text stringlengths 0 1.82k |
|---|
• Dangers of deleting and re-cloning repositories |
• Benefits of contributing to open source projects, even with limited skill level |
• The "Help Wanted" label on GitHub for finding beginner-friendly contributions |
• The value of learning by doing and getting feedback from others through pull requests |
• The difficulty of understanding complex code, such as the Go raft library |
• The importance of domain knowledge when reading source code |
• Strategies for learning from code, including starting with the Go docs and then reviewing the code |
• The value of community involvement in software development, particularly in the Go community |
• Ways to get involved in a local Go community, including attending meetups and contributing online |
• The importance of volunteering and helping out with events and projects beyond just coding |
• The importance of having multiple people involved in organizing meetups, including a host, sponsor, and organizer. |
• Reaching out for help when needed, rather than trying to do everything alone. |
• Organizing regional conferences, such as the GothamGo and Gopherfest, and the need for more regionals. |
• Starting conferences, with Brian and Erik offering advice and guidance to those interested. |
• The value of community members traveling to participate in and help with meetups outside their local area. |
• GopherJS team thanked for helping with frontend code fear |
• Haxor News: a command line tool to access Hacker News without browser distractions |
• Vim Go: praised as essential tool for Go development and Vim users |
• Kubernetes: touted as game-changing technology for container orchestration |
• Open source community shoutouts, including Patreon support for Vim Go developer Fatih |
**Erik St. Martin:** It's Go Time! A weekly podcast where we discuss interesting topics around the Go programming language, the community and everything in between. If you currently write Go or aspire to, this is the show for you. |
This is episode two, and with us on the show today we have Brian - say hello, Brian. |
**Brian Ketelsen:** Hello Brian. |
**Erik St. Martin:** \[laughs\] We also have Carlisia on the call. |
**Carlisia Thompson:** Hello everybody. |
**Erik St. Martin:** And we have a special guest today, Cory LaNou, who is a developer for InfluxDB, and also highly active in the open source community, and runs I don't know how many open source meetups any more. Cory, why don't you tell everybody hello. |
**Cory LaNou:** Hey, how's it going? |
**Erik St. Martin:** So you want to give everybody maybe a short high-level background on yourself and then we'll kick this show off? |
**Cory LaNou:** Sure. I work at Influx Data, working on the core team for InfluxDB, so that's what I do during the day. I spend a lot of time with the community, as many of you know. As far as the amount of meetups I run, I'm slowly removing myself from the Denver meetup, which is great, but I have a Chicago meetup tha... |
**Erik St. Martin:** Yeah, that's beyond busy. |
**Brian Ketelsen:** That's all? What do you do on the weekends? |
**Cory LaNou:** Well no, I live on a 320-acre estate up here with my family, and my wife has horses, so that means any free time whatsoever that I have, I'm out fixing a fence or chasing horses, and all that kind of good stuff. |
**Erik St. Martin:** So where's the time for the fireworks? |
**Cory LaNou:** Luckily that's seasonal, and that pretty much just dictates every single waking moment of my day once that season starts. |
**Erik St. Martin:** Seasonal? There's always time for fireworks. |
**Cory LaNou:** \[laughs\] I wish everybody believed that. |
**Erik St. Martin:** Have you written anything to launch your fireworks in Go yet? |
**Cory LaNou:** No, I would really like to. I think all the firings systems out there are really expensive and don't work so great, and I don't think it'd be that hard to use a arm device to go ahead and start doing that kind of stuff, so I might have to look into that. |
**Erik St. Martin:** I'm in, I'll help. |
**Brian Ketelsen:** An Arduino and a little bit of power, I think you're good. |
**Cory LaNou:** As long as it can ignite the match we're good. |
**Brian Ketelsen:** Yeah. |
**Erik St. Martin:** Alright, so typically the way we do this show is we'll start out with some news. We'll kind of roundtable it and see what everybody has that might be interesting to talk about, we'll chat about that, and then we'll move into Brian's fun projects that he discovers on GitHub because he downloads all ... |
**Brian Ketelsen:** I think the biggest news recently is the Go 1.6.2 update, although it's a tiny little update from what I can tell. The only thing that really caught my eye in there is the net/http client getting http2 by default. That's big, and I know a lot of people were looking forward to that. The rest of the f... |
**Erik St. Martin:** Yeah, It seemed that way to me too, that the main thing that people would be interested in is the fact that http2 is defaulted now, which I believe was a mistake. It was meant to be default first. |
**Brian Ketelsen:** I think that's correct, I think this fixes that. |
**Erik St. Martin:** One thing that I'm particularly interested in chatting about is GopherChina just completed - was it two days ago, a day ago? |
**Brian Ketelsen:** Yeah, I think so. |
**Erik St. Martin:** So the GopherChina stuff, there seemed like there were some interesting talks there, but one particularly caught my eye, which was Dave Cheney's talk on performance, and I happened to see the slides for that - the videos obviously aren't released yet, but the slides look good. There is a lot of stu... |
**Carlisia Thompson:** For me, I never think about it from the beginning. It's usually when the bottleneck happens, then yeah, let's take a look and see what we can do here. I think it is not the ideal, I think a middle ground should be reached there. I think you should think about it a little bit as you develop things... |
**Erik St. Martin:** It's rough, I think we wanna beat ourselves up when we don't have time to benchmark, but at the end of the day we're also expected to deliver products, right? So if performance isn't an issue for the product at the moment, can we justify the time to do it? So as much as we would like to make everyt... |
**Cory LaNou:** Yeah. For me, I find that, I don't optimize right away or think about performance when I write the code for two reasons. One, because I don't typically have the time, and two because I'm almost certainly wrong every single time when I do get the time to actually profile it. |
**Brian Ketelsen:** Yeah, isn't that really what Dave said in his slides? If you performance is good, stop, don't optimize it. Only optimize when there's a problem. For me, coming from languages that have been much slower than Go, I almost never profile anything unless I find a specific problem. I'm already 30 times fa... |
**Erik St. Martin:** I think I tend to look at my code, just kind of quick-eyeing it. If I can remove an allocation or things along that line that are really obvious at the time, I know that a slice is gonna grow beyond some amount, I'll set it from the beginning; at least a base level to prevent the copying every time... |
**Carlisia Thompson:** My recollection too is that anytime, or most of the time that I did have performance issues, they were related to database queries. I don't even think of a time when I had to go back and rewrite some logic because of performance issues. But database queries definitely, that's usually where I find... |
**Erik St. Martin:** Yes, queries, and interestingly I think some of the issues that I've found too are bugs with goroutines kind of growing unbounded, where they don't properly close out and memory continues to grow because of it. We've had a few of those over the years. |
**Cory LaNou:** We've definitely seen that several times on the Influx Core. We've gotten bug reports, and it's almost always been one of those things where we get a bug report and we look, and "Oh, there's a goroutine that was just unbounded." We never thought about actually shutting it down, and it should have been. ... |
**Erik St. Martin:** Oh, no... |
**Brian Ketelsen:** The penalty for success! |
**Cory LaNou:** Absolutely. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.