040
when you're using uv node.
00:36:53.470 --> 00:36:54.240
So I thought that was pretty cool.
00:36:54.420 --> 00:36:54.940
You know what?
00:36:55.320 --> 00:36:56.280
Just bookend it with uv.
00:36:56.670 --> 00:36:57.080
Well done.
00:36:57.540 --> 00:36:57.840
Well done.
00:36:58.860 --> 00:36:59.220
Astral.
00:37:00.120 --> 00:37:00.880
Open AI team.
00:37:03.760 --> 00:37:09.480
so far they're still doing again the good work i appreciate their what they're putting out there
00:37:09.480 --> 00:37:14.360
for the community and keeping the again staying on top of us like that's really cool that
00:37:14.370 --> 00:37:19.680
this has we went from one week being announced to next week it's in use and you're getting it
00:37:19.910 --> 00:37:25.220
yeah that's super cool and i think i do think they're doing a great job so yeah yep so grab
00:37:25.400 --> 00:37:30.359
the latest version of uv yeah i love it all right are you ready for a joke let's close it out do it
00:37:30.360 --> 00:37:32.940
This joke is called beware of dog.
00:37:33.720 --> 00:37:33.840
Oh.
00:37:34.080 --> 00:37:35.600
We've all seen the beware.
00:37:36.740 --> 00:37:38.640
Yeah, Mushu made his appearance last week on the show.
00:37:38.660 --> 00:37:39.500
Yes, exactly.
00:37:40.060 --> 00:37:40.180
Exactly.
00:37:40.740 --> 00:37:42.140
We've all seen the signs.
00:37:42.540 --> 00:37:45.540
Do you really, is it just a, be careful, there is a dog here?
00:37:45.540 --> 00:37:48.100
Or is it like a dangerous, aggressive dog?
00:37:48.220 --> 00:37:48.560
We don't know.
00:37:49.020 --> 00:37:49.780
But this one is aggressive.
00:37:50.080 --> 00:37:50.760
Oh, let me tell you.
00:37:51.720 --> 00:37:54.980
So there's a big sign that says, warning, beware of dog.
00:37:55.360 --> 00:37:56.440
He uses Kali Linux.
00:38:00.140 --> 00:38:00.820
So you're laughing.
00:38:01.160 --> 00:38:03.120
Tell people what Kali Linux is for those who don't know.
00:38:03.380 --> 00:38:06.180
For those security folks who are looking to do some pen testing,
00:38:06.340 --> 00:38:08.780
you may have installed Kali Linux in the past.
00:38:09.140 --> 00:38:13.460
So it provides some attack tools to let you go after other people's servers,
00:38:13.940 --> 00:38:14.640
or your own, hopefully.
00:38:15.080 --> 00:38:15.560
Yeah, yeah.
00:38:15.640 --> 00:38:15.980
Not others.
00:38:16.380 --> 00:38:20.760
It basically is a distribution that bundles a whole host of things
00:38:20.860 --> 00:38:24.160
you would separately install if you were a pen tester, hacker sort of thing.
00:38:24.200 --> 00:38:26.240
And it comes ready to go, right?
00:38:26.680 --> 00:38:28.380
That needs to be the new logo for Kali Linux,
00:38:28.560 --> 00:38:35.320
like a big junkyard dog with a big like spiked collar exactly love it terminal with the uh the
00:38:35.520 --> 00:38:43.020
name and address on it there's no place like one two seven zero zero one exactly tattooed on his arm
00:38:44.960 --> 00:38:49.000
you know i'm a sucker for dad jokes that i will laugh every time i know these are these are
00:38:49.070 --> 00:38:54.559
pretty solid these are pretty solid so beware the dog he's his cali Linux all right well that is it
00:38:54.580 --> 00:38:56.380
for the show, Calvin,
00:38:56.970 --> 00:38:58.080
thank you for being here with all of us.
00:38:58.230 --> 00:38:58.700
Yeah, you bet.
00:38:58.880 --> 00:39:00.980
And check out Pydantic Logfire
00:39:01.520 --> 00:39:03.080
at pythonpys.fm slash logfire.
00:39:03.280 --> 00:39:05.100
Really appreciate them supporting the show.
00:39:05.440 --> 00:39:06.260
Update your Pythons.
00:39:06.720 --> 00:39:06.860
Yes.
00:39:07.340 --> 00:39:09.260
uv, Python, even had an upgrade.
00:39:09.370 --> 00:39:11.680
So for in-place point release upgrades for uv,
00:39:11.790 --> 00:39:13.280
you can just, there's a single command now
00:39:13.280 --> 00:39:13.920
to update all of them.
00:39:14.060 --> 00:39:14.760
So just do that.
00:39:15.020 --> 00:39:15.600
Stay safe out there.
00:39:16.020 --> 00:39:16.300
That's right.
00:39:16.390 --> 00:39:16.660
Bye, y'all.
00:39:16.870 --> 00:39:17.100
See ya.