I was the BanklessDAO website project champion during season 3 (last season). In this post, I want to give an update and a retrospective of the project.
A Humble Beginning
We started season 3 off with the idea to add a few blog articles to the website and shore up a few UI bugs. The site was built on a very popular blogging platform called Ghost. A copy of the open-source version of Ghost was running on Github and Vercel in addition to the hosted Ghost Pro service. This allowed us to show custom code (the Web3 wallet) and use the main Ghost admin login to post articles. Users were able to login to the ghost domain and add custom blog-style content. When new articles were added, it triggered a deployment and the final result is all the ghost blog stuff along with the Web3 wallet code.
An Ambitious Vision
There were several setbacks with the above setup. Firstly there was a huge reliance on Ghost as a service. Being this was a Web3 project we felt we should reduce our reliance on this third-party service. There was also the very practical matter of the scalability of the setup. The login to add articles to the ghost website was broadly shared and it was not clear what kind of moderation policies we should employ going forward. Should we be the ones scrutinizing every new projects and guilds content and auditing them for accuracy and safety? We didn’t feel that should be or could be our job so we landed on the following scheme.
Firstly, we would remove our dependence on Ghost pro as a service. The site is a static website and static websites are virtually free to deploy and host globally via Github and CDNs so this seemed like a no-brainer. This meant moving to a JAMstack setup. This setup basically means that when the source code changes in version control, that new version gets automatically deployed to a production CDN.
Secondly, we would begin exploring decentralization options. Assuming our new JAMstack configuration, we could attempt to replace the centralized components with Web3-friendly ones. This initiative came to be known as the FRAG stack. The acronym stood for Fleek, Radicle, Akash, and Gnosis Safe + ENS. We brought in and met with Fleek, Radicle, and Akash directly. What was interesting was that despite the advertising, most of these services and products were not as decentralized or production-ready as they were portrayed to be. Despite this, we increasingly dialed in on a solution we felt would not only work for us but could be a critical piece of Web3 infrastructure. More to come on FRAGstack in a moment.
Another consequence of moving to this new setup was going to be a change in the way we handled website content and content moderation. Rather than treating the website as a destination for community content, we felt we could treat it as a springboard linking out to all the diverse projects and guild content. This meant focusing on the core information of the DAO that was subgroup agnostic and providing at least one landing page for all subgroups that could link out to their individual websites or wiki pages.
One additional avenue we felt was very promising was the Content Gateway project. As I understood it, it would provide an API feed of content that was moderated by a community. Think API feed where the composition is determined democratically. Armed with this, I felt we could get the benefits of a fully decentralized front end via the JAMstack but also the dynamic content feeds of a content-rich website.
Execution
Given the above goals, we needed a new design and a rewrite of basic content. We initially sought outside help from Rumfoords and Cthdrl. Rumford’s proved to be a dead-end because they were more interested in a full Bankless rebrand and revealed that even they did not do their own web design and development. Cthdrl showed interest but was in such demand that we could not ultimately secure their services. My hope was to be able to circle back to them for some website “spice” at a future point. After both of these options fizzled out, Design Guild said they could take the job. This is the design they delivered: Figma. The whole team was happy with the design so we started to code up the site.
Current State
The FRAG stack is still in flight and making great progress. We feel that we can have a solution to the decentralized front-end puzzle that will be of enormous value to the entire Web3 ecosystem. Radicle is a big piece of that and they are working towards the same end. It is our hope to be the first community to demonstrate this solution. We continue to work extremely close with Radicle and build out our own custom piece of the solution.
Content Gateway is live but further development has been passed. I still believe in the vision behind the product and hope to see it revived after we successfully deliver on our current goals. To paint a picture of how these technologies can and should come together, consider the following picture.
- DAOs are built on Miltisigs.
- FRAGstack is trying to put front-end development and deployments behind multisig.
- Content Gateway can put API data feeds behind multisig.
Do you see the connection and importance of these things now?
Lastly, we have the web design. You can see the current state of development here. We need the community to give us feedback and permission to go live!! Not everyone is impressed with it. That’s ok. The website will always be improving and we plan to continue the inclusion of subpages dedicated to individual projects and guilds over the course of this season. We don’t need to be perfect at launch. It’s just the beginning. Please leave feedback on this form here and we will work to correct any content errors and make the necessary design changes to satisfy community sentiment. Feedback form here