client participation beta

[ballPointPenguin]

Stacked on #2302

TODO

• Update all libraries
• Convert js to ts, jsx to tsx
• Set up eslint
• Lint fixes
• Improve typings throughout
• Set up formatting (prettier)
• Format everything
• Centralize API calls in src/api/
• Extract svg markup into components
• Break up PCA visualization into smaller components
• Validate a11y
• Validate custom translations
• Validate 3rd party translations
• Unit tests
• Component tests
• Github Workflow for CI
• Validate Embeds
• E2E Cypress tests
• Validate mobile layout
• Screen Reader affordances
• Custom Themes (future work)

[tevko]

cc @ballPointPenguin can we get a screenshot of performance differences between this pr and client-participation-alpha survey page on prod? just want to ensure we've preserved our performance metrics, particularly with regard to initial page load size and time

[ballPointPenguin]

@tevko
This is the results from some basic web app perf evaluations with lighthouse.
The script for this is in branch br/perf-compare.
Results are very promising. We also capture a current prod evaluation for before/after comparison.

Performance Comparison Report

• Date: 2026-02-17 01:10:46
• Conversation: 2jztymipfp
• Runs per target: 5
• Tool: Lighthouse 13.0.3
• Mode: Desktop preset, simulated throttling

Branches

Target	Branch / URL	Git SHA
Production	https://pol.is/alpha/2jztymipfp	(deployed)
edge (local)	edge	584d5f50b
beta (local)	br/client-participation-beta	d0c8a7ba1

Results (Median of 5 runs)

Metric	Production	edge (local)	beta (local)	Δ edge→beta
Performance Score	78.0	93.0	100.0	+7.0
First Contentful Paint	230.7ms	464.0ms	464.2ms	+0.3ms
Largest Contentful Paint	234.7ms	646.0ms	646.3ms	+0.3ms
Total Blocking Time	475.0ms	0.0ms	0.0ms	0.0ms
Speed Index	867.7ms	464.0ms	464.2ms	+0.3ms
Cumulative Layout Shift	0.1	0.2	0.0	-0.1

Individual Run Scores

• production: 77, 77, 78, 78, 79
• edge: 93, 93, 93, 96, 93
• beta: 100, 100, 100, 100, 100

Notes

• Local runs used the Astro standalone preview server (SSR) with the Docker API backend.
• Production results include network latency and CDN effects; local results do not.
• The Δ column shows the change from edge→beta. Negative values for timing metrics indicate improvement.

---

Generated by scripts/perf-compare.sh

[xeeg]

Voting on the new embed fails with 500 when topic either does not exist, or is not selected.
See https://www.showra.org/conversations/testing for 2jztymipfp (same one used here for testing i.e with topics) and https://www.showra.org/conversations/on-democracy for one without topics.

Also XID, ucsh, ucst, ucsh, ui_lang are all broken.