Tabs become unreadable with many open tabs

[Firnschnee]

Description

When too many tabs are open, tab widths shrink below a usable threshold. At that point:

• Tab titles become unreadable
• The dynamic active-tab expansion (--uc-active-tab-width) can no longer take effect because there's no space left to redistribute
• The visual distinction between active and inactive tabs breaks down

The entire dynamic tab sizing concept stops working once the tabs are squeezed too tight.

Possible approach

Auto-switch to scroll mode when individual tab width drops below a minimum threshold. Firefox supports overflow: scroll on the tab strip. The idea would be to trigger it based on tab count or computed width, so tabs maintain a readable minimum size and the active-tab expansion keeps working.

Open questions

• What's the right minimum tab width before scrolling kicks in?
• Pure CSS solution (e.g. min-width on tabs forces overflow naturally) or does this need a clamp()/container-query approach?
• How does this interact with --uc-active-tab-width / --uc-inactive-tab-width?

Needs investigation before implementation.

Labels

enhancement

[Firnschnee]

Closing as resolved indirectly.

The two concrete problems described here are now covered:

1. Active-tab distinction breaks down – fixed in 963e793. The inactive
   tab max-width ceiling was lowered (200px to 120px) so inactive tabs cap
   sooner and the active tab grows visibly larger across a much wider range
   of tab counts and window widths, instead of only at very wide tab strips.

2. Tabs shrink below a usable threshold – already handled by
   --uc-tab-min-width (76px default, added via Added min tab width modifier to userChrome.css #20). Firefox falls back to
   native overflow-scroll once tabs can no longer fit at their min-width, so
   they stop at a readable floor rather than shrinking to unreadable. Users
   who want to cram more tabs in can lower the variable (opt-in).

The originally proposed "auto-switch to scroll mode based on computed tab
width" turned out unnecessary: the native min-width floor already produces
exactly that behaviour, with no custom threshold to maintain.