In this interview conducted at DevConf.cz 2014, I discuss my work as an ext4 filesystem developer at Red Hat’s Brno office. The conversation explores why ext4 continues to receive new features rather than transitioning to a new version like ext5, explaining that recent additions such as Bigalloc and metadata checksumming aren’t substantial enough to warrant a completely new filesystem. I compare ext4’s development philosophy with emerging filesystems like Btrfs and F2FS, noting that Btrfs isn’t primarily focused on performance but rather on unique features like copy-on-write and snapshotting, while F2FS faces challenges due to lack of detailed information from flash memory manufacturers. I emphasize that filesystem stabilization is a lengthy process taking 10+ years, making ext4’s maturity a significant advantage over newer alternatives.

Key Takeaways: Ext4 remains viable because new features don’t require *fundamental format changes that would justify creating ext5; filesystem *stabilization requires a decade or more, making mature filesystems like ext4 *more reliable than newer alternatives like F2FS for production use.