Sun Microsystems began developing gettext in the early 1990s and the GNU Project began GNU gettext development in 1995 for this widely-used internationalization and localization system commonly for multi-lingual integration. While GNU gettext is commonly used by countless open-source projects and adapted for many different programming languages, only an hour ago was GNU gettext 1.0 finally released.
GNU gettext has crossed the symbolic “v1.0” milestone over 30 years after its development began. GNU gettext 1.0 succeeds the v0.26 release from last July.
GNU gettext 1.0 brings PO file handling improvements, a new “po-fetch” program to fetch translated PO files from a translation project’s site on the Internet, new “msgpre” and “spit” pre-translation programs, and Ocaml and Rust programming language improvements.
With this v1.0 release in 2026, the “msgpre” and “spit” programs do involve…. Large Language Models (LLMs) in the era of AI:
“Two new programs, ‘msgpre’ and ‘spit’, are provided, that implement machine translation through a locally installed Large Language Model (LLM). ‘msgpre’ applies to an entire PO file, ‘spit’ to a single message.”
And when dealing with LLMs, added documentation warns users to look out for the licensing of the LLM in the spirit of free software.
More details on the GNU gettext 1.0 changes via the NEWS file.
GNU gettext 1.0 can be downloaded from GNU.org.
