I'm backtracking on Supabase. Again. Here's the scoop. And I'm walking back a lot of decisions in the interests of making something fast.

I've heard many warnings that building a cluster is painful at early startup stages and that we want to avoid doing that. The easiest workaround is to use hosted PaaS, and both Supabase and Vercel give us everything.

The Updated Plan

I'm moving the initial plans for Enterprise, which is something we can consider when we are bringing in paid enterprise clients.

Supabase

for supabase, I'm making a call, because it's going to work for us in a simple fashion for two Ladysmith test campaigns, plus a demo site.

Supabase SaaS is fine-tuned, region-locked (meaning we can host it in Canada), compliant for most security purposes, and finally, it has solutions for:

The pricing will net us abourt 25 usd a month. We will aim to start the pro plan somewhere closer to August in order to support the campaigns.

• authnetication | authorization
• Task Runners
• File Storage (https://supabase.com/docs/guides/storage/management/pricing)

Tenancy

Data Security

This is something I haven't considered. We are going to have to use database encyrption and row level security, rather than having gaps between the two campaign instance. this can be done.

App entry points

Vercel allows the apps for platforms, which allows us to manage multitenancy. Imagine you have candidateA.com and candidateB.com. They would both point to the same winning ground application. The root domain would point to the homepage for visitors, and the login of canadiateA.com/signin would go to the campaign.candidateA.com, with the underlying data separated by row-level protection and encyrption at the database level.

Burning Questions

Should I use a different platform of volunteer managemnet and campaigners? Can this be managed by feature toggles? This would simplify the notion of homepage management and setup.

Other Notes

While I am building out the bones, we need to make sure to build this app for more people than just Echo and Brian. This is why I've asked Echo with writing a small set of user personnas, although I will be asking her to fall on her personal experience of helping others out throughout past campaigns and volunteer experience, rather than going through the difficulity of running interviews herself. That can come later. much later. Rough user personnas gives us a common framework of who we are building for which we can increment and refine on in the future. \

(although there are a few people I could talk to and grab their opinions).