Attending: Marsh G. , Agueda Gama, Jesse, Amber Bennett, Chris Wood, Kevin Swiber, Neal Caidin, Stu Waldron
This document is a work in progress
- Marsh: We are still working on what makes the community resonate
- What is a healthy community? Engagement? Numbers of new members? We need more involvement and people participating. AsyncAPI and GraphQL have done a good job, they are co-evolving with tooling. We took the spec out of Swagger, so we are a standard, which is different from software.
- Kevin: difference between open source community and open standards. Open source software tends to have more iterations. A lot is done via email. Should we bring in more tools?
- Stu: We are still struggling with finding relevance meaning the value to the various types of members. Evident relevance to tool and related solution vendors, limited relevance to developers at large and their management (funding authority). The latter sees benefit to using the spec but not to actively contributing in the ongoing development.
- Neal: thoughts on OAI challenges - perhaps can lead to goals?
- Swagger vs OAS - is the terminology still obfuscating?
- Stu: Not sure if this is a goal, more of a challenge. As an organization, we've talked about efforts of value beyond the spec. There are many issues concerning the creation and publication of APIs to address. Possible solutions stray beyond the spec. While SIGs do have some latitude, do we state as a goal to address issues beyond the spec?
- Neal: First step is putting ideas on paper, being able to share and get feedback from the community
- People carving out time of their busy day. Organizations not allocating time for them to work on OAI/OAS
- Stu: Goes with the value prop, they are making them a greater investment by contributing labor, should have a positive ROI for them
- We haven't been as open as we need to, so conversations happening elsewhere
- Things have improved over the last year. We've opened up our Slack workspace and started a Discord channel. But these might need tending.
- Tooling - dependency on ecosystem tooling inhibits adoption of latest spec version
- Stu: This is a value prop problem again. Value not perceived by vendors that are making the tools. The cost/risk of upgrading apparently not justified with sufficient ROI.
- Value proposition of being an OAI member. Are we not articulating the value proposition well enough, or do we need to improve the value prop for membership?
- Stu: We should be looking at clarifying the value prop for upgrading to 3.1. Is there something we can do to make the path to adoption easier?
- Kevin: Idea for partnering with events - mutually beneficial - We could focus on things like an OpenAPI track versus sponsorship, which would be more useful and cost less money.
- Marsh: We should be shining a spotlight - help draw attention to tracks in conferences. We have had good success doing that in the past.
- Is there a list of API events?? - None for 2023 - Current list is out of date, produced prior to Covid.
- API communities outreach
- Lorinda and team are looking to do some of those activities
- Marsh: We should connect with AsyncAPI (async nature connects well with OpenAPI), and GRPC should be transcoded to OpenAPI
- Chris: Pick a vertical! OpenBanking has real interest and is wedding to APIs
- Marsh: Challenges with gathering data, what do we want to learn? Need a repeatable process.
- Kevin: Invite tooling devs. They should have a say in building the spec.
- Amber: BetterCloud challenge is maintaining other community APIs, staying up to date
- Kevin: Very interested in hearing what Linux Foundation is doing with open standards
- Marsh: Yes, to case studies since incentives drive case studies
- How APIs are being used? Is a huge project, previously considered hiring a firm to do this, problem is how to you get the data, we don’t know tooling, can’t tell - We could do survey but it needs to be more than just a 100 responses What is the adoption rate of each spec?
- Chris: massive undertaking in terms of gathering data from the community. Not specific of what they are trying to find on who uses the OpenAPI spec and how - have a basis for a database
- More notes here
- With the direction to not have a big once a year meeting and look to tag along with other conferences we need OAS in a box. I mean that literally. We talk about soft collateral like handouts or presentations anyone can grab. But we also need easy to use booth material such as backdrops, signage and yes, tchotchkes. Something you can ship to me when I'm off to a travel conference or Amber when she's at an IT conference. There are of course costs and logistics to work out. But if we really want to get more exposure by being present at many conferences, we need more than handouts.