text stringlengths 0 1.82k |
|---|
**Carlisia Thompson:** Let's not end this show... Can we just keep going? |
**Brian Ketelsen:** Let's just keep going, just let it roll. |
**Erik St. Martin:** Unfortunately it has to end sometime. I think we would all get hungry. |
**Brian Ketelsen:** Well, soon my kids will be knocking on the door. "Feed us, feed us, feed us!" |
**Erik St. Martin:** Alright. |
**Brian Ketelsen:** Well, it's been a very awesome show. Thanks so much for coming on, Katrina. |
**Katrina Owen:** Thanks so much for inviting me, it's been a lot of fun. |
**Brian Ketelsen:** I think we've really covered a wide gamut of things on this show. That's atypical, but we've covered a lot of territory in this show. |
**Erik St. Martin:** Choosing the headline for the episode will be fun, like which things to mention. But yeah, definitely thank you for coming on, this has been a lot of fun. |
**Katrina Owen:** Thanks so much for having me. |
**Erik St. Martin:** And thanks to all the listeners and everybody who's participating live in the Slack channel. For next week, if anybody else wants to participate, we're on the Gophers Slack in GoTime FM. We are @GoTimeFM on Twitter, we are GoTime.fm online if you want to subscribe if you haven't already subscribed.... |
**Carlisia Thompson:** Thank you, goodbye! |
**Brian Ketelsen:** Goodbye! |
**Katrina Owen:** Thanks, bye-bye! |
• Introduction to Go Time podcast and its focus on Go programming language |
• Guest introduction: Brian Kettleson, Carlisa Campos, and Rafael Simon (creator of Goa framework) |
• Background on Rafael Simon's experience with RedScale cloud management platform and designing APIs |
• Challenges in designing consistent and standards-based APIs for distributed services |
• Creation of Praxis framework in Ruby to address these challenges |
• Development of Goa framework in Go to generate HTTP APIs from a design language (DSL) |
• Benefits of using code generation and DSL in API design, thanks to the simplicity and power of the Go language |
• Design of API and data structures |
• Code generation using GoaGen tool |
• Automatic validation and binding of request body |
• Generation of client package and client tool |
• Consistent and efficient development process |
• Documentation of API through Swagger and JSON schema |
• Sharing design with others through documentation |
• Efficient way to develop and consume APIs |
• The speaker finds the generated code in Goa to look like handwritten Go code |
• Design goal was for generated code to be idiomatic and not feel like it was generated |
• Speaker worried about reception from Go community due to their preferences for geometric code |
• DSL (Domain Specific Language) itself has received some comments trying to make it look more like Go, but speaker defends its design as a separate language implemented in Go |
• The DSL (Domain Specific Language) in Goa is not intended to be idiomatic Go, but rather a tool for generating clients in various languages, including JavaScript. |
• The language of the DSL should be agnostic and independent of any target it generates. |
• The approachability and readability of the DSL in Goa are notable features that distinguish it from other DSLs. |
• The speaker mentions that literal data structures were initially used to define a DSL, but it was later replaced with an anonymous function as an argument. |
• The speaker credits the Goa community for its adoption and contribution to the project's development. |
• The speaker expresses surprise at the rapid growth of the Goa community and its impact on the project. |
• The speaker talks about a blog post they wrote that discussed their personal research project, which eventually became part of the Free Software Friday movement. |
• Discussion of a person's enthusiasm for the Goa framework |
• Mention of proposing to a programmer who created Goa on Twitter |
• Apology for making someone uncomfortable with a tweet |
• Acknowledgment of support from contributors and developers |
• Explanation of a refactoring process in Goa to allow pluggable plugins |
• Description of a plugin called Gorma that allows model definition in DSL |
• Development process for Goa |
• Gorma plugin and its importance |
• Code generation and maintenance |
• Surprising contributions to Goa |
• Use cases for storing requests in a database |
• DSL (Domain Specific Language) discussion limitations due to audio format |
• Maintaining generated code is not necessary |
• Generated code and user code should never mix |
• A clear interface (such as a Go interface) should exist between the two |
• Regenerating code should have no side effects on existing code |
• Changing design or adding new features should only require regeneration of code, without affecting existing code |
• Auto-generated code ownership |
• Testing generated code |
• Code generation lifecycle management |
• Separation of auto-generated and custom code |
• Integration testing vs. functional testing |
• Scaffolding code for bootstrapping services |
• Code ownership and maintenance |
• The difference between Go's generated code (low-level handlers) and the code written by the developer (controllers) |
• Similarities between Go and Rails |
• The potential for Go to be used as an alternative to Rails for backend or API development |
• The speaker discusses the goal of the Goa project to keep things simple and achieve a balance between simplicity and practicality. |
• The speaker compares Rails to Goa, stating that while Rails is easy to get going with, it can be overly complex due to numerous plugins and gems. |
• The speaker notes that using Goa provides direct control over what's happening in the application and makes everything simpler. |
• Goal of simplifying things and hiding complexity |
• Importance of simplicity in user experience and developer experience |
• Anecdote about Raphael, the "godfather" who prioritizes simplicity in DSL design |
• Principle of keeping complexity hidden from users and developers |
• Need for a tool to be approachable for developers of all levels |
• The importance of understanding and leveraging a tool to its full potential, without requiring extensive knowledge of how it works. |
• Swagger for creating API specifications and generating Swagger UI for free. |
• Inspiration from JSON schema for designing abstractions in the language. |
• Easy mapping between the design language and Swagger's representation of path objects. |
• Goa is a tool that can generate Swagger definitions from DSL (Domain Specific Language) |
• The process could be reversed, where Swagger definition is used to generate DSL |
• This could potentially create an endless loop of generating and regenerating the same specifications, making it Turing complete |
• A project or add-on that combines these two features would be interesting to see how the Swagger evolves over time |
• Discussing API representation using views |
• Defining media types and their fields once |
• Creating multiple views to represent a single resource in different ways |
• Arbitrary field definition for each view |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.