Notes from Application Creator

# Raw Notes from Application Creator Call **Wednesday, 6 December 2023** ## ALR Introduction and Framing - [ ] Ecosystem development remains the principal goal for TBD - with both partners and our OSS community - [ ] And it turns out - Developer Experience is the best avenue I know for building affinity and passion for any engineering project. - [ ] A critical part of that is the onboarding experience, again, for both our white-globe treatment as well as for folks to self-serve in community efforts. - [ ] Several of us have been talking about or prototyping early efforts at a bootstrapping project, a Creator application, going on a year now. Mic, Frank, me, Leo - [ ] And with the emergence of more defined use cases in both tbDEX and Web5, I think we’re in a good position now to coordinate those. - [ ] It’s an avenue for DevRel to publish working examples ## Team Discussion - [ ] Mic - the idea of having one place to go to get started might make sense - less to maintain - [ ] Things that create-react or start.spring.io are good, we can even use AI to spit out a whole working app or choose which one to do. People may not know what they’re coming in for; or we may not know - an AI system can train that. - [ ] Other approach is to go with standard polished experiences - [ ] Leo - could do via content catalog, ie. Like Vercel - [ ] https://vercel.com/templates - [ ] Allows you to create Next.js applications from a set of templates. - [ ] The cool thing is that they’re created both by Vercel and the community, which fosters an environment of collaboration - [ ] ALR: so long as they’re tested, they can be supported - [ ] And this could feed the AI model - [ ] Another cool thing about the template is that this is the strategy they use to promote their main framework, NextJS - like we could for Web5 and tbDEX - [ ] Becca - yes we’ll be doing point to point integrations for tbDEX for awhile but as we pivot, can we identify who the target audiences are for tbDEX? - [ ] Mic - Yes, might be a timing thing - [ ] Moe - For now, out the gate, less helpful on the PFI side of the fence. If you scaffolded a wallet codebase with all necessary dependencies and a small amount of code to make a wallet and send a tbDEX message, this could be the starting point for a use case. Probably the most appealing to the broadest audience. - [ ] Wallet devs would be - [ ] Angie - A starting point just for the wallet work with the SDK may be a great start, ready for them to integrate with their existing application. They want to know: “how do I speak tbDEX”? - [ ] Moe - like the wallet file we coded for folks in the workshop - [ ] Nick - DevRel does a great job at getting interested audiences - I think having these create apps ready to go reinforces their messaging - [ ] Ace - Been working on this a bit too, would like to paint a picture of a CLI that’s interactive, and then can be evolved as we go. Approach I’d recommend now is a CLI to host the examples and scaffoldings, and then folks can plug that in and use. And the template universe can update over time. - [ ] https://github.com/withastro/astro/tree/main/packages/create-astro - [ ] Reinforces that the system pulls from a store of templates or repos - [ ] Angie - There’s 3 parties: - [ ] Wallet - we discussed earlier - [ ] PFI - we’re not dismissing this yet, Mic prototyped one as a good place to start - [ ] Issuer - Solve the use case of how do I issue? - [ ] Moe - In my mind the most impactful thing about out demo is that we actually sent real Bitcoin to a wallet. Having that same experience online at all times can be costly, but it does drive the point home: in 72 lines of code. - [ ] Becca - not in terms of templates just in terms of building dashboards for ourselves for the first time, but would be really cool to not have ppl repeat work bc it will all be pretty standard needs based on tbdex messages being roughly standard rn - [ ] Angie - Each use case needs a starter thing and helps folks get the information - [ ] Moe - This is the proof that things actually work, and lets us test out with devs that they find it valuable. If you’re coming at it as a beginner, it’s difficult to give them a win without teaching them a lot. - [ ] ALR - this is about lowering the barrier to entry in getting that dopamine hit. - [ ] Angie - Yeah, people who aren’t already invested aren’t looking to learn the protocols and backing tech before they get their first win. I don’t think the audience is a generic developer who doesn’t have some attachments to payments as a business. The Web5 case is more broad. - [ ] Leo - ^^ Yeah exactly! If the PFI apis are an open standard we could pretty much create a frontend dashboard cli where they could deploy in their internal network just to have a nice dashboard/monitoring default tool - [ ] Moe - i wonder if it'd be possible for us to run our own publicly accessible PFI that could exchange tbdollars -> very very smol amt of btc for people to dev against. obvious considerations here would be legal and regulatory - [ ] Mic - Even a sandbox/demo is fine - [ ] Angie - Folks who were talking to us on the ground were like “we have a payments service; what can you do for us?” Answer - we are the gateway to many payments providers. Speak tbDEX once, commerce everywhere. - [ ] Angie - I want to make sure we’re not losing the very basic premise: nail the “get me started with a quick template” use case - [ ] Ace - I do have a vision - starting with a npm package CLI solution that we can always expand with a template catalog. npm create web5@latest my-web5-app -- --template web5-ecommerce npm create web5@latest my-web5-app -- --example book-reviews npm create web5@latest my-web5-app - this opens up an interactive CLI - [ ] Nick - Could be a nice opportunity for convenience libraries - helps to get Web5 and tbDEX in frameworks like React - [ ] Moe - too early to invest in something expansive for tbdex. still have some figuring out out to do. until then, the smallest possible sample to get people their first dopamine hit would b guud - [ ] Mic - While we could do for Web5, for tbDEX still in example phase. - [ ] Angie - So maybe we start with system for Web5, paving the way for when tbDEX is ready. Have immediate need for Web5: DIDs, Credentials, etc - having people do that themselves is not a delightful experience.