#495: Banned

00:00 Hello and welcome to Python Bytes, where we deliver Python news and headlines directly to

00:04 your earbuds. This is episode 495, recorded on Tuesday, September 8th. I'm Calvin Hendricks

00:11 Parker. And I'm Michael Kennedy. Glad to have you, Michael. If you want to support the people who

00:16 make this possible, make sure you support our sponsors, the fine folks at, oh, Talk Python.

00:21 Make sure you check out Michael's courses. He is doing incredible work over there. It's a great way

00:26 to brush up on some technology you weren't sure about,

00:29 try something new.

00:30 He's got some really cool notes.

00:31 You just released a couple of new courses, right?

00:33 Yeah, we just released a Rust course, which I think is really nice.

00:37 Yeah.

00:38 Very cool.

00:38 And then also check out the fine folks over at Six Feet Up.

00:41 They are the ones who love solving hard problems.

00:44 They do a lot of Python work.

00:45 They've been around for 27 years.

00:47 And we'd love to talk to you about doing some cool projects.

00:50 If you want to connect with the hosts online, you can find Michael, myself, and the show

00:54 on Mastodon, Bluesky, and X.

00:56 And you can also find Michael and I on LinkedIn.

00:59 But make sure you join us on YouTube for the live stream at pythonbytes.fm/live to

01:03 be part of the audience.

01:04 Usually it's Tuesday mornings, Pacific time.

01:07 Older video versions are available there too.

01:09 And finally, if you want our artisanal digest of every week's show notes in email form,

01:15 add your name to our friends of the show list and we'll never share it, but you'll get a

01:19 really cool recap of everything we talked about.

01:21 Michael, thinking about recaps, what you got?

01:23 I always look forward to when the videos of the conferences come out.

01:28 How about you?

01:29 Same. I do too.

01:30 I mean, sometimes I'm at the conference and I just miss a great talk.

01:34 And I'm so glad that many of the good ones are recorded.

01:37 Exactly the same.

01:38 And sometimes you don't miss it because you just want to...

01:42 There's trade-offs, right?

01:42 Like, I'm going to go to that talk instead of that one.

01:44 Or, oh, there's a really interesting group of people having this conversation.

01:48 And the talk is starting now.

01:49 You got to decide, do I join the group and have the experience that's not going to be recorded or do I go to the talk?

01:56 So, EuroPython2026 has all of the videos out.

02:02 And I realize I need to update what I put in the show notes because this is just my subset.

02:08 So, if you want to watch all the talks for EuroPython, 117 videos are now online on YouTube, which is great.

02:17 That is fantastic.

02:18 And quick.

02:18 Yes.

02:19 Because the conference just happened.

02:20 Yes, I know.

02:21 It's really, really good.

02:22 And I went through and said, hmm, what might make an interesting thing that I personally

02:29 would be interested in?

02:30 This is not a voting.

02:31 This is not a ranking.

02:32 This is not a good, bad.

02:33 This is Michael wants to watch this for two reasons.

02:35 Either one is relevant to me, or I might be considering it for a Talk Python episode.

02:40 Like, for example, the very first one by Giovanni Bairoliari is already out over on Talk Python

02:47 Oh, nice.

02:48 As the latest, he did a cool project rewriting asyncio

02:54 from a perspective of what if there was only free-threaded Python?

02:58 You wouldn't write asyncio exactly the same because you have the potential to split that work

03:03 across multiple threads, whereas asyncio is single-threaded, et cetera, et cetera.

03:07 So I'm finding these really interesting.

03:09 We've got Marlene Magami.

03:12 Her talk, her keynote is up.

03:14 Updates on the JIT, update on free threaded Python by Thomas Warters.

03:20 Paul Everitt is in here somewhere because it's Paul Everitt.

03:23 Yeah, look, I haven't watched, I have not watched his talk yet, but I really want to,

03:26 because he talks a lot about kind of the future of AI-driven development and local models and controlling your destiny and your data.

03:33 Yeah, exactly.

03:34 Stefan Ostrowski deploying Python web apps by Will Vincent, probably with a lean towards Django, hinted to your topics, maybe.

03:43 Maybe.

03:43 But anyway, I thought I would make my, like this morning I did it, made my watch list public because I don't know.

03:51 I went through all 117 videos and said, okay, maybe, probably.

03:55 And now I'm down to 29.

03:57 So people can check that out or they can just look at the full list.

04:00 I love that you make it a watch list so you get Michael's curated version.

04:04 And so I can go enjoy what you think might be important because maybe I follow similar topics.

04:09 So I think more folks should do that with conference videos because it's so hard to go.

04:13 I mean, there's, you said what, 117 videos, what you said?

04:16 Yeah, that's a lot.

04:17 That's a lot of time to invest.

04:19 I would venture there's probably not a single person that watches all videos all the way through, you know?

04:26 Yeah, that'd be hard.

04:27 That would be super hard.

04:28 That's just a lot of time.

04:29 That's a lot of time.

04:31 Well, and I just got back from DjangoCon, which was in Chicago.

04:34 And same time, just shortly afterwards, the PyCharm folks and the Django Software Foundation have

04:41 released the 2026 State of Django. And I love so that William Vincent also published this over on

04:48 the PyCharm blog. I love the fact that it's called Boring is Back. If you've ever read the blog post

04:56 about boring tokens, like you only get to have so many new shiny technologies, shiny tokens,

05:02 You only get so many new shiny technologies in your project.

05:05 The good news is you can have Django and you can have Postgres and they're both considered very boring and very stable.

05:11 If you look through here, it's very thriving.

05:14 Postgres is actually the database of choice for almost 80% of the respondents who are doing Django work out there, which I think is a really telling mark.

05:22 Another nice thing to see here is 43% are already on Django 6.0 months after it shipped.

05:28 So folks are staying up to date and pretty well on the stable side of things.

05:33 It was 3,500 Django developers across more than 40 countries.

05:36 So it's a fairly good representation of the world's uses of Django.

05:41 I'm sure there's a lot more.

05:42 Statistically, I think that that's a pretty big sample size.

05:45 Yeah.

05:45 I mean, they did a good job of promoting it across multiple podcasts and conferences and

05:51 things like that.

05:51 And I think it's been really well received.

05:54 They got a nice little infographic here.

05:56 What I thought was interesting is this relates to maybe a future topic of yours.

06:00 HTMX is closing in on React for the front end.

06:04 You see here we've gone from 2025, 26% of projects maybe using, or we're using HTMX to

06:10 now 34% catching up to React, which is only at 38% on the front end.

06:16 I'm pro that.

06:17 I think that's a really great sign of maybe people are looking for the simpler, more boring

06:22 side of the web.

06:23 And Django is a great place to do it.

06:25 folks obviously using AI. So it's interesting, they had a couple of special sections in here.

06:29 AI is mainstream, but no workflow is one. So whether they're using it all the time,

06:34 several days a week, once or more often, AI is dominating the workflow, but it's still supervised,

06:40 which I think it should be. The AI is here to make us better developers, not replace us as

06:45 developers. I think we've seen that time and time again, as we talk about AI on the various shows.

06:50 And so the report here, the stats definitely show it.

06:53 I mean, that's the way to get the most out of these tools.

06:57 VM is still the top of the list for managing your Python environment,

07:01 not necessarily your Python dependencies, but the Python environment.

07:04 I have a feeling uv will overtake here real soon

07:07 because once you see uv work for getting your Python version,

07:11 it's hard to go back seeing how fast and easy that is to be managed.

07:15 Docker at 31%.

07:16 So I'm also quite excited to see the fact that folks are containerizing and just using the Python coming out of the containers. That's also

07:22 a very simple and easy way to just get started on the right foot with Python and Django is use

07:28 Python that's already in a Docker container. Yeah, yeah. And we're both on team Docker.

07:32 Oh, yeah, for sure. And in fact, I'm on team VE and V in Docker for consistency across the dev

07:40 workflow. Me and Hennig. Right. I am on, actually, I use Docker, but a standard container

07:48 and install my own uv python via uv for consistency exactly same like i don't yeah i don't want to

07:53 pick some pre-made thing like well python comes with it's like it's so easy to install python

07:58 into it pick pick yeah specifically i think and you can you run into weird edge cases if you start

08:04 just kind of coloring slightly outside the lines with one of the standard python ones it could be a

08:08 c dependency or just a weirdness of you want to install something and it gets it gets hard so i

08:18 to install this way to go. Rough top of the list for code tooling. No surprises there. I think we

08:25 should see more pre-commit coming up given what we'll see down here later. There's lots of testing

08:30 going on. AI is write more code. Verification matters more. You see right here, pytest 45%.

08:37 So we will, we see lots more testing happening, which I think is also a very good sign. And I'll

08:42 talk a little bit about that in my second article as well. So I think, yeah, go check it out.

08:48 This was also part of a fundraiser.

08:50 So if you are renewing your PyCharm Pro license or looking for one, you can get 30% off.

08:56 And then you also get 100% of that purchase price goes to the Django Software Foundation.

09:00 So not an ad for PyCharm necessarily, but they are doing good out there in the world.

09:04 I like a lot about this.

09:06 Specifically, I also think this is a deeper view into Python in general.

09:11 When I saw the headline for the survey, I thought, oh, it's going to survey Django things.

09:16 Like who's using this feature, the ORM, and then who's deploying this way versus that way.

09:21 And no, it's more of a general thing.

09:23 Just let's ask the Django people what they're up to.

09:27 Yeah.

09:27 Yeah.

09:28 Nope.

09:28 Really good stuff.

09:29 It is good stuff.

09:29 I'm glad you covered it.

09:31 Speaking of good stuff.

09:32 Well, I mean, boring is good.

09:33 But you mentioned HTMX.

09:35 Well, HTMX 4 is now out.

09:39 And that's officially released.

09:40 Officially.

09:41 And it's a Jeep.

09:43 No, it's not.

09:43 Excuse me.

09:44 No, it's a horse.

09:45 It's a Land Cruiser, I believe, not a Jeep.

09:49 Anyway, there's a big picture on the announcement of some Land Cruiser jumping.

09:54 That's the same cover art for the Game Boy Color cartridge that they produced in the announcement as well, which I have.

10:01 You have? Okay.

10:02 I have. I've been playing it too.

10:03 So thank you, Carson, for doing that.

10:05 In the release, HTMX 4.000, yay for non-zero ver.

10:11 It says, after eight months of work, plus a game, and if you go to the game,

10:16 this is a real Game Boy cartridge you can get, yeah?

10:19 Yep, I have said cartridge is not made up.

10:21 It is legitimately a cartridge you can play through.

10:24 If you play all the way through, your reward is to see the source code for HTMX 4.0.

10:31 Incredible.

10:32 It is.

10:32 All right, so there's a lot of interesting changes here.

10:36 I think, first of all, HTMX has been really broadly adopted

10:40 by people who are just like, I don't want to take on a React front-end framework or something

10:45 like that. Even though all of the AI tools leap at using React for stuff, it's still really popular,

10:53 which I think people are choosing. Why do you want to develop your models twice?

10:58 Yeah. Why do you want a bunch of build steps and all that stuff just to get your app out?

11:05 Yeah. If you don't need it, don't go all in. Also, a lot of these

11:10 front-end frameworks, not HTML, but many of them encourage you to treat your back-end as a dumb

11:16 pipe that just routes data to the front-end where the smarts live. And it's fine if you want to write

11:22 your code on the front-end in JavaScript, but if you want to leverage the things that are really

11:26 nice about Django or Flask or whatever you're writing against, and just Python in general,

11:31 well, then it kind of undermines that, right? Yeah, but it's the last you just stay in the

11:35 batteries included world especially with Django like all the html like filters and helpers that

11:42 are built in to Django the framework are all still very useful with a tool like this they become

11:47 useless if you're on react like you're not going to use it right exactly exactly so major changes

11:53 attribute inheritance is now explicit it's a little is on a python isn't it um so it used to be you

12:00 could do like an hx-confirm and if some action was going to happen first a dialogue would pop up

12:06 and say are you sure you want to delete all the data and you would say yes or no but you could put

12:11 that in say a div that contains some elements and then there was a button way down if you click the

12:15 button the confirm would fire that no longer happens so you have to explicitly say at the top

12:20 in the confirm and i want this to cascade down to the children biggest part yeah except for that it

12:26 it stops working. That's the only problem. On the older code, so there's an upgrade path. It's not

12:30 major, but there is, I mean, it's kind of a find and replace upgrade path, so it's not a huge deal.

12:35 The event names have been changed a little bit. Yeah, we'll go through some examples, I guess.

12:40 So yeah, there's like, you could do an hx confirm and then a button with an hx delete, and now you've

12:46 got to say colon inherited. The events are also used to be able to say disinherit, so you could

12:52 block a child from having stuff inherited upon it? Leave this part alone.

12:56 Get away from me.

12:58 Now that's gone. Now that's gone because you just that's the default is not to inherit.

13:02 So you don't have to say don't cascade it down.

13:04 Some events were renamed. Here's the find and replace. So it used

13:08 to be htmx before request now it's htmx colon

13:12 before colon request and so on. What else? History. This is

13:16 actually, I didn't know this was happening.

13:18 This is not ideal. So history support allows you go forwards and backwards, even if, so your forwards and backwards can navigate

13:26 between HTMX events, not through just the page loads, right? So if I click a button, yeah,

13:33 and it does a load, it changes something else. And I click another, maybe another row in a grid and a

13:38 master detail, like another detail thing loads, you can go backwards, forwards between those.

13:43 I didn't know this. It used to be that in HTMX2, the way backwards and forwards worked is it would

13:49 save a copy of the pages for restoration and local storage.

13:53 I forget the storage part.

13:54 Like, wait, what if that data changed?

13:58 You know, there's two hard things.

14:00 Yeah, dangerous.

14:02 Exactly.

14:02 There's like two hard things in computer science.

14:05 Caching, naming things, and off by one errors, right?

14:08 Love that joke.

14:09 Yeah, it's there for a reason.

14:11 The caching is super hard.

14:12 And this was so, like, I thought it was just reloading it.

14:15 What's going on?

14:15 So anyway, that's gone.

14:17 Nice.

14:17 It just literally reloads it like I always thought it did.

14:20 So there's that.

14:21 And the last one is, well, there's like two things.

14:24 There's morph swaps.

14:25 So what this is, I had to look this up, folks, is a morph swap is that instead of saying this portion of the page is changing,

14:35 so we're going to get the new partial or fragment from the server

14:38 and just replace that, a morph swap will rewrite the DOM to match the new segment.

14:44 So for example, it doesn't jump around as much and those kinds of things, right?

14:49 If maybe only some tiny element, even though you replaced a whole section,

14:53 changes, it's like, okay, we'll just change the little element

14:56 because, again, master detail, like there's a whole structure of the HTML there,

15:00 but really as you click from page to page, it's just the values inside that structure.

15:04 That sounds hard as a framework, but much easier for the end user

15:07 who's building with the framework.

15:08 Exactly.

15:10 Let them suffer.

15:11 Apparently they did the suffering for us, which I love.

15:13 Yes.

15:14 I mean, honestly, it's not the point of a framework.

15:15 It is.

15:16 I mean, why would you do a framework if you're not going to leverage that?

15:20 And then the HX partial tag.

15:23 So this for outer band changes.

15:26 This is where HTMX used to be really brutal.

15:29 And honestly, I've worked on a project before.

15:31 And in the end, I'm like, oh, I kind of wish I didn't choose HTMX for this.

15:34 Because I want to update something over here and something over there.

15:37 And this part here.

15:38 And it was tough.

15:40 So this lets you specify partials.

15:43 then um elements of it that change right so pretty pretty neat um that's slick it allows you to to

15:49 sort of glue that stuff together very nice yeah so there there it is i mean and they definitely get

15:56 best launch swag gifts or whatever the game boy cartridge over the top and real i love it

16:03 i know the website you can go to the website and play it too but i didn't even have the

16:08 eminid version that's cool so everyone can enjoy it if you don't have a game boy

16:12 but that's crazy right it is super good it works yeah it's a beautiful cartridge really well done

16:17 like the game was surprisingly well done for i can't imagine what how much time they spent on it

16:22 well carson definitely is known for um interesting and catchy stuff so well played well well played

16:30 well played i love it that's awesome yeah let's head over you mentioned zen some zen of python

16:36 things. So for the final news article for me, this is actually a local IndyPy person. He presented

16:44 about two IndyPys ago, Kyle from Testable. He gave a little nice little blog post about

16:49 functionally Zen. And I kind of like going back over Zen of Python things. It's one of my favorite

16:55 talks to give at PyCon and other conferences is something more around these initial tenants of

17:02 Python and simplicity and why we love it. And for those of you who don't know, you can go to your

17:07 command line, you can type Python minus m this, and you will see the little Easter egg, which is the

17:12 Zen of Python that is in Python. Everyone should print it out, they should frame it, and you should

17:16 put it in your, above your sink when you brush your teeth at night so you can read it every night.

17:21 And you'll become a better human being instantly because you've got the Zen of Python there

17:25 available to you. But Kyle took the Zen of Python and he actually put, took kind of added into it

17:30 like seven extra tenants for Python simplicity.

17:33 So he took specifically the simple is better than complex tenant of Python

17:39 and then added these seven additional pieces that are about idiomatic code, data,

17:45 pure functions being simpler than impure functions.

17:48 It kind of goes on like in class functions being maybe simpler than classes.

17:53 So it's just a way to think about how to keep your code again simple.

17:57 There's no reason to get overly complex if you don't have to.

18:01 And Python gives you all the tools that are right here.

18:04 So you can talk about, again, he talks about being idiomatic.

18:07 There's things you need to watch out for.

18:08 For example, side effects, especially with classes that have constructors,

18:12 they can have side effects in subclasses that you aren't aware of.

18:16 So it just adds to the complexity and overhead as a developer.

18:20 Some good examples in here.

18:21 So if you looked at the idiomatic code is simpler than non-idiomatic code,

18:25 some functional programmers might be tempted to use the map function of Python to extend a list.

18:32 But there's actually a simpler way.

18:33 Just using a list comprehension is actually a clearer, probably cleaner way

18:38 for writing what you're looking for there.

18:40 So this actually loops through and does the x times 2

18:43 for x in a list of 1, 2, 3.

18:45 So now you've got 2, 4, 6.

18:46 And sometimes I think even better performance.

18:48 Not hugely, but a little bit.

18:50 Well, now the docs should be able to tell us, right?

18:51 Because I've got the notation for those things in there.

18:55 It's right in the docs.

18:56 Data is simpler than functions.

18:58 This one's a little harder to do.

18:59 I think the example was a little contrived because if you have an object that has age

19:04 with a sub-element of weight and a sub-element of lab measurement,

19:08 I don't know if you'd have that structure necessarily all the time,

19:10 but if you can get your data structured that way,

19:12 it's probably easier to use, again, a dictionary lookup,

19:15 and actually it's probably faster to use a dictionary lookup

19:18 than it is to run a function that would be taking in arguments

19:21 and looking up things based on those arguments.

19:23 So I mostly agree with that example.

19:27 This is about the impure functions or pure functions versus impure. So this pure function, for example

19:33 You're certain is always going to return an integer like it has a reliable return to it

19:39 This the second example where they're talking about fetch doubles

19:42 The first one was like a double where you just to return x times two

19:45 So it doubles the number the second one which was like fetch

19:48 Under double it actually uses requests to get the response and then it tries to return the body

19:54 looking for an answer key, that is not as stable.

19:58 It could be it returns two or four.

20:01 It could be the returns of 500 server error.

20:03 Could be the returns some other, if a developer makes a mistake someplace,

20:06 some other not expected answer or a string.

20:09 So that's just an example of what It might even return an exception,

20:13 like network not available.

20:15 And then the pure functions versus classes.

20:18 I mean, if you look at the example here, a function you can do in as simply as two lines

20:23 with a definition of the function and a return.

20:26 When you start talking about classes, you just get a lot more boilerplate going on inside the class.

20:31 So if you don't need the classes, there's probably no reason to go that direction.

20:35 And it kind of goes on and on and on here with more and more examples

20:39 that just show some simpler ways to do that.

20:42 This is the last one I'll talk about, which is the side effects.

20:45 If you have a class that has a constructor in it,

20:48 so you've defined a __init__() method, that __init__() can affect subclasses

20:52 and you may not be aware of it because you've not looked at the code for it.

20:55 So they have a time client example here with a __init__() that initializes a token

21:02 using a request post.

21:03 So it actually makes an API call in the base class.

21:06 These other ones for time service and time controller

21:09 are calling a constructor that uses that client,

21:12 which means that every one of these will also make an API request.

21:16 So you may want to manage that a little cleaner

21:19 and reduce them or say, for example, external calls

21:21 you're making in that case. The last bit I'll call out in honor of Brian Okken and his Lean TDD book,

21:28 tests with no mocking are simpler than tests with mocking. I think we can all agree, but I think

21:33 there's a world that needs to hear it. And there's a book about it. So if you check out Brian's Lean

21:37 TDD book, he talks about this as well as part of that book. So I thought that was pretty apt and

21:43 timely that this came up. And thanks, Kyle, for helping remind folks of the simpler ways and the

21:49 nice idioms that Python gives us out of the box. I really like it. I remember when I first got into

21:53 Python, I hear Pythonic. I'm like, but I need examples. These words are great, but give me

21:58 examples so much. Yeah. I learned by example. And these are just great examples of like, oh,

22:03 that's how I should do that. Okay. I can follow that pattern every time. Yeah, absolutely. That's

22:07 good stuff. Speaking of following a pattern, your screen's already up. Maybe you just take the extras.

22:11 Yeah, I will. I'll go with the extras first. And continuing in my three-week streak of talking

22:17 about uv releases and new features. The latest one released four days ago, uv 0.12.10, now ships the

22:26 trusted publisher token revocation, and it has Python 3.15 support. So uv is on a terror of keeping

22:33 up with all the latest bits that are going on, including the Python 3.15 releases coming up. So

22:40 that's really cool that they're doing that. And then the other bit, I saw a fun little

22:44 right release here making a python interpreter in 1024 bytes so kind of a little example of

22:51 of if you've ever done code golf and do something in in such and such a number of steps this person

22:57 implemented a full python interpreter in 1024 bytes of good old c code so that's pretty wild

23:05 he i think he first attempted with 512 bytes uh which seems kind of crazy yeah good for you austin

23:12 the subtitle was 512 bytes. It's not enough. Like, right. No kidding. Wow.

23:18 So yeah, he, he chronicled his whole journey. I just thought that was fun. not necessarily a

23:23 full news article, but it's pretty cool. The people out there doing still cool stuff like this.

23:27 That's some minified Python, not something you normally see.

23:30 Right. But yeah.

23:32 Wow. Okay. I just got a one quick one. Steering council elections are going on. So

23:41 Oh, two week nomination period followed by a voting period.

23:46 When is this actually?

23:48 I saw that the votes were coming up, but I don't see the actual.

23:50 Anyway, it's previously been November through December or November through November.

23:54 I think it's not yet up or just started.

23:57 Anyway, this is the 2026 one.

23:59 And so candidates, the nominees are here.

24:02 Lots of folks I see.

24:04 Yeah, it's a good list.

24:05 So check it out, folks.

24:07 And yeah.

24:08 So anyway, I just want to give a shout out to that being a thing.

24:10 Hey, Michael, you got any jokes for us?

24:12 I have a band joke.

24:14 A band joke?

24:16 Oh my God.

24:17 You mean like band as in not allowed?

24:18 We're going to be kicked off of YouTube and all the

24:21 Yes, as in disallowed.

24:22 Disallowed.

24:23 We'll only get kicked off Reddit.

24:25 This is not like a trombone versus trumpet.

24:27 Like, what can you do with a trombone?

24:29 You can't hit him with a little slider thing?

24:31 I don't know.

24:32 I used to play the trombone.

24:33 I never really enjoyed it though.

24:37 It was a requirement.

24:39 No, the joke is, keep in mind, there's not programming exactly it's it but i think we can all just like enjoy a little it uh

24:45 stuff you know and it says what single word in it makes you look like this and there used to be a

24:50 picture that had a ah like i can't take it anymore and sorry this post was removed by reddit's filters

24:56 and all it was was a picture of somebody making a weird face like honestly i don't know what was

24:59 wrong with it but someone got up in arms this one this one took me a minute to understand what the

25:05 the heck it was. It said, hello. Do you know what this is about? Why would this make IT people wig

25:12 out? Maybe, you know, I didn't either. It's apparently on Teams or Slack. If somebody just

25:18 says hello, and then they're going to write more, but then they kind of don't, you're like, hi.

25:22 Okay. That could make me a little, okay, let's go. Okay. All right. Not a single word, but anytime

25:28 I get a direct message along the lines, you got a minute? Or can I ask a question? Yes, exactly.

25:35 Oh, let's see.

25:36 So I got to collapse them because they all go down at like,

25:39 this person says printer.

25:41 Just no, no.

25:43 Can't do it.

25:44 What year is this and we still can't reliably print?

25:46 I know.

25:46 I just, this other, just like last week, I print maybe twice a year and I just needed to print.

25:52 And it's like, the printer's out of ink.

25:53 Like, I just bought an ink.

25:54 I haven't printed since the last time I bought an ink.

25:56 Anyway, but as in, not to interrupt your lunch, but.

26:02 Could you add in a well actually?

26:04 Yeah.

26:05 Rico or scan to mail.

26:09 Now I get hello.

26:11 That's pretty good.

26:12 Yeah.

26:13 All right.

26:13 Now I'm in my zone now.

26:15 DNS.

26:16 There's even like a little like Japanese style like water painting.

26:20 It says it's not DNS.

26:22 There's no way it's DNS.

26:23 It's always DNS.

26:24 Like some philosopher.

26:25 Yeah.

26:26 In the early days of Six Feet Up, Gabrielle made a comic about DNS because it was so hard to explain to folks.

26:32 And every time I say there's a problem, she's like, is it DNS?

26:35 and so this is my co-founder partner in life and she totally gets it dns is absolutely the problem

26:41 this one might be for me outlook crowd strike i'll just not ever since you owe this one oh my god

26:47 ever since you xyz this doesn't work anymore um update i'm surprised yeah i'm surprised update's

26:54 not in there oracle i'm on that bandwagon um sharepoint let's see what else we got cve or just

27:01 Citrix, teams, SAP, you got my vote on SAP.

27:07 Oh my gosh.

27:08 Anyway, I thought this was a really fun list.

27:10 I don't understand why.

27:11 That was a good thread.

27:12 I don't understand why they banned it, but whatever.

27:14 Now that makes it even better.

27:15 Because it's Reddit apparently.

27:17 Yeah, I know.

27:17 Awesome.

27:18 I put stuff up there and it'll be like, well, it looks like you're trying to share an announcement

27:23 and that's not allowed.

27:24 So blocked.

27:25 Like just, I wrote a blog post.

27:27 I thought this community would think was interesting.

27:28 So I wanted to share it.

27:29 It's like, I'm not selling you anything.

27:31 I promise.

27:32 Not in this case anyway.

27:33 Not on my blog.

27:35 Sorry, Michael.

27:36 Not allowed.

27:36 Banned.

27:37 Banned.

27:38 Well, that was a good episode.

27:39 Thanks for hanging out.

27:40 And we'll see you all next week.

27:41 I'll still be in France next week, I believe.

27:43 Au revoir.

27:44 Au revoir.