AetherCutChangelogHow it worksPricingOpen the editor

Home / Changelog / v1.9.90

v1.9.90

Three next-action fixes: analytics, WebM fallback, shareable /six-tools

AetherCut v1.9.90 shipped on Aug 13, 2026.

What's new in v1.9.90?

  • Project Mode Analytics — VideoEditor.handleSetSidebarMode now fires trackFunnelOnce('project_mode_selected', [mode]). Backend already whitelisted the event but the frontend never dispatched it, so the admin distribution card was stuck at 0/0/0/0 since the mode-picker shipped.
  • Showcase WebM fallback — /showcase page adds an inline <video> preview with <source type='video/webm'> before <source type='video/mp4'>, so Firefox and headless-Chromium builds that reject unlicensed H.264 still play the demo. Each MP4 download row also exposes a companion WebM download button.
  • Six-tools shareable preview — /six-tools ships a dedicated OG card (og-six-tools-1200x630.jpg, generated via /app/scripts/generate_six_tools_og.py) with the '6 tools no other editor has' hook + 6 tool chips + brand footer, so the URL renders punchy in Slack/Twitter/LinkedIn instead of the generic hero card.
  • /six-tools?utm=share tracking — arrivals via that query bump a separate seo_landing_hits doc (six-tools_utm_share) and render a small green 'someone shared this with you' banner above the H1, doubling perceived credibility on first impression.
  • Pricing correction — /six-tools SEO copy + OG card footer originally advertised a single '$4.99/mo Pro tier' for the six AI flagships. The Production Suite ($4.99) is NOT AI; AI flagships unlock via the AI Layer ($9.99/mo) or Full Pro ($14.95/mo). Corrected every mention (intro, section body, FAQ, flagship block intro, OG card footer)
  • Regression suite grew to 413 (added v1990NextActions.regression.test.mjs locking in all three fixes)

How these fixes show up in real use

The easiest way to understand this release is to follow one real path through it. A new visitor sees a shared /six-tools link in Slack, gets a page card that actually explains why the URL matters, clicks through, and lands on a version of the page that acknowledges it was shared rather than discovered cold. If that visitor opens the page in Firefox, or in a browser build without H.264 support, the showcase preview still plays instead of silently failing. If they continue into the editor and try different project modes, those selections now reach analytics instead of disappearing between UI and backend. None of that is glamorous on its own, but together it closes three leaks in the same funnel: weak previews before the click, playback failure after the click, and missing product signals after activation.

A practical way to verify the analytics side is to test behavior rather than just event presence. Open the editor, switch project modes one by one, then refresh and repeat with a second mode order. The point is not merely that project_mode_selected fires, but that it fires from the actual sidebar interaction that users take, and only when the mode is selected. A common mistake with product instrumentation is counting renders, defaults, or restored state as deliberate choices, which inflates adoption and makes an admin card look healthy while telling you nothing. The fix here matters because the backend already accepted the event; the blind spot lived in the frontend dispatch. That kind of mismatch is especially easy to miss when dashboards show zeros, because teams often assume the report is empty due to low usage rather than a wiring gap.

The WebM fallback is similarly more important than it first appears because media bugs often hide behind machine-specific codec support. On a developer laptop with system codecs available, an MP4 preview can look perfectly reliable. Then the same page breaks in Firefox for some users or in headless Chromium during automated capture, QA, or embedded environments. Putting the WebM source first gives those environments a format they are more likely to decode without extra licensing baggage, while preserving MP4 for the mainstream path. If you want to test this fix properly, don’t stop after seeing the video play in one browser. Check source selection in Firefox, confirm the preview still loads when H.264 is unavailable, and verify that the companion WebM download is visible anywhere an MP4 download row appears. A fallback only helps if both preview and download paths stay coherent.

The /six-tools changes are best judged by how much context survives outside the site. Generic OG cards waste one of the highest-intent moments in distribution because they force every shared link to look like every other page. A dedicated card with the “6 tools no other editor has” hook gives the recipient enough reason to click before they ever read body copy. The ?utm=share treatment then continues that context after the click instead of dropping the visitor into a generic landing experience. One mistake teams make with share pages is tracking the campaign but failing to adapt the page state, which means the attribution helps reporting but not conversion. Here the banner does both soft social proof and message continuity. That continuity became even more important after the pricing correction, because a strong share card is only useful if the promise it amplifies is accurate.

How do I get this update?

Open the editor and you have it. AetherCut runs in Chrome, Edge, Safari and Firefox. There is nothing to install, and you can edit without an account. Open the editor or see how it works.

What shipped just before and after?

See all 111 releases · Follow the RSS feed