2023-02-08 launch planning meeting

General update and Q&A

Kriti gave an update on

• Launch goals
• Funding plan
• Usability testing strategy

Launch goals

• Get early adopters that are really excited about Mathesar and can help us alpha-test it
• Post Mathesar on HN, Reddit
• Get people to install and use it
• Website ready
• Documentation for at least one way to install
• Only have stuff in Launch blockers that block user from deciding whether to install Mathesar
  • Small usability issues and bugs don’t block users from deciding to install Mathesar
• Initial target users: Software engineers
  • Reducing target audience size
  • Only aiming now for HN-user types, engineers (since they are the ones deciding whether to install Mathesar)

Funding

• For Non-profit projects, the important thing is impact
  • Theory of change
• Can’t do anything until we launch
• Need to show that we’ve shipped something
• Need to show people are interested in it.
• Kriti wants to do pitches in March/April

Usability testing

• Using tool: Playbook UX
• We provide tasks. People record themselves doing it.
• We also ask questions via the tool.
• Goal is to find people similar to initial target audience (HN denizens)
• Main focus is on high-impact changes which might affect adoption
• Also provides insights which we can do post-launch
• Current testing strategy is unmoderated. We might want to do some moderated tests later.
• We’ve been refining tests. First test didn’t go great.
• We’re also focusing on the right segment of users for the tests.
• Sean thinks usability tests on installation would be quite useful.
  • Kriti plans to find people on Upwork for that.
  • It’s on the plan for launch.
• For more open-ended instructions, we’d need a bigger sample size to get good info.
• The test is designed as a narrative
  • Exploring existing Mathesar Con data set
  • Adding a table called ‘Sponsors’
  • Assigning Sponsors to talks
  • Creating an exploration at prompt
• Each test is giving us a 20-minute video of someone using Mathesar

Launch plan

We’re going to launch on Feb 27.

Blockers NEED to get done before launch.

Blockers

• Users & Permissions (Pavish, Rajat, Sean)
  • Pavish & Kriti met Tue to cut scope a bit
• Deployment (Brent, Dom, Mukesh)
  • Brent & Kriti met Tue to cut scope
  • Only “Deployment Type 1”
• Usability testing issues (Ghislaine)
  • https://github.com/centerofci/mathesar/issues/2413
  • More incoming – no major changes planned
• Deployment usability testing (Kriti)
• Demo video (Kriti)
• Update README (Kriti)
• Complete website (Kriti & Ghislaine)
  • Link to docs
  • Integrate demo video
• Reset migrations (???)
• Create and publish release (Kriti)
  • Think about versioning (semantic versioning may not be appopriate for applications)
• Document upgrade process (Brent)

Nice to have

• Upgrades in UI (Sean)
• Additional deployment types (Brent, Dom, Mukesh)
• “High” items in Backlog milestone (Pavish, Rajat, Sean, Ghislaine)
• Pre-launch publicity (Kriti)
• Process for publishing releases on DockerHub

Process

• Currently ad-hoc process for launch
• We’ll discuss improving velocity in a more permanent way after launch
• Considering meetings 2-3 times a week to stay focused.
  • Everyone agrees, we’ll set these up.

GSoC / community work

• We’ll stop reviewing community work until launch.
• Anish will handle community communication until launch.
  • Kriti will support as needed.
  • Everyone else should ignore community work until we launch.
• If we get into GSoC, we’ll review and update project ideas after launch.

Next steps

• Update launch milestone in GitHub to be accurate.
• Kriti will set up meetings

Current status

• Users and permissions
  • Try to get this done before any other frontend work, it’s in blockers
  • Rajat’s PR is open for styling
• Deployment
  • There’s a preliminary writeup in #2034
  • New Docker setup for prod is finally merged
    • Brent will send email to mailing list with dev changes
  • install script is looking good, almost done (Mukesh)
  • Documentation is pending
    • Port needs to be documented so that people can run things on different ports if needed
    • Make sure we’re clear about using WSL in Windows
    • Make prerequisites of bash and Docker clear
  • Install script will be friendlier for Windows too
  • Documentation for happy path - ETA Friday or early next week
  • Next steps:
    • Get install.sh done and merged
    • Get Watchtower (upgrade library) done and working with the current setup
    • Documentation
  • Upgrade UI is blocked on discussion
    • Mukesh, Brent, Dom, Sean, Kriti should meet
  • Docker won’t pull the new image unless you give it a “build” argument
  • Installation flow
  • Write up for user flow is in #2034
  • We need to make sure we’re not building frontend files in the npm container
    • Use pre-built files
    • Double check run.sh - it might be cruft
    • We need to revert package lock changes
  • Make sure to involve frontend engineers in installation review (probably Pavish)
• Usability testing
  • Next steps
    • Finish writing notes for completed tests
    • Write up tasks for what we need to fix before launch
      • Critical
      • Easy to implement
    • Write up other, non-blocking tasks
• Deployment usability testing
  • Wait until we have docs
• Demo video
  • Going back and forth on feedback
  • Doing a new voiceover
  • We seem to have a voiceover sample that works
  • Should have v2 soon (hopefully final)
• Need to update README.md - not started
• Website
  • Changes implemented
  • Updating illustrations - in review
  • Once we have docs and demo video, need to integrate
• Not started:
  • Reset migrations
  • Create and publish release (Kriti)
  • Document upgrade process (Brent)