text stringlengths 0 1.24k |
|---|
**Steve Francia:** I think there's an interesting thing here, too... We often reduce things to things that we currently have. The question you asked, Mat, was "Is it better to have more dependencies or less?" and there's not an answer to that question. It's really an "It depends." If you have more dependencies, but the... |
So I think really the question is not "How big is the dependency tree?" but "What is the quality of your dependency tree?" And are we really evaluating -- anytime you import a dependency, it really becomes a part of your project, and I don't think we really think about that as much as we should... That as soon as you d... |
So I don't think quantity is necessarily the right measurement of that. I think quality is really the right measurement, and then multiplied by quantity. |
**Jon Calhoun:** That's a good point, because the standard library really is just a set of third-party libraries, it's just one we know that is maintained at a very high standard... Whereas anything on GitHub or whatever - we don't know really know what their standards of quality are... So it's just kind of a gamble at... |
**Mat Ryer:** Yeah. This also helps package developers, if we think about what criteria people are gonna use, what they're gonna look for in packages... If someone out there wants to roll their own package for something, then there's kind of now a nice little -- I mean, Julie, your talk is great for people that also wa... |
**Steve Francia:** I'll take that one. I'll start by saying, if a package solves the problem really well, then there isn't much of a need for another package. The standard library is a great model. When I first started learning Go, I looked at that as the pillar of excellence, and I tried to strive for that in the pack... |
So if a package is stable and well-tested, and does the job well, then we don't have a need for an alternative. We need an alternative when there might be a fundamental shift in design philosophy, or there's things that it doesn't accomplish, and that's a very natural thing that happens. We have a similar debate about ... |
There's a big political and philosophical debate around that, but the reality is that over time we've seen that startups do emerge, and they're there to fill gaps, and the larger a company is, and I will translate to packages or libraries - the more established it is, the less it has a chance to migrate and move, and i... |
**Jon Calhoun:** I think the JSON package is a good example of this, where the one in the standard library is great, but there have been ones that emerged that sort of solved a slightly different problem. If you don't wanna build a struct to get something that's like six levels nested, there's a couple packages out the... |
And even if you look at front-end frameworks for JavaScript, you'd think at some point one of them would have just won out and everybody would have stopped... But they all solve different goals, and if something comes along and it's unique enough and solves a different enough problem, I think we see that that tends to ... |
**Break:** \[19:59\] |
**Mat Ryer:** The Context package, when that happened - that was kind of one of those shifts you talked about, Steve, where suddenly now people expected to be able to cancel things that they couldn't cancel before... For example like copying. Is there a context-aware copy operation in the standard library? If you do io... |
**Steve Francia:** I am not aware of it in the standard library. |
**Mat Ryer:** So use cases like that are -- there's still plenty of opportunity for anyone really to contribute, that's the thing... So I'd hate for people to be put off because standards are going up; that doesn't mean you can't contribute something... And especially like -- Steve, your point about "find the thing tha... |
**Jon Calhoun:** One of the goals you guys had mentioned was helping companies understand what other people are doing with Go, and I guess learning from that, deciding whether or not to adopt it. I guess this is something that a lot of us probably don't see; at least me, because while I want people do adopt Go, I don't... |
**Steve Francia:** We've spent a lot of time meeting with different companies, trying to understand what their needs are and what they're struggling with. I'll also say, one of our big goals as the Go team is adoption. We want Go to be used as widely as possible. And as part of the adoption journey, which we're all fam... |
We really got lots of adoption on startups and hobbyist very early, and so we recognize that the enterprise is important to get Go adoption to really fulfill its potential. As part of that we're doing lots of research, and we've been talking to lots of companies, from all different walks - from retailers, to banks, to ... |
And through those conversations we've heard two things that almost every company asked us, two questions almost every company asked. These are companies that are either thinking about using Go, or they've done a prototype with Go, or they've adopted it for some projects... So the two questions that everybody asked was ... |
So a lot of those stories stop there. But some of those stories continued, and we had people that felt confident enough that it was a good decision that they kind of went rogue, and they built a prototype in Go, and it ended up being phenomenally successful. |
So those pioneers in those companies really provide the things that they wanted before, which were these stories... So we're really happy that we're able to tell some of these stories. We've been working with these companies for many months to be able to capture these stories that they shared with us, and make them pub... |
It wasn't just in these internal meetings, we also did different surveys, and this was one of the top things that people asked for in our surveys. Now, if you're coming from a small company or you might be an individual consultant, this probably sounds foreign to you. But if you're working at a big company... You know,... |
**Mat Ryer:** Yeah. It's interesting, if you go to Go.dev, there are a lot of logos, but they're not just -- sometimes you see these on websites and you feel like they're sort of brags, or something... But these are actually -- you can click these and go and read about the actual ways in which people have used it, and ... |
**Carmen Andoh:** You sure can. |
**Mat Ryer:** You can, you can. |
**Steve Francia:** You definitely can. |
**Carmen Andoh:** Yup. |
**Steve Francia:** I think they're very relatable stories, and I think they're very inspiring. As you click through and read these articles and these different stories... I remember back -- I first adopted Go when I was working at MongoDB, very early on. And there was not a lot of these stories that existed yet, but in... |
These stories shared a similar thing. If you've been using Go a while, read them; it'll remind you why. It'll take you back to that moment when you had the first a-ha and fell in love with the language. |
**Carmen Andoh:** But the other thing that I wanted to point out about these case studies that I think are really useful is when I learn how to code, I am not taught how to influence my manager, or influence my higher-ups on choosing Go. That's just not something that they teach you in school... And this is just the pe... |
I think so many people, when I had been in contact with them over the years, like "How can I convince my manager?" or "How can I convince my CTO?" or whoever in the organization, to use Go. Certainly, there's the rogue tactic that Steve mentioned, but this now doesn't -- you don't have to worry about going rogue, or ta... |
And what's great about these is that some of them are more detailed than others, and they can give you a blueprint for how you might wanna do it, to certain extents... Whether you wanna go whole hog with Go, or whether you wanna maybe just instrument some of your observability toolset with Go, or you wanna do some of y... |
So I really, really liked that, and I tell people now that certainly learning how to influence is a vital skill that they don't teach in school, but this is super-helpful. |
**Mat Ryer:** Yeah, I spoke to somebody at the last London Go meetup who -- they were reluctant, because learning a new language sounds like quite a bit thing, especially for people that maybe only know one language; that was this case. But learning Go, especially if you are already familiar with the kind of C base lan... |
And I also like this idea of using it to solve a real problem you have, even if you're not 100% certain of what it's gonna turn out like. When you learn the context that you have, when you're trying to learn something, if you've got context around a real problem you're trying to solve, it really focuses the mind. You d... |
**Carmen Andoh:** And to add further to that context, I also like the idea -- when we talk about adoption, there are different mindsets, depending on where you're at within your company and where that company is at. Sometimes what I often heard was "I don't wanna just learn about theory" or "I wanna see what it looks l... |
Some people like to just go around and play and find what works for them, but others just need to see what it looks like in practice, and see how that matured over time. I think the case studies is one of my favorite pieces. Well, I like the whole site, but I really sing the case studies praises, for lots of reasons. |
**Steve Francia:** Moving forward with these case studies - we launched with a handful of case studies and articles that were published on external websites as well, and we're excited to tell these stories more; to tell additional stories, to tell deeper stories, and we're hoping that some of the listeners today are fr... |
One is you don't need us to tell the story. A lot of these companies, like Capital One, published several stories on their own blog about their journey to using Go. And we link to them here. So feel free to tell your own story; we'd love to link to it and give more exposure to it. Also, we're happy to meet with you and... |
The most important thing - that's an anonymous button, so if you want us to actually get a hold of you, you have to put some identifying information in there, so that we have a way to reach back out to you, or else \[unintelligible 00:35:24.27\] Which we've had a few companies do. "Oh, we're really interested in the ca... |
**Mat Ryer:** Okay... \[laughs\] So put your name and email in, or something. |
**Steve Francia:** Do it someway. |
**Julie Qiu:** And also, if you wanna just send us a non-confidential email, you can feel free to email go-discovery-feedback@google.com. In case you can't remember all of this, we have an About page on Go.dev. Go.dev/about has all of this information. |
**Mat Ryer:** Thank you. We'll also put some information in the show notes, too. |
Will you have videos on there, and talks, and things from conferences, do you think? |
**Steve Francia:** As we said in our blog post when we launched this - it was really us getting what we felt was like a minimum viable product out to the community, as early as possible. You'll notice on the website we say it's for the Go ecosystem, by the Go ecosystem, but it doesn't have a ton of community resources ... |
**Mat Ryer:** There's an Event section on the homepage, isn't there? |
**Steve Francia:** Yeah. And it only gives you those three events... And I think they're only meetups. It's not conferences yet. So it's a start, and we're really happy that it's there, but we have so many meetups around the world that -- if you look on it today, it says January 11th, it has three meetups, but there ar... |
**Mat Ryer:** Yeah, that's great. That's gonna be really great to have that, because it's difficult even for those of us that have been in the community for a while to know what's going on. I think it's not just gonna be a good resource for new people; I think for everybody. |
**Steve Francia:** It's also partly solving the problem that kind of crept up on us... If we look back two years, there was -- I don't remember the exact number, but around a dozen (maybe a little more) Go conferences. And this year, in 2020, there's scheduled to be over 30 conferences throughout the year. That's three... |
We're not gonna have that delivered in the next month or two, but it is on our roadmap and it's things that we're intending to do. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.