text
stringlengths
0
1.49k
**Liz Rice:** \[19:56\] Yeah. I should really understand Flannel and Calico better than I currently do, so actually that's a really good idea. I should probably dig into those more. There's so many things, there's a world of things out there to learn about.
**Erik St. Martin:** Yeah. Finding the time for all of them is the challenge.
**Liz Rice:** Yeah.
**Brian Ketelsen:** That's exactly the problem, finding the time. It's impossible, there's so many things to learn.
**Erik St. Martin:** I think also, another cool topic for people would be understanding Linux schedulers. I don't know how you would implement that in Go code, but I think that that's something that... I find having discussions with people too, that it's like a dark magic that just occurs; you're like 'Run this thing',...
**Liz Rice:** Yeah, and I think it's quite widespread that people have misunderstandings about threading and concurrency and— well, there's been some amazing talks about concurrency and books about concurrency, and they try and explain some of those things. But it always used to bug me when people would say 'Oh, I am g...
**Erik St. Martin:** Yeah. Then you also have contending for resources, right? Like if both of them need access to the same resource, one can be blocking waiting for another... And it actually reminds me a little bit of the reverse of that, when hyper-threading came out and multi-CPU motherboards, and then multi-core.....
**Liz Rice:** Yeah. It might. \[laughs\]
**Erik St. Martin:** It just depends on the way it was written. No doubt you could run more than one thing at one time if you had two completely isolated processes, but people just automatically assume that some CPU hog program that they use was gonna be magically faster because they have more cores.
**Liz Rice:** Yeah.
**Erik St. Martin:** And that's all part of our growth, right? I think that you don't know what you don't know.
**Liz Rice:** Definitely. And the other thing that I always feel is in the same category is locking. You see so many people who throw locks at code without necessarily really having a strategy for it.
**Brian Ketelsen:** That's my coding plan - add a mutex.
**Liz Rice:** Yeah. \[laughs\] What can possibly go wrong? Let's throw another lock at it.
**Erik St. Martin:** So do you have somewhere, a poll or something where you're having people submit suggestions? Because we could definitely link to it in the show notes and have people suggest topic ideas.
**Liz Rice:** Oh, that's a really good idea. Yeah, I could have a poll, I could have... Yeah, I will set something up so that in the show notes we can have a poll. Great idea.
**Erik St. Martin:** So I think it would be really interesting, especially because... I think a lot of us came from dynamic languages and things like that, and Go is a lot closer to C in the sense of being a systems language, so I think there's a lot of people learning a lot of these things that I think a lot of the pe...
\[24:02\] And there's a much larger group of people probably migrating to Go that this is all new information for. So I think it's really valuable that people like you and Julia are trying to help educate people on these things in ways that they can easily understand.
**Liz Rice:** Hopefully, yes.
**Erik St. Martin:** Kernel man pages are not the best way to learn stuff. \[laughter\] They are there, in some cases it's the only option, but some of those things are extremely hard to learn on your own unless you are already familiar with the domain, or really wanna do a lot of digging and playing and learning throu...
**Liz Rice:** Yeah. Which again, takes time.
**Brian Ketelsen:** So, you've done some Alexa apps, haven't you?
**Liz Rice:** Oh, yeah. So over New Year I was at a house party with some friends over a few days, and somebody had an Alexa... What do you call them? Echo. And I ended up to hacking together the beginnings of a skill, and I wrote about it on Medium and then that has actually ended up being [one of my most popular post...
**Brian Ketelsen:** So...
**Carlisia Thompson:** So, I'm curious... Go ahead, Brian.
**Brian Ketelsen:** I was just going to ask what the app did and then, as a follow-up question, did they invite you back afterwards?
**Liz Rice:** Yeah, it's a group of friends, we've always had New Year together for the last many years. And now a lot of people have got kids as well, so the group's got bigger and bigger. And my skill, the idea... What I thought would be really fun would be I could have a list of the names, and particularly for the k...
**Brian Ketelsen:** Yeah, a little bit creepy. That's cool.
**Carlisia Thompson:** Yeah, I was going to ask if there was any particular reason why you didn't choose Go. And nothing against you having them done that in another language, but I'm more curious to know if in the case you did it \[unintelligible 00:27:31.11\] in Go for this project and Go wasn't up to par?
**Liz Rice:** So, I don't know if this is still the case, but Amazon, and certainly at the time, supported Node and Python, and I think maybe one other. Certainly, those two, of which Python I was by far the most familiar with, so that was the natural choice to hack something together.
\[28:02\] But now there are some schemes and some libraries and tooling out there to let you implement the real function in whatever language you like. There's one called [Apex](https://github.com/apex/apex) that I've heard is really good to let you write your functions in Go, and then you've got this little scheme thi...
**Brian Ketelsen:** Turtles all the way down.
**Liz Rice:** Yeah. \[laughs\]
**Erik St. Martin:** So in our email that we send out, where we talk about the things you're interested in working on and all that good stuff, we were talking about a couple of tools that you've written. One of them was Kubernetes Security Benchmarking.
**Liz Rice:** Yeah, [kube-bench](https://github.com/aquasecurity/kube-bench). Yeah. So, I talk there about that.
**Erik St. Martin:** Yeah, and I'm really interested in that. I know Brian and I are big [Kubernetes](https://kubernetes.io/) fans. Carlisia, is [Fastly](https://www.fastly.com/) using Kubernetes at all?
**Carlisia Thompson:** Not that I know of. At least the work that I do doesn't use any Kubernetes. Docker, yes, Kubernetes no. But in some other parts of the company maybe, I don't know.
**Erik St. Martin:** See, you make the names sound so much better. Say Kubernetes again.
**Carlisia Thompson:** Kubernetes.
**Erik St. Martin:** Ah, see, the first time I think you said Kubernetee, you had a nice inflection on it and it sounded...
**Carlisia Thompson:** I think I'm mispronouncing a letter or two the first time. It sounded exotic to you...
**Erik St. Martin:** It did.
**Carlisia Thompson:** ...but it was just wrong. \[laughter\] The second time I pronounced it right.
**Erik St. Martin:** And I am gonna go with the first time was right. I liked that, it sounds more elegant.
**Carlisia Thompson:** Whenever you want me to say it, I'll say it.
**Liz Rice:** I think I flip between Kubernetes and Kuberneetees, I'm like really haven't... I really don't know which is right. It's a bit like schedule and schedule - I genuinely don't know which of those two is the right pronunciation.
**Carlisia Thompson:** I don't think it matters in the end. Let's come up with more different ways of pronouncing it. \[laughter\]
**Liz Rice:** Anyway, I was gonna talk about kube-bench, wasn't I?
**Carlisia Thompson:** Yeah.
**Liz Rice:** Yes, so there is an organization called the [Centre for Internet Security](https://www.cisecurity.org/), and they write guidelines they call benchmarks for how to configure software to implement best practices for security. And they've got a -- I guess over the last two or three months they released a ben...