Love to see it, great work everyone.

On Tue, Oct 13, 2026 at 5:21 PM John Benediktsson <[email protected]> wrote:

> “You don’t want just to develop for ARM, you want to actually use ARM in
> your day-to-day work on the desktop.” - Linus Torvalds
>
> I'm very pleased to announce the release of Factor 0.102! You can find
> download links on the Factor website:
>
>     https://factorcode.org
>
> This release includes 1,068 commits from the following contributors:
>
> Aleksander Sabak, Andrii Puhachenko, Chris Double, Dmitry Matveyev, Doug
> Coleman, Eric Willigers, Giftpflanze, John Benediktsson, Leo Mehraban, Rudi
> Grinberg, Sebastiaan Speck, Zoltán Kéri, @knottio, @ronnie-w
>
> Highlights of this release include:
>
> * A new GTK4 UI backend, now the default on 64-bit Linux; 32-bit Linux
> continues to use GTK3, which remains a build option on other supported
> Linux architectures.
> * New 32-bit and 64-bit RISC-V Linux ports, with optimizing compiler
> backends, foreign calls and callbacks, and support in both the C++ and Zig
> VMs.
> * DirectWrite text rendering on Windows, including color glyphs,
> typography controls, and better support for large output and complex
> scripts.
> * More responsive rendering and selection of large text across platforms,
> with fixes for display scaling, listener scrolling, and adjacent text
> backgrounds.
> * Expanded native ARM64 SIMD support, half-precision and bfloat16
> operations, and improved foreign calls, variadic callbacks, and native
> va_list support.
> * Experimental register allocators and optional compiler optimizations,
> alongside fixes to stack inference, generated code, and GC root tracking.
> * Updated Unicode data to 18.0.0 and timezone data to 2026c, with bindings
> for SQLite 3.53.4, Raylib 6.0, and Lua 5.5.
> * Faster process launching and native process-exit notifications on Linux,
> macOS, BSD, and Windows, with polling as a fallback.
> * Extensive fixes to HTTP parsing, TLS identity checks, cooperative waits,
> resource cleanup, signal handling, and VM memory safety.
>
> Compatibility notes
>
> * 64-bit Linux builds now select GTK4 by default; use -ui-backend=gtk3
> when bootstrapping if GTK3 is required. 32-bit Linux uses GTK3 and does not
> support GTK4.
> * Windows UI text now uses DirectWrite by default, so text metrics and
> appearance may differ from the previous Uniscribe backend.
> callable is now a mixin that includes words; code that assumes all
> callables support the sequence protocol should use quotation-like instead.
> * Fixed-array struct fields now validate input buffer lengths, and
> stricter HTTP and multipart parsing rejects malformed or incomplete input
> previously accepted.
> * PEG repetition and list parsers now omit ignore values. Minimum counts
> refer to successful matches, so repeat1 and list-of-many can return empty
> output when all matched items are hidden.
> * The buffered HTTP client now decompresses gzip responses before
> character decoding and rejects unsupported content codings.
> with-http-request still provides bytes before content decompression.
> * Ambiguous Objective-C selectors now require a class-qualified name such
> as Class.selector when imported signatures differ.
> * Tuple literals reject duplicate slot names, and minimum and maximum now
> throw for empty iota and repetition instead of returning a value.
> * Command-line option names and aliases must now be unique, including the
> default --help option. Exact names take precedence over no- negation.
> * SQLite connections now use persistent WAL mode, which requires a local
> filesystem. with-transaction uses BEGIN IMMEDIATEand acquires the write
> lock at transaction start, including for read-only work. Write
> serialization is per process; wrap raw SQL and DDL writes in
> with-transaction or serialize-transaction to participate.
> * Building the experimental Zig VM now requires Zig 0.17.0.
>
> For more details, please see the full announcement at:
>
> https://re.factorcode.org/2026/10/factor-0-102-now-available.html
>
> _______________________________________________
> Factor-talk mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/factor-talk
>
_______________________________________________
Factor-talk mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/factor-talk

Reply via email to