The PHP internals mailing list is discussing the RFC for deprecations planned for PHP 8.6. Gina P. Banyard announced the annual deprecation RFC, noting each deprecation will be voted on individually and the RFC must be frozen by July 13th. A notable debate emerged around whether proposals include proper impact analyses. Juliette Reinders Folmer criticized the lack of impact analysis across the ~25 proposals, while Tim Düsterhus argued that deprecations are not breaking changes and that impact analyses based on public code (e.g., Packagist) are inherently biased and incomplete. Rowan Tommins sided with Juliette, stating proposals without impact data would receive automatic 'No' votes. One concrete example shared was the 'return-in-finally' deprecation, where an analysis of ~5000 top Composer packages found only 12 affected sites across 9 packages (0.18%), with 3 being actual latent bugs.

48m read timeFrom externals.io
Post cover image
129 Impressions