Product decisions
What changed, and what I chose not to do.
Opinionated software is only worth anything if the opinions are any good. This is the running log — every decision, and the option I turned down beside it.
Not a blog. If I'm going to claim this software respects the people using it, these entries are the evidence.
This week
Declining a suggestion now requires a reason
The field is validated server-side with a minimum of one character, so a no can't be submitted empty.
I could have made it optional. A silent decline tells the suggester nothing and turns into a follow-up message anyway. Better to ask for one sentence now than a Slack thread later.
This week
An admin can no longer remove the team owner
Ownership is checked before removal, and self-removal is blocked separately.
I could have left it to good behaviour. Anything a data model permits will eventually happen at 6pm on a Friday.
Last month
Private todos don't survive you leaving
When an account closes, team todos stay so the work isn't stranded. Private ones are deleted outright.
I could have kept everything — more data is easier to build against. But a private todo was private for a reason, and that reason doesn't expire when you quit.
From the start
There is no priority field
Order is the priority, and only the person doing the work sets it. A suggestion can't carry urgency, position or a date.
I could have added a P1/P2/P3 flag. Everything becomes P1 within a month, and then you've taught your team that your urgency outranks their judgement.
You already know how you want to run your team.
This is the tool that stops arguing with you about it. Free for two, forever.