Another Day, Another Modal
Spent most of today inside Figma. Again. Cleaning up inconsistencies in the guided modals. Turns out 2px padding differences start to haunt you once your brain is in flow mode. Polished the tooltip positions, updated type hierarchy, and created a quick checklist to catch things like button alignment before testing.
We’ve crossed into “testable” territory. The current prototype isn’t flashy, but it flows. More than I could say yesterday.
Firebase vs Supabase
Took some time to deep dive on backend options today. Did a side-by-side between Firebase and Supabase. Might seem premature without an MVP, but real-time updates are kind of central to what we’re building.
Here’s the short version:
- Firebase: Polished, but proprietary. Great docs. Sketchy when you want custom role-based permissions.
- Supabase: Open source. Built on top of Postgres (which we’d probably want anyway). Auth and row-level security look solid.
Leaning Supabase for now. Especially if we want to support custom integrations or user flows down the line.
User Feedback Is (Still) King
Went through notes from 5 early interviews + comments via IH and landing page. Clear signals:
- People want Notion + Linear integrations.
- No one wants another generic onboarding tool.
- Handoff sucks. Everyone agrees.
Working with my product advisor to tighten positioning. Less “onboarding banner,” more “real-time collaboration layer for handoff.” It feels… sharper.
Navigation Glitch = QA Time
Found a weird bug with button behavior inside the guided modals. Somehow not all of them were navigating properly. 🙄
Fixed it. Then built a small QA checklist to avoid that happening again. Stable = testable. Unstable = untestable. That’s the rule now.
A Call That Made My Day
Got pinged through a friend by a potential angel investor. Quick exploratory Zoom. Honestly, I didn’t even have a pitch deck open. Just walked them through the vision and what’s coming.
Surprisingly positive reaction. No commitment, but they said they’d be interested in joining the pre-seed if MVP shows promise post-testing. That alone made the 11pm bug fix feel useful.
Ongoing Debate: MVP Scope
The whole team’s split between:
- Build enough infra to make it "expandable"
- Keep it painfully lean and solve one nagging pain
Right now, lean is winning. We'll ship the smallest possible thing that actually helps a designer-to-developer handoff, test, and go from there.
Lesson of the Day: Picking the Right Backend (Without Overbuilding)
If you’re pre-MVP and choosing backend infra, here's my 3-part filter:
- Can it scale if your thing works? Don’t optimize for now. Optimize for “post-success.”
- Does it have sane defaults for security + auth? You can’t afford to build brittle permission logic this early.
- How painful is rollback or migration? If it locks you in hard, bake in budget to switch later. Or pick open source now.
That’s why Supabase is looking like the right bet—for both flexibility and realism.
Clarity is sharpening. Momentum’s slow but steady. If the next 10 days go as planned, there’s a real MVP forming. Time to keep pushing.
