Skip to content
Spectre iOS v3.1.0, build 47 (beta)

UPDATED
-------

- More efficient and smoother updating of site and user cells.
- Rather than blocking insecure URLs, try to load them securely instead.
- Reduced performance sampling rate.

FIXED
-----

- Keyboard appearance animations are now smoother.
- Under rare circumstances, technical site metdata could find itself in the pasteboard.

CHANGELOG
---------

 - Bump to 3.1 for next public release.
 - Fix keyboard animation.
   [FIXED]     Keyboard changes' animation settings were not delegated.
   [FIXED]     Keyboard callbacks are now fully animated.
 - Observer unregistration improvements.
   [FIXED]     A few missing unregistrations of observer registrations.
   [IMPROVED]  Avoid a few unnecessary unregistration/registrations.
   [IMPROVED]  Avoid unnecessary post-registration callbacks when already registered.
 - Apple Silicon build compatibility.
   [UPDATED]   Apple Silicon has a simulator arm64 & a device arm64: platforms need to be separated.
               This should also yield smaller library binaries and faster builds (that don't include unnecessary platforms and architectures).
 - Fixes to URL handling and site name size.
   [FIXED]     URL to Spectre Community for Questions.
   [FIXED]     Scheme slash handling when converting absolute URLs into https:// URLs.
   [FIXED]     Convert all image URLs to HTTPS for increased privacy & to conform with ATS.
   [IMPROVED]  Increase font size of site cell names to callout.
 - Code format update.
   [UPDATED]   Apply code formatting rules.
 - Implement DataSource using NSDiffableDataSourceSnapshot
   [UPDATED]   More reliable data source updates through iOS 13's NSDiffableDataSourceSnapshot.
 - Move to iOS 13.
   [UPDATED]   Minimum supported iOS is now 13 to facilitate new API adoption.
 - Fix leaking preview data out via the pasteboard.
   [FIXED]     Left-over debug code caused site preview data to be placed in the pasteboard.
 - Reduce performance sampling rate.
   [UPDATED]   Performance sample only 1/10 sessions, randomizing which sessions are impacted and reducing the transaction load on Sentry.