WEBVTT 00:00.000 --> 00:10.000 Welcome, everyone, to the 11th edition of the Go Dev Room. 00:10.000 --> 00:14.000 It's been 11 amazing years, two of which were in the pandemic, we don't count those. 00:14.000 --> 00:16.000 But it's been 11 amazing years. 00:16.000 --> 00:19.000 And as always, we're starting with the state of go, 00:19.000 --> 00:22.000 where we're going to talk about what has changed in the past years 00:22.000 --> 00:27.000 since last two years, one year ago, which was go 1.22. 00:28.000 --> 00:31.000 First of all, we are my, in case you're new here on Omits. 00:31.000 --> 00:32.000 My name is Machia. 00:32.000 --> 00:35.000 I'm a training conductor and training, but that's what's interesting for today. 00:35.000 --> 00:36.000 I'm also a software engineer. 00:36.000 --> 00:40.000 I restylical co-fight, where I work on cloud native security. 00:40.000 --> 00:43.000 And I'm going to talk to you about a few subjects today. 00:43.000 --> 00:46.000 I'm going to talk to you about change in the language. 00:46.000 --> 00:48.000 Change to the new standard library and go. 00:48.000 --> 00:49.000 Change to the go tooling. 00:49.000 --> 00:54.000 As well as an interesting design draft, and of course, a community update. 00:54.000 --> 00:58.000 So you'll be wondering, what has been new since go 1.22? 00:58.000 --> 01:00.000 Of course, go 1.23. 01:00.000 --> 01:05.000 It was released in August 13, 2024, and go 1.24, 01:05.000 --> 01:08.000 which is about to be released in February 2025, 01:08.000 --> 01:13.000 which if you don't realize is starting from today any moment now, 01:13.000 --> 01:16.000 if you're in the go team and have the button press it now. 01:16.000 --> 01:19.000 Nobody, okay, do bad. 01:19.000 --> 01:22.000 Wait a minute, go 1.24. 01:22.000 --> 01:27.000 Means we have been having 25 releases of go 1.0. 01:27.000 --> 01:29.000 Wow. 01:29.000 --> 01:35.000 People, that's 13 years of go 1.0. 01:35.000 --> 01:40.000 If go was a person, it's a teenager, it's going to high school now. 01:40.000 --> 01:43.000 Wow, that's all. 01:43.000 --> 01:50.000 Which follows with the question, is go still that cool new upcoming language we also go learn? 01:51.000 --> 01:55.000 Probably not cool and upcoming. 01:55.000 --> 01:58.000 We should be learning it, but it's not a new fancy thing 01:58.000 --> 02:02.000 that's going to be the next language, because it is the next language already. 02:02.000 --> 02:05.000 So let's talk about anywhere to come in language. 02:05.000 --> 02:08.000 What's anywhere to rest? 02:09.000 --> 02:16.000 Sirius, what are the top programming languages? 02:16.000 --> 02:21.000 And I found better metric than the stack overflow survey or the state of GitHub. 02:21.000 --> 02:24.000 I use the only true metric in the world of programming languages, 02:24.000 --> 02:28.000 which is the size of a dev room and falls down. 02:28.000 --> 02:34.000 This is the go dev room, the past year since 2019 since I had been starting counting this chart. 02:34.000 --> 02:36.000 I didn't want to enter the whole 11 years. 02:36.000 --> 02:37.000 Oh, that's lazy. 02:37.000 --> 02:42.000 Are you seeing our nice chart flowing here between 400 and 600 people? 02:42.000 --> 02:46.000 Let's add a true well tested popular language in the world. 02:46.000 --> 02:47.000 Next to it. 02:47.000 --> 02:49.000 Let's add Python. 02:49.000 --> 02:51.000 We always have been neighbors of Python. 02:51.000 --> 02:54.000 Except for this year, this year we got the bigger size. 02:54.000 --> 02:59.000 We've been Python. 02:59.000 --> 03:04.000 So now let's see where a new up and coming fancy new language. 03:04.000 --> 03:07.000 Also probably learn because it's the best words. 03:07.000 --> 03:12.000 Rust is here. 03:12.000 --> 03:15.000 Rust this year got a very big as dev room. 03:15.000 --> 03:16.000 You can get a falls down. 03:16.000 --> 03:23.000 They went from 170 people to 570 people in their dev room. 03:23.000 --> 03:26.000 Well, congratulations, Rust. 03:27.000 --> 03:35.000 However, 30 years ago I'm going to go learn a lot. 03:35.000 --> 03:39.000 Go learn that we have APIs, the ability of APIs, the ability is important. 03:39.000 --> 03:43.000 You can compile your 30-year-old go-go today and it might still work. 03:43.000 --> 03:45.000 It probably still work. 03:45.000 --> 03:47.000 Officially it should work. 03:47.000 --> 03:49.000 We have a strong standard standard library. 03:49.000 --> 03:52.000 We have everything in the standard library, which you want to use in your life, 03:52.000 --> 03:54.000 in the standard library of your language. 03:54.000 --> 03:57.000 No more choose your framework, stories, standard library. 03:57.000 --> 03:59.000 We have fixed dependency management. 03:59.000 --> 04:00.000 Look some years. 04:00.000 --> 04:01.000 We have fixed it. 04:01.000 --> 04:04.000 We even had generics, which was impossible to think about 04:04.000 --> 04:05.000 13 years ago. 04:05.000 --> 04:08.000 And go even learn to try not to track its users. 04:11.000 --> 04:14.000 Which brings me back to go is a more mature language 04:14.000 --> 04:15.000 than it was 13 years ago. 04:15.000 --> 04:19.000 So what has been changed in this new mature language? 04:20.000 --> 04:23.000 Well, first of all, even in a new year old language, 04:23.000 --> 04:26.000 we still have two major changes to the language. 04:26.000 --> 04:30.000 Which is this year that we can now loop over three new types. 04:30.000 --> 04:33.000 Which is a function, which takes a function, which returns volume, 04:33.000 --> 04:37.000 a function, which returns function, which has a value k, 04:37.000 --> 04:39.000 which returns volume, a function, which contains a function, 04:39.000 --> 04:42.000 which contains k, v, and a Boolean. 04:42.000 --> 04:44.000 What is this nonsense? 04:44.000 --> 04:47.000 It's actually quite simple once you know what it means. 04:47.000 --> 04:49.000 Because I talked about this out here. 04:49.000 --> 04:52.000 The slide from last year is a range-run experiment. 04:52.000 --> 04:55.000 To quickly recap during a range-run experiment, 04:55.000 --> 04:58.000 we have a new interface. 04:58.000 --> 05:01.000 And we have these new ones defined as an interface, 05:01.000 --> 05:04.000 which is Internet Sequence, and Internet Sequence 2, 05:04.000 --> 05:06.000 lack of creative naming. 05:08.000 --> 05:09.000 And what can we do with this? 05:09.000 --> 05:11.000 Well, let's say I have my file struct. 05:11.000 --> 05:14.000 And I want to look over all lines in the file. 05:14.000 --> 05:16.000 We're all the old missing a CSV. 05:16.000 --> 05:20.000 And what I will do is I just take a local copy memory of my data. 05:20.000 --> 05:23.000 And I then return a function, which takes a yield function, 05:23.000 --> 05:25.000 which is a with function in the sequence. 05:25.000 --> 05:29.000 And this one takes a slide of byte, I return the Boolean. 05:29.000 --> 05:31.000 In there, I take my data. 05:31.000 --> 05:33.000 I check, okay, is there still data left? 05:33.000 --> 05:36.000 If so, I take it till the next new line. 05:36.000 --> 05:38.000 I return it to this yields function, which go, 05:38.000 --> 05:40.000 what we give in me, and I test against it. 05:40.000 --> 05:42.000 And then I take the rest of the data. 05:42.000 --> 05:44.000 So this complex structure here. 05:44.000 --> 05:46.000 We go take over in that loop. 05:46.000 --> 05:48.000 And now I have my own lines. 05:48.000 --> 05:50.000 And instead of doing those weird cuttricks, 05:50.000 --> 05:53.000 I can now do range file, a lab file, 05:53.000 --> 05:56.000 that lines, and can range over my files. 05:56.000 --> 05:59.000 You can now easily make your own range functions. 05:59.000 --> 06:04.000 The second new language is that we now have generic type aliases. 06:04.000 --> 06:06.000 When I write that last, because like, 06:06.000 --> 06:07.000 ooh, that's complex. 06:07.000 --> 06:09.000 That's kind of the local slides. 06:09.000 --> 06:11.000 Turns out it's just this. 06:11.000 --> 06:13.000 It's just a generic type. 06:13.000 --> 06:16.000 So now allow it to be used in type aliases. 06:16.000 --> 06:18.000 So for example, I have a set here of p, 06:18.000 --> 06:19.000 which is a comparable. 06:19.000 --> 06:21.000 It's just equal to map p Boolean. 06:21.000 --> 06:23.000 Honestly, I thought this already worked. 06:23.000 --> 06:24.000 Okay. 06:24.000 --> 06:26.000 We now have this in go 1204. 06:26.000 --> 06:28.000 Those were my two new changes to the language, 06:28.000 --> 06:30.000 which takes us to tooling. 06:30.000 --> 06:32.000 And I always talk about govets. 06:32.000 --> 06:34.000 Why do I talk about govets? 06:34.000 --> 06:37.000 Because govets is being run when you run tests. 06:37.000 --> 06:39.000 And if you get that you go versioning your CI, 06:39.000 --> 06:41.000 instead of in your test format, 06:41.000 --> 06:43.000 it might be because of a new govets warning. 06:43.000 --> 06:45.000 And you all will know why. 06:45.000 --> 06:47.000 That's why I'm talking about it. 06:47.000 --> 06:49.000 There are many new warnings in govets. 06:49.000 --> 06:53.000 In go 1.23, it will now warn if you use a function of the standard library 06:53.000 --> 06:55.000 that has been added beyond the function. 06:55.000 --> 06:59.000 Beyond the version, you defined in your go module file. 06:59.000 --> 07:01.000 Very handy. 07:01.000 --> 07:03.000 If now we'll also warn again, 07:03.000 --> 07:07.000 if you don't use any parameters in a printf or any f printing function. 07:07.000 --> 07:10.000 For example, here, this is written by an AI. 07:10.000 --> 07:12.000 Many AIs have made this mistake, 07:12.000 --> 07:14.000 and you haven't corrected them. 07:14.000 --> 07:17.000 So gov will now correct this for you. 07:17.000 --> 07:19.000 If you're now also report, 07:19.000 --> 07:22.000 if you build a parameter in for the patch version, 07:22.000 --> 07:24.000 which is probably what you don't want to do. 07:24.000 --> 07:26.000 If you're now also report more issues 07:26.000 --> 07:27.000 in the declarations of tests, 07:27.000 --> 07:30.000 further benchmark and example code. 07:30.000 --> 07:33.000 Let's look forward to go come out. 07:33.000 --> 07:35.000 And I'm introducing a new category, 07:35.000 --> 07:38.000 which is the CI life changes. 07:38.000 --> 07:42.000 This new changes will change your life in the CI forever. 07:44.000 --> 07:48.000 First of all, go to link now support JSON output. 07:48.000 --> 07:51.000 We have go test, minor JSON, go in some minor JSON, 07:51.000 --> 07:53.000 and go build minor JSON. 07:53.000 --> 07:55.000 Yes, JSON output, parsable, logging, 07:55.000 --> 07:58.000 collecting knowledge and elastic search, perfect. 07:58.000 --> 08:00.000 How does this work? 08:00.000 --> 08:02.000 Well, I'm working on this co-fight as decating, 08:02.000 --> 08:04.000 and of course, I'm a coot developer. 08:04.000 --> 08:06.000 I write tests. 08:06.000 --> 08:07.000 So I run my tests. 08:07.000 --> 08:09.000 As we all do, we go test. 08:09.000 --> 08:11.000 And I have here okay, okay, okay. 08:11.000 --> 08:14.000 If I want to cause metrics out of this for my CI, 08:14.000 --> 08:17.000 I can do a magic part on the word okay. 08:17.000 --> 08:19.000 That's going to be very complex. 08:19.000 --> 08:21.000 Or can use this new JSON output, 08:21.000 --> 08:23.000 which will output something like this. 08:23.000 --> 08:26.000 The exact time output, which package has been built, 08:26.000 --> 08:29.000 the exact test has been run on this, 08:29.000 --> 08:31.000 and the output test has just passed. 08:31.000 --> 08:34.000 Okay, the output is still somewhere to do. 08:34.000 --> 08:38.000 At least we now have JSON parsable outputs. 08:38.000 --> 08:42.000 Well, I'm also building my co-fight CTL. 08:42.000 --> 08:44.000 Well, yeah, binary. 08:44.000 --> 08:45.000 So I'm doing this. 08:45.000 --> 08:48.000 And we all know if you do co-ocolon build, 08:48.000 --> 08:50.000 and if you succeed, we don't have outputs. 08:50.000 --> 08:51.000 We're all used to that. 08:51.000 --> 08:54.000 What do you get if you do co-ocolon build, 08:54.000 --> 08:56.000 there's JSON? 08:56.000 --> 09:00.000 Well, your output is, nothing. 09:00.000 --> 09:03.000 Maybe not useful if you want to pass locks. 09:03.000 --> 09:06.000 Still output something. 09:06.000 --> 09:09.000 Another big change is that goes set the binary version 09:09.000 --> 09:12.000 based on your version control system. 09:12.000 --> 09:14.000 Really? No, boss? 09:14.000 --> 09:19.000 Wow. 09:19.000 --> 09:25.000 13 years for this basic feature. 09:25.000 --> 09:27.000 What we're going to do? 09:27.000 --> 09:29.000 Well, let's set the version in your binary, 09:29.000 --> 09:31.000 based on your tag and or commits. 09:31.000 --> 09:34.000 It will add a dirty suffix if you have a committed changes. 09:34.000 --> 09:37.000 Of course, you can disable it with this build flag. 09:37.000 --> 09:39.000 How does this work? 09:39.000 --> 09:41.000 We have all done this before in our lives. 09:41.000 --> 09:43.000 We added a dirty variable somewhere 09:43.000 --> 09:46.000 or gives environment variables of some crazy tricks. 09:46.000 --> 09:48.000 How do we do this now? 09:48.000 --> 09:50.000 We have here my main function, 09:50.000 --> 09:52.000 of which I call the debug package. 09:52.000 --> 09:55.000 And I fetch the read build info. 09:55.000 --> 09:56.000 I handle my error. 09:56.000 --> 09:57.000 I'm a good developer. 09:57.000 --> 09:59.000 And I print out my version number. 09:59.000 --> 10:01.000 So I have this. 10:01.000 --> 10:03.000 So I wrote this program. 10:03.000 --> 10:05.000 It doesn't work from goal run by the way. 10:05.000 --> 10:07.000 Only on goal build. 10:07.000 --> 10:09.000 I do goal build and I run my program. 10:09.000 --> 10:11.000 My output version is empty. 10:11.000 --> 10:13.000 I forgot getting it. 10:13.000 --> 10:15.000 Mark anyway. 10:15.000 --> 10:17.000 What do I do whenever getting it? 10:17.000 --> 10:19.000 I get my version is development. 10:19.000 --> 10:21.000 Oh, I forgot to commit my changes. 10:21.000 --> 10:23.000 I commit my changes. 10:23.000 --> 10:25.000 I write a very good commit message. 10:25.000 --> 10:29.000 And I get the version number with the current build date 10:29.000 --> 10:31.000 and the get shot. 10:31.000 --> 10:33.000 Perfect. 10:33.000 --> 10:35.000 What do I do when I get it? 10:35.000 --> 10:37.000 I get the version. 10:37.000 --> 10:39.000 We try to act. 10:39.000 --> 10:41.000 V0.1.0. 10:41.000 --> 10:43.000 My build now finally without any dirty tricks in your CI 10:43.000 --> 10:45.000 has the build version built in. 10:45.000 --> 10:49.000 There is also the new book goal tool directive. 10:49.000 --> 10:53.000 If you're using terpory libraries for goal generation, 10:53.000 --> 10:55.000 the open API library, you previously had to define 10:55.000 --> 10:57.000 them in the goal model file as an actual dependency, 10:57.000 --> 10:59.000 which is not correct because you're not 10:59.000 --> 11:01.000 depending on that code. 11:01.000 --> 11:03.000 So there is now this dash tool flag. 11:03.000 --> 11:05.000 Right here in the goal gets. 11:05.000 --> 11:07.000 And you can then, 11:07.000 --> 11:09.000 project tool will be saved in the tool 11:09.000 --> 11:11.000 right in the goal module file. 11:11.000 --> 11:13.000 And you can use it with the goal tool commands 11:13.000 --> 11:15.000 and goal goal on general rate will automatically use this. 11:15.000 --> 11:19.000 Very handy for any build dependencies. 11:19.000 --> 11:21.000 Now let's look at the goal command. 11:21.000 --> 11:25.000 What's been new apart from CI lab changes? 11:25.000 --> 11:29.000 There was a big change to goal telemetry. 11:29.000 --> 11:31.000 And if I'm giving this talk at false them, 11:31.000 --> 11:35.000 the community which is about all about open data, transparency, 11:35.000 --> 11:37.000 no tracking, no tracing. 11:37.000 --> 11:39.000 And I'm talking about goal telemetry. 11:39.000 --> 11:41.000 This is controversial. 11:41.000 --> 11:43.000 Oops. 11:43.000 --> 11:45.000 And yes, it's controversial. 11:45.000 --> 11:46.000 I looked at that. 11:46.000 --> 11:47.000 They started back in 2023. 11:47.000 --> 11:49.000 And the first thing I could find on hacker news 11:49.000 --> 11:51.000 at 234 comments. 11:51.000 --> 11:53.000 I read them all. 11:53.000 --> 11:54.000 No, it didn't. 11:54.000 --> 11:55.000 Even that hidden on hacker news, 11:55.000 --> 11:57.000 the register picked it up with the name 11:57.000 --> 11:58.000 subline. 11:58.000 --> 12:01.000 That's devs do not trust mountain view. 12:01.000 --> 12:02.000 Wow. 12:02.000 --> 12:03.000 It's mainstream. 12:03.000 --> 12:05.000 That media picks up on a program in line 12:05.000 --> 12:07.000 which is terrible. 12:07.000 --> 12:11.000 So goal makes telemetry opt in only. 12:11.000 --> 12:13.000 We should of course result to the 12:13.000 --> 12:16.000 contradicting post by goal making my making telemetry opt in. 12:16.000 --> 12:17.000 It wasn't mistake. 12:17.000 --> 12:19.000 And again, theta by hacker news, 12:19.000 --> 12:21.000 300 comments and people were angry again. 12:21.000 --> 12:25.000 So what is magical to say on stage for 25? 12:25.000 --> 12:27.000 First of the clear, 12:27.000 --> 12:29.000 I am not involved with the goating. 12:29.000 --> 12:31.000 I am not working at Google. 12:31.000 --> 12:33.000 I do not have any involvement in this. 12:33.000 --> 12:35.000 So please do not kick me. 12:35.000 --> 12:38.000 The good news is, 12:38.000 --> 12:40.000 it is still opt in. 12:40.000 --> 12:41.000 Phew. 12:41.000 --> 12:43.000 We now have an easy way to opt in. 12:43.000 --> 12:45.000 We just go telemetry on and go telemetry off. 12:45.000 --> 12:46.000 What has been changed? 12:46.000 --> 12:48.000 There is a new design. 12:48.000 --> 12:50.000 Fancy, more efficient. 12:50.000 --> 12:52.000 And it will now automatically, 12:52.000 --> 12:54.000 even if you did not enable, 12:54.000 --> 12:56.000 store data locally. 12:56.000 --> 12:58.000 For, okay, it crashes. 12:58.000 --> 13:00.000 You can use this with telemetry data in about report. 13:00.000 --> 13:02.000 There will be data on your desk. 13:02.000 --> 13:04.000 It will not be sent. 13:04.000 --> 13:05.000 Don't be scared. 13:05.000 --> 13:07.000 That's what is new. 13:07.000 --> 13:09.000 There are also desk technology prompts. 13:09.000 --> 13:12.000 If you use Visual Studio Code and you are in the lucky 2% 13:12.000 --> 13:13.000 which you roll out, 13:13.000 --> 13:15.000 you will now get a friendly notice. 13:15.000 --> 13:17.000 It will only enable telemetry. 13:17.000 --> 13:19.000 Only 100 people enable this. 13:19.000 --> 13:21.000 We need more. 13:21.000 --> 13:23.000 So you now have this bag in prompts. 13:23.000 --> 13:25.000 And why did they put this bag in prompts? 13:25.000 --> 13:27.000 Because the go-to truly believes that telemetry 13:27.000 --> 13:29.000 will drive their stability forward. 13:29.000 --> 13:31.000 This is a quote by the go-to-f team in a blog post about this. 13:31.000 --> 13:33.000 Go-turns 15 this November. 13:33.000 --> 13:36.000 And both the language and the ecosystem continue to grow. 13:36.000 --> 13:39.000 Telemetry will play a critical role in helping go-contributes 13:39.000 --> 13:42.000 move faster and more safely in the right direction. 13:42.000 --> 13:45.000 They truly believe that telemetry will help them. 13:45.000 --> 13:48.000 It is still your own choice to label a lot. 13:48.000 --> 13:50.000 I want if you want to. 13:50.000 --> 13:52.000 You can do so. 13:52.000 --> 13:55.000 That was the go tooling. 13:55.000 --> 13:57.000 Now let's talk about the standard library. 13:57.000 --> 14:00.000 Now usually a few very interesting changes in there. 14:00.000 --> 14:03.000 Last year I introduced these lies in max package. 14:03.000 --> 14:04.000 Thanks to genetics. 14:04.000 --> 14:07.000 Well we now have this either interface. 14:07.000 --> 14:10.000 So we now have helper functions for iterations. 14:10.000 --> 14:13.000 We have all key values in search for it. 14:13.000 --> 14:16.000 For example, I have your slice with three random names. 14:16.000 --> 14:18.000 And I want to look over it backwards. 14:18.000 --> 14:20.000 In the past, I would have to do credit tricks like reversing the 14:20.000 --> 14:23.000 length or reversing everything and doing weird tricks. 14:23.000 --> 14:28.000 I can now just use slices.backwards, names and range over it. 14:28.000 --> 14:31.000 So we have a lot of helper functions thanks to this new 14:31.000 --> 14:32.000 range from experiments. 14:32.000 --> 14:36.000 And as you would suspect, it will return nicely 14:36.000 --> 14:38.000 than in backwards order. 14:38.000 --> 14:42.000 That also means very welcome yay security changes. 14:42.000 --> 14:45.000 The treaty aside for what's removed. 14:45.000 --> 14:46.000 Yay security. 14:46.000 --> 14:49.000 If you had been using it, it is a breaking change for you. 14:49.000 --> 14:52.000 If you have been using it, stop so now. 14:52.000 --> 14:54.000 More security. 14:54.000 --> 14:58.000 We now have supports in 1.24 for the encrypted client hello in 14:58.000 --> 14:59.000 the LATAN checks. 14:59.000 --> 15:01.000 As well as quantum proof key exchanges. 15:01.000 --> 15:03.000 Yay security. 15:03.000 --> 15:04.000 More yay security. 15:04.000 --> 15:06.000 The RSA package will now break. 15:06.000 --> 15:09.000 If you ask the key, less than 10, 24 bits. 15:09.000 --> 15:11.000 Which again, you shouldn't be doing. 15:11.000 --> 15:15.000 If you do so, this is a breaking change and stop doing that. 15:15.000 --> 15:17.000 More cryptoneuse. 15:17.000 --> 15:18.000 HQDF. 15:18.000 --> 15:19.000 TBKF2. 15:19.000 --> 15:22.000 As the HH3 moved from the experiment to critical library to the 15:22.000 --> 15:23.000 stable library. 15:23.000 --> 15:25.000 Yay security. 15:25.000 --> 15:27.000 Also, we now have a new package which 15:27.000 --> 15:32.000 will complement the 503 key exchange with quantum proof key 15:32.000 --> 15:33.000 exchanges. 15:33.000 --> 15:35.000 More yay security. 15:35.000 --> 15:38.000 It is in the OS package. 15:38.000 --> 15:40.000 Are you tired of this classic attack, 15:40.000 --> 15:43.000 where somebody enters dot dot slash dot dot slash dot 15:43.000 --> 15:44.000 slash dot dot slash dot slash dot slash dot slash 15:44.000 --> 15:47.000 slash slash EDC passwords in your application. 15:47.000 --> 15:51.000 We will see is guilty of warranty such an application. 15:51.000 --> 15:53.000 Oh, nobody wants to admit it. 15:53.000 --> 15:55.000 We all have done this. 15:55.000 --> 15:56.000 It's a very easy attack. 15:56.000 --> 15:58.000 It happens on every IoT device. 15:58.000 --> 15:59.000 No worries. 15:59.000 --> 16:01.000 Go got a fixed after 13 years. 16:01.000 --> 16:02.000 You're stuck in 30. 16:02.000 --> 16:04.000 What do they now have? 16:05.000 --> 16:08.000 Is the OS dot open root function. 16:08.000 --> 16:11.000 Which gives you a secure root and own and 16:11.000 --> 16:14.000 doesn't allow any backwards movement into the file 16:14.000 --> 16:15.000 system. 16:15.000 --> 16:16.000 It's even that secure. 16:16.000 --> 16:18.000 It protects you against symbolic links out of 16:18.000 --> 16:19.000 this specific root. 16:19.000 --> 16:22.000 Are you dealing with files that you just upload a user 16:22.000 --> 16:24.000 can access in your go application. 16:24.000 --> 16:27.000 Switch to open root right now. 16:27.000 --> 16:30.000 It's also been a change at the CP package. 16:30.000 --> 16:33.000 Go now support the multi-party CP flag on the 16:33.000 --> 16:35.000 left and enable it by default. 16:35.000 --> 16:37.000 What is multi-party CP? 16:37.000 --> 16:40.000 That's everybody but the network engineer right there. 16:40.000 --> 16:42.000 We implemented. 16:42.000 --> 16:45.000 Well, this is an RFC thing that depends it's simple. 16:45.000 --> 16:48.000 You can now have the TCP flow go over multiple 16:48.000 --> 16:50.000 network interface and multiple parts of the internet. 16:50.000 --> 16:52.000 Why should you care as developer? 16:52.000 --> 16:54.000 Well, mobile network targeting. 16:54.000 --> 16:57.000 Even fallback to mobile network targeting. 16:57.000 --> 17:00.000 Also, you can use multiple parts for better performance. 17:00.000 --> 17:02.000 Also, TCP will stay active over any IP changes. 17:02.000 --> 17:06.000 Oh, it's just super cool. 17:06.000 --> 17:10.000 The JSON package now also has a new omit zero flag, 17:10.000 --> 17:13.000 which is just like omit empty but for the zero value. 17:13.000 --> 17:15.000 I have an example, or I have a transition. 17:15.000 --> 17:17.000 I have a name and if it's empty, I want to 17:17.000 --> 17:19.000 omit it in my JSON module output. 17:19.000 --> 17:20.000 We all do this. 17:20.000 --> 17:23.000 But I also have a UIC code, which is an integer. 17:23.000 --> 17:25.000 An integer cannot be empty. 17:25.000 --> 17:26.000 But it can be zero. 17:26.000 --> 17:28.000 And a UIC code is probably never zero. 17:28.000 --> 17:30.000 So I just do omit zero. 17:31.000 --> 17:34.000 Same thing, I have a date, an opening date of the station right there. 17:34.000 --> 17:39.000 I can now omit zero if the station opens at the gear zero, 17:39.000 --> 17:41.000 which I hope it doesn't. 17:41.000 --> 17:43.000 We're also talking about time. 17:43.000 --> 17:45.000 There is a brick and change in the parts and 17:45.000 --> 17:47.000 parts and location function and go. 17:47.000 --> 17:50.000 If you know, we submitted the time zone that doesn't exist, 17:50.000 --> 17:52.000 it will add or. 17:52.000 --> 17:57.000 More about time, there is also a new thing test 17:57.000 --> 17:59.000 and run experimental package. 17:59.000 --> 18:02.000 This one will run your test in a time bubble. 18:02.000 --> 18:05.000 We'll run your gorethins in a time bubble with a fake 18:05.000 --> 18:08.000 clock, which if you have code at a present time, 18:08.000 --> 18:09.000 it's very handy. 18:09.000 --> 18:11.000 It's still experimental. 18:11.000 --> 18:14.000 We have one new package in the standard library, 18:14.000 --> 18:16.000 which is the unique package. 18:16.000 --> 18:20.000 Which can not recognize its value for faster comparison. 18:20.000 --> 18:23.000 It's a very fancy way if you want to impress your partner 18:23.000 --> 18:25.000 to more way, connect, I learned about 18:25.000 --> 18:28.000 the organization of failures across the comparisons. 18:28.000 --> 18:30.000 Which is actually quite simple, 18:30.000 --> 18:31.000 once you put it into practice. 18:31.000 --> 18:34.000 I have two very complex trucks. 18:34.000 --> 18:36.000 They both contain a number. 18:36.000 --> 18:38.000 And I want to compare those two. 18:38.000 --> 18:40.000 How could I compare those two right now? 18:40.000 --> 18:41.000 Very hard. 18:41.000 --> 18:43.000 Well, with the unique package, 18:43.000 --> 18:45.000 I can just make them unique values. 18:45.000 --> 18:47.000 I do unique that makes that one and unique. 18:47.000 --> 18:48.000 Let's try to make two. 18:48.000 --> 18:49.000 And you can get it. 18:49.000 --> 18:50.000 I can just compare. 18:50.000 --> 18:52.000 If you two equals wheel one, 18:52.000 --> 18:54.000 is the same. 18:54.000 --> 18:58.000 Super simple to compare your very complex trucks. 18:58.000 --> 19:02.000 Now, let's talk about changes to go runtime. 19:02.000 --> 19:05.000 First of all, time the timer and then the ticker 19:05.000 --> 19:07.000 now will be gobbish collect, 19:07.000 --> 19:09.000 even if you didn't cancel it. 19:09.000 --> 19:11.000 If you forgot to cancel your timer, 19:11.000 --> 19:14.000 it will be coverage collected if there are no more references. 19:14.000 --> 19:17.000 Go also, implement Swiss maps. 19:17.000 --> 19:21.000 And gosh, I am an expert of Swiss-related on maps. 19:21.000 --> 19:23.000 Just a wrong kind of Swiss maps. 19:23.000 --> 19:26.000 If you want to try to play your train travel in Switzerland, 19:26.000 --> 19:27.000 yeah, come to me. 19:27.000 --> 19:31.000 But I don't know zero shit about official ghost 19:31.000 --> 19:32.000 with maps. 19:32.000 --> 19:34.000 Coating Brian Board, I'm in the room. 19:34.000 --> 19:38.000 And it will come back at 11.30 to give you an in-depth view 19:38.000 --> 19:40.000 about how Swiss maps are the new amazing thing 19:40.000 --> 19:44.000 and go that makes your applications go faster. 19:44.000 --> 19:46.000 That causes speed. 19:46.000 --> 19:50.000 We have two to three percent faster on the average 19:50.000 --> 19:53.000 go benchmarks, yeah speed. 19:53.000 --> 19:57.000 Most of you thank you Swiss maps and the new unique 19:57.000 --> 20:00.000 implementation. 20:00.000 --> 20:05.000 Now, let's take a look at goports as the last technical subject. 20:05.000 --> 20:10.000 Go 1.23 at six experimental supports for OpenVSD 20:10.000 --> 20:12.000 on the risk five, 64 bit per four. 20:12.000 --> 20:16.000 And one guy right in the back is very excited about this, I see it. 20:16.000 --> 20:21.000 Go 1.25 will require macOS 12, all later. 20:21.000 --> 20:26.000 Go 1.24 has no support for Windows R32 bits, 20:26.000 --> 20:28.000 which is only the Windows R key tablet. 20:28.000 --> 20:31.000 For obvious back which they kind of fix. 20:31.000 --> 20:34.000 It just is broken, don't try it right now. 20:34.000 --> 20:37.000 There also have this on well, we'll change it to WebAssembly. 20:37.000 --> 20:39.000 We now have a WASMAX course function, 20:39.000 --> 20:42.000 which can export you from the function to WebAssembly host. 20:42.000 --> 20:45.000 We also now import allow more data types 20:45.000 --> 20:47.000 to be imported with a WASMAX port. 20:47.000 --> 20:51.000 But again, I go the next part of that is. 20:51.000 --> 20:55.000 At 4pm I will have a talk here by Eduardo is 20:55.000 --> 20:57.000 which is about WebAssembly for GoFresh, 20:57.000 --> 21:02.000 which will take you to all things I just said just to way better. 21:02.000 --> 21:05.000 And there also have these changes to the Linux port in Go. 21:05.000 --> 21:07.000 Wait, what? 21:07.000 --> 21:09.000 I never have changed the Linux port. 21:09.000 --> 21:10.000 Never. 21:10.000 --> 21:14.000 Well, go 1.23 will be the last release of supports. 21:14.000 --> 21:17.000 The Linux kernel 2.6.30 to your later. 21:17.000 --> 21:19.000 The Linux developer is now looking very like, 21:19.000 --> 21:23.000 wow, that's released in 2010. 21:23.000 --> 21:25.000 Yes. 21:25.000 --> 21:30.000 Go support a kernel that's basically older than itself. 21:30.000 --> 21:32.000 Because I've read that. 21:32.000 --> 21:36.000 Go 1.24 will now require 3.20 later, 21:36.000 --> 21:39.000 which was released just in 25. 21:39.000 --> 21:41.000 They deprecate a macOS version 21:41.000 --> 21:43.000 which was released yesterday. 21:43.000 --> 21:45.000 But the next version is still table 21. 21:45.000 --> 21:49.000 We all know where Go runs most, right? 21:49.000 --> 21:52.000 So that has been the new updates to go. 21:52.000 --> 21:55.000 What's there for the future of Go? 21:55.000 --> 21:57.000 I always looked through some design drafts, 21:57.000 --> 21:59.000 which is going to be put in together, 21:59.000 --> 22:01.000 and want to find something interesting. 22:01.000 --> 22:04.000 This year I found one very interesting thing, 22:04.000 --> 22:06.000 which is the concept of memory regions, 22:06.000 --> 22:09.000 which builds on a lot of early expansion 22:10.000 --> 22:14.000 with the goal to reduce the risk of cost of garbage collection, 22:14.000 --> 22:18.000 and return more controlled developers, 22:18.000 --> 22:21.000 just like in Rust. 22:21.000 --> 22:23.000 It's built on an early proposal, 22:23.000 --> 22:25.000 which was the arena package. 22:25.000 --> 22:29.000 What does this thing do in very simple terms? 22:29.000 --> 22:31.000 I have a function, 22:31.000 --> 22:35.000 and I am a developer who works with very big complex prototypes. 22:35.000 --> 22:38.000 My job I work with very big complex prototypes. 22:39.000 --> 22:43.000 And I have my region, and I will parse my very big complex prototypes, 22:43.000 --> 22:46.000 which takes a lot of memory with the very big complex prototypes. 22:46.000 --> 22:48.000 So I am marshallet, 22:48.000 --> 22:53.000 but all the program cares actually just if it errors or not. 22:53.000 --> 22:56.000 Okay, so I will set this top level error, 22:56.000 --> 22:59.000 which is my function here above on my screen. 22:59.000 --> 23:01.000 Done, perfect. 23:01.000 --> 23:03.000 Why is this more efficient now? 23:03.000 --> 23:05.000 Once this feature is released, 23:05.000 --> 23:07.000 once this region functions ends, 23:07.000 --> 23:09.000 my function in region.do, 23:09.000 --> 23:13.000 it immediately removes our memory without having to run garbage collection, 23:13.000 --> 23:15.000 which gives me control over my memory, 23:15.000 --> 23:17.000 and it's way more efficient. 23:17.000 --> 23:20.000 Another example is I define a bunch of new variables, 23:20.000 --> 23:22.000 meaning immediately assigned use, 23:22.000 --> 23:25.000 whatever types I only want to keep the results, 23:25.000 --> 23:27.000 once this function exits, 23:27.000 --> 23:29.000 I now have returned all the memory, 23:29.000 --> 23:32.000 read out to in the complex garbage collection mask. 23:32.000 --> 23:34.000 It just removes the whole region. 23:34.000 --> 23:38.000 It gives us the developers complete control over our memory. 23:38.000 --> 23:39.000 Is this more efficient? 23:39.000 --> 23:41.000 Of course, the go down run with a few tests. 23:41.000 --> 23:44.000 They tested against co-crojDB and SCD, 23:44.000 --> 23:46.000 jabbo databases, 23:46.000 --> 23:48.000 which located a lot of memory. 23:48.000 --> 23:51.000 And they found the 5% improvement in time 23:51.000 --> 23:53.000 in the garbage collector, 23:53.000 --> 23:55.000 and even a 2% to 1% improvement in time, 23:55.000 --> 23:59.000 using allocations spent less than before. 23:59.000 --> 24:01.000 These are still early tests. 24:01.000 --> 24:03.000 This is a really handy feature, 24:03.000 --> 24:07.000 and maybe we see it next year, maybe we will never see it. 24:07.000 --> 24:09.000 If you are interested in this, 24:09.000 --> 24:10.000 this long-lange, 24:10.000 --> 24:13.000 and I will post it on the fourth on website later, 24:13.000 --> 24:16.000 has all information about this. 24:16.000 --> 24:19.000 There was even a very interesting implementation details, 24:19.000 --> 24:21.000 which I don't have time to talk about today, 24:21.000 --> 24:23.000 but it's also pretty interesting. 24:23.000 --> 24:26.000 And at last, I want to talk about my favorite objects, 24:26.000 --> 24:27.000 the co-community. 24:27.000 --> 24:30.000 I haven't met here of all go-beats in the world, 24:30.000 --> 24:32.000 which is used to be more, 24:32.000 --> 24:35.000 but the general meat of community is just way less active 24:35.000 --> 24:36.000 than people in them. 24:36.000 --> 24:38.000 It's just a true fact. 24:38.000 --> 24:40.000 Do we move going go-beats meet-ups, 24:40.000 --> 24:42.000 are still stable around 41 chapters, 24:42.000 --> 24:44.000 with the thing that many of those chapters 24:44.000 --> 24:45.000 are currently inactive, 24:45.000 --> 24:49.000 it's just hard to define which numbers these are. 24:49.000 --> 24:52.000 But I like the frozen community the most. 24:52.000 --> 24:54.000 You all here in this room. 24:54.000 --> 24:57.000 We've been here for 11 years, 24:57.000 --> 24:59.000 and I clicked too fast, 24:59.000 --> 25:01.000 and now the Wi-Fi is loading my slides. 25:01.000 --> 25:04.000 And the time is ticking away. 25:04.000 --> 25:06.000 And so I'm going to talk about how Google slides 25:06.000 --> 25:09.000 is a very amazing product by the people of Google, 25:09.000 --> 25:12.000 which never blocks even on bad Wi-Fi. 25:12.000 --> 25:13.000 There it is. 25:13.000 --> 25:16.000 Yep, 2017, and loading again. 25:16.000 --> 25:23.000 Again, Google slides are an amazing product. 25:23.000 --> 25:29.000 Maybe I should have used an open source solution. 25:29.000 --> 25:31.000 It's small images, maybe. 25:31.000 --> 25:33.000 I don't know. 25:33.000 --> 25:35.000 Small images suck. 25:35.000 --> 25:37.000 Come on, Google. 25:39.000 --> 25:42.000 Wow. 25:42.000 --> 25:43.000 Yes, please. 25:43.000 --> 25:46.000 Show anyway. 25:46.000 --> 25:50.000 Thank you. 25:50.000 --> 25:53.000 Last year, we came from last year for yourself. 25:53.000 --> 25:56.000 I can. 25:56.000 --> 25:59.000 And this year, we have an AI chat box. 25:59.000 --> 26:04.000 Our blockchain support got removed. 26:04.000 --> 26:06.000 And I am pretty nervous. 26:06.000 --> 26:08.000 This amazing schedule, which, by the way, 26:08.000 --> 26:11.000 is 100% billion average. 26:19.000 --> 26:21.000 If any of you happen to be working on a billion, 26:21.000 --> 26:24.000 you can just quickly come give me a donation of a million dollars 26:24.000 --> 26:26.000 to have the code for next year. 26:26.000 --> 26:28.000 So this is our main schedule. 26:28.000 --> 26:30.000 I really had to reject amazing talks. 26:30.000 --> 26:32.000 I wish I had two days. 26:32.000 --> 26:34.000 Well, actually don't, because it's too heavy. 26:34.000 --> 26:37.000 And I made this year, I made five tracks. 26:37.000 --> 26:39.000 The first one is about go. 26:39.000 --> 26:42.000 And now that we're all fresh and not yet drunk, 26:42.000 --> 26:44.000 we have an hour of deep dives. 26:44.000 --> 26:47.000 Very deep that even talks about the inner workings of go. 26:47.000 --> 26:49.000 We will slowly transition out of it. 26:49.000 --> 26:52.000 And I will go to from 13, 13, 16, 26:52.000 --> 26:54.000 we will have talks about using go. 26:54.000 --> 26:57.000 Using go better, using go in different projects. 26:57.000 --> 26:59.000 And from 16, 30 to 18 o'clock, 26:59.000 --> 27:02.000 we will have the one and two tiny go conference. 27:02.000 --> 27:05.000 And the most interesting hour is the last one, 27:05.000 --> 27:08.000 which is all about you, about the lightning talks. 27:08.000 --> 27:10.000 Submit one. 27:10.000 --> 27:13.000 I see you all got a taste of go conferences. 27:13.000 --> 27:15.000 Well, well, that's an interesting go conference to go to. 27:15.000 --> 27:16.000 At the go, definitely force them. 27:16.000 --> 27:18.000 Very interesting, please say. 27:18.000 --> 27:19.000 Quite amazing talks. 27:19.000 --> 27:21.000 If you want to travel to Latin America, 27:21.000 --> 27:24.000 you can do go for go for go in Latin America Brazil in May. 27:24.000 --> 27:27.000 Go for go on Europe for the health in Berlin this year. 27:27.000 --> 27:28.000 In June. 27:28.000 --> 27:32.000 And go for go in Europe this year on August. 27:32.000 --> 27:35.000 Lightening talks. 27:35.000 --> 27:41.000 And again, the last hour of the day is about lightning talks, 27:41.000 --> 27:44.000 which will be eight minutes of talks this year. 27:44.000 --> 27:46.000 And the CFP will be open to five feet, 27:46.000 --> 27:48.000 which is one hour before the talks, 27:48.000 --> 27:51.000 which will quickly notify you if you had been accepted. 27:51.000 --> 27:52.000 How you submit one? 27:52.000 --> 27:54.000 You go to go for the last slash light. 27:54.000 --> 27:56.000 Oh, you scan one of the QR codes in this room. 27:56.000 --> 27:59.000 You have to be at force them to submit one. 27:59.000 --> 28:02.000 Don't submit one if you're watching home from your computer. 28:02.000 --> 28:04.000 Do not do that. 28:04.000 --> 28:07.000 So you can submit a talk for lightning talks 28:07.000 --> 28:10.000 and get eight minutes of this amazing stage here 28:10.000 --> 28:12.000 to prove how awesome you are. 28:12.000 --> 28:17.000 And I can all see that each and every one of you is awesome. 28:18.000 --> 28:20.000 You can follow us on social media. 28:20.000 --> 28:22.000 You can post about us on social media. 28:22.000 --> 28:23.000 And if you do, please do so. 28:23.000 --> 28:25.000 And use the hashtag hashtag for them. 28:25.000 --> 28:27.000 You can follow us on the Fediverse. 28:27.000 --> 28:30.000 And you can also follow us on our blue sky bridge. 28:30.000 --> 28:34.000 You can not follow us on X for very obvious reasons. 28:34.000 --> 28:36.000 That was my talk for today. 28:36.000 --> 28:38.000 Thank you very much for enjoying this. 28:38.000 --> 28:40.000 And I see you all of the fall. 28:40.000 --> 28:41.000 See you next year. 28:47.000 --> 28:50.000 Thanks for watching.