text stringlengths 0 1.82k |
|---|
**Raphaël Simon:** Thank you. It has been great. Thank you very much. |
**Erik St. Martin:** We have the Godfather of code generation on the show. \[laughter\] |
**Raphaël Simon:** This has been really great. Thank you for the opportunity, I really appreciate it. |
**Erik St. Martin:** And we'll have links to everything we've talked about in the show notes, or if you happen to be following us on Twitter @GoTimeFM, most of the stuff should be linked there, or the Slack Channel, the Gopher Slack or Goodtime FM, there as well. So I think that is about it and with that, I guess, we'l... |
**Brian Ketelsen:** Awesome, thanks everybody. |
**Carlisia Thompson:** Thank you, goodbye. |
**Raphaël Simon:** Bye. |
• Introduction to Dave Cheney and his background |
• How Dave got started with writing about Go on his blog and its impact |
• Dave's contributions to the Go project, including hosting Arm builds and proposing language changes |
• Dave's experiences as a developer at Canonical and traveling for Go conferences |
• Gratitude from the hosts and guests for Dave's efforts in promoting Go and building community |
• Discussion on the importance of design principles for long-term maintainability of Go code |
• Critique of the focus on "good code" as being subjective and not actionable |
• Proposal for a more objective approach to design using guidelines rather than rules |
• Value of discussing design at an abstract level, focusing on goals rather than specific patterns or solutions |
• Maturity model for Go, including potential growth phases and lessons from other programming languages |
• Discussion about the Gang of Four book and its influence on software design patterns |
• Limited number of fundamental software design patterns, with 30-odd being considered sufficient for most scenarios |
• Comparison to laws of nature, implying a finite set of underlying principles |
• Debate on algorithmic complexity and trade-offs (time vs. space) |
• Meta-language for discussing algorithms (big O notation, time and complexity) |
• Design decisions in software development, including coupling, lookups, and package layout |
• Critique of the standard library as an example of inconsistent design |
• Evolving knowledge and code design over time |
• Discussion on error handling and a new approach being advocated by Dave Cheney |
• Evolution of functional options in Go |
• Error handling design: fail-fast, fail early |
• Importance of decoupling and simplicity in error handling |
• Use of interfaces for modular design and loose coupling |
• Considerations for retrying operations and idempotency |
• Information hiding and encoding extra information into errors |
• Sticking additional context to errors using fmt.Errorf |
• The standard library in Go has a pattern of returning errors with descriptive messages |
• Checking for specific error values can be problematic and lead to issues with stacking errors |
• A proposed solution is to give errors a method that allows getting the underlying error and undoing stacking |
• Using sentinel error values based on type can become problematic |
• Tagged logs only help in log messages, not when passed back up the stack |
• Handling errors once at each level of the call stack can lead to excessive logging |
• Proposed solution is to return the error with annotations to the caller and handle it there |
• Structured logging is seen as unnecessary for operator use cases, but useful for developers during development. |
• Different personas for logging (developers vs operators) |
• Structured logging and its limitations |
• Use cases for counters and metrics instead of logs |
• Distributed tracing and request IDs |
• Ordered logs and their importance |
• Instrumentation and monitoring versus logging |
• Trade-offs between logging, performance, and storage costs |
• Go's approach to error handling is a key factor in its success for writing server software |
• Error handling in Go does not use exceptions but rather requires explicit checks for errors |
• The use of error handling in Go encourages developers to think about potential failures and handle them at the point of failure |
• The "errors" package can simplify error handling by allowing returns of error values with nil indicating no error |
• The verbose nature of error handling in Go is a design decision that prioritizes reliability over convenience |
• There are parallels between designing interfaces in Go and error handling, both require thinking about potential failures and handling them at the point of failure |
• A lack of clear guidance on when to use channels and how to structure concurrent code is an open question in the Go community |
• There is a growing interest in discussing language design and best practices for writing successful Go code. |
• The hosts discuss their time constraints and decide to skip over certain topics |
• Brian Ketelsen talks about his experience with rsync, a UNIX tool for synchronizing files |
• Dave Cheney mentions the connection between Samba and rsync, and recommends pt (Platinum Searcher) as a faster search alternative to Ack or AG |
• Carlisia Thompson shares her experience using Sourcegraph, which she finds much faster than grep |
• Erik St. Martin talks about Asciidoctor, a tool for generating documentation with features like table of contents and source code highlighting |
• The hosts also discuss their personal preferences for text editors and search tools |
• Show submission and guest suggestions via GitHub |
• Wrap-up and goodbye |
**Erik St. Martin:** We are back for another episode of GoTime. Today on the show we have myself, Erik St. Martin, Brian Ketelsen is also on the show today... |
**Brian Ketelsen:** Hello! |
**Erik St. Martin:** And Carlisia Campos. |
**Carlisia Thompson:** Hi, everybody! |
**Erik St. Martin:** And our special guest today, although he needs very little introduction, is Dave Cheney. Say hello, Dave. Why don't you give everybody a brief introduction? |
**Dave Cheney:** Hello there! My name is David, I am a Go enthusiast from Sydney, Australia. I've been involved in Go for many years, and I love the opportunity to be involved with the language and get involved so early with something I'm so passionate about. Hello, everyone! |
**Brian Ketelsen:** It's kind of amusing having you introduce yourself, while one of the things we found at GopherCon just a month or so ago was that nearly everyone that we talked to learned Go from your blog. Out of the 1,500 people that were there, 1,499 of them listed your blog as one of the resources they used to ... |
**Dave Cheney:** Yeah, it was great to meet so many people at GopherCon. It was really touching the way that everyone was like, "Oh, I love to read your blog!" It was really touching. Five years ago, I never started out to be an author, a blogger or a public speaker. Just like every good engineer, you think "Oh, I've r... |
**Brian Ketelsen:** A question on that - did Go encourage you to find your voice, or did just becoming a better engineer encourage you to find that voice? |
**Dave Cheney:** It's kind of like the question, "Does the tail wag the dog, or does the dog wag the tail?" I joined Canonical about four, five years ago, and we were encouraged to communicate on IRC all the time, so it was a great opportunity to moonlight in the IRC channel for Go. At the time Go 1.0 was just barely o... |
I was so excited about it, I wanted to tell people "Hey, you can run Go on this kind of hacked up NAS." That was one of the first things I wrote on my blog. Then it just went from there. |
**Erik St. Martin:** In addition to all the blog writing, you've also contributed quite a bit to Go. |
**Dave Cheney:** Yes. As I said in my introduction, it was kind of like this opportunity to get involved at the ground floor. All you needed to be involved was spare time, and it's just snowballed since then. You know, that kind of "see a need, fill a need" kind of thing. |
**Brian Ketelsen:** \[04:01\] For years, you've hosted the unofficial builds for the Arm platforms on your website, and I think it was not too far in the past when Go started making the Build Dashboard and actually having builders that were hosted in-house. But I think for the longest time, all of us that used Arm boar... |
**Dave Cheney:** I really think Arm is really special. I know Intel rules most of the server world, but Arm has a really special place. It's so simple, it's such a beautifully clean instruction set, rather than the Intel mess. Just like I appreciate Go for being simple and minimal, I really appreciated Arm for being al... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.