Created attachment 254333 [details] Python-3.14.0a1 New port python-3.14.0a1 Result of `make test`: 10 slowest tests: - test_tools: 14 min 14 sec - test_subprocess: 1 min 42 sec - test_pathlib: 1 min 38 sec - test_zipimport: 1 min 33 sec - test.test_multiprocessing_spawn.test_processes: 1 min 26 sec - test.test_multiprocessing_forkserver.test_processes: 1 min 17 sec - test_ssl: 1 min 12 sec - test_peg_generator: 1 min 5 sec - test.test_multiprocessing_fork.test_processes: 1 min - test.test_concurrent_futures.test_wait: 58.8 sec 30 tests skipped: test.test_asyncio.test_windows_events test.test_asyncio.test_windows_utils test.test_gdb.test_backtrace test.test_gdb.test_cfunction test.test_gdb.test_cfunction_full test.test_gdb.test_misc test.test_gdb.test_pretty_print test_android test_dbm_gnu test_dbm_sqlite3 test_devpoll test_epoll test_free_threading test_idle test_launcher test_msvcrt test_perf_profiler test_perfmaps test_sqlite3 test_startfile test_tcl test_tkinter test_ttk test_ttk_textonly test_turtle test_winapi test_winconsoleio test_winreg test_winsound test_wmi 1 test skipped (resource denied): test_zipfile64 3 re-run tests: test.test_concurrent_futures.test_thread_pool test_dtrace test_shutil 2 tests failed: test_dtrace test_shutil 445 tests OK.
Created attachment 256731 [details] python-3.14.0 alpha4 update to alpha4, `make test`: 1 test skipped (resource denied): test_zipfile64 3 re-run tests: test_dtrace test_embed test_shutil 3 tests failed: test_dtrace test_embed test_shutil 448 tests OK.
Created attachment 260279 [details] Update to 3.14beta1 Update to 3.14beta1
Created attachment 260990 [details] Update to python-3.14.0beta2 Update to python-3.14.0beta2
Created attachment 262383 [details] Update to 3.14.0 RC1 Update to 3.14.0 RC1 the result of `make test`: Total duration: 7 min 16 sec Total tests: run=45,753 failures=16 skipped=2,063 Total test files: run=493/491 failed=3 skipped=32 resource_denied=1 rerun=3
Created attachment 262982 [details] Update to 3.14.0 RC2 Update to 3.14.0 RC2
Created attachment 263918 [details] Python-3.14.0rc3 Update to 3.14.0RC3
Created attachment 264409 [details] Update to 3.14.0 release Update to 3.14.0 release
Why don't you just commit both this and 3.13? Can we expect the addition of the new python versions to ports tree at least in 2035? [I understand that I am too optimistic there]
result of `make test`: ... 3 tests failed: test_dtrace test_posix test_shutil ... Total duration: 11 min 40 sec Total tests: run=45,807 failures=16 skipped=2,076 Total test files: run=493/491 failed=3 skipped=32 resource_denied=1 rerun=3 Result: FAILURE then FAILURE
wen@ it would really help to post details from make test. Are there detailed test logs?
A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/ports/commit/?id=60648732604548051b170d9e599e05b870f9d008 commit 60648732604548051b170d9e599e05b870f9d008 Author: Matthias Andree <mandree@FreeBSD.org> AuthorDate: 2025-11-01 20:02:17 +0000 Commit: Matthias Andree <mandree@FreeBSD.org> CommitDate: 2025-11-01 20:11:56 +0000 lang/python314: get port ready to ship claim maintainership on the new port - not doing that with the previous commit by wen@ to have a clear distinction who contributed what. - require archivers/zstd since it's part of the Python standard library: https://docs.python.org/3.14/whatsnew/3.14.html#whatsnew314-zstandard - refresh Makefile.pre.in patch - skip test_gdb without WITH_DEBUG test_gdb has test_pretty_print, which requires debug symbols. Skip it if WITH_DEBUG is not defined. - drop --with-system-ffi configure option, which is no longer supported and its behaviour is now the default - move sqlite3 extension back into port because the separate port fails to build (mark the external python:-3.13) PR: 282176 Mk/Uses/python.mk | 2 +- Mk/bsd.default-versions.mk | 2 +- databases/py-sqlite3/Makefile | 2 +- lang/python314/Makefile | 18 ++++++++++-------- lang/python314/files/patch-Makefile.pre.in | 6 +++--- lang/python314/files/pkg-message.in | 1 - lang/python314/pkg-plist | 2 ++ 7 files changed, 18 insertions(+), 15 deletions(-)
A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/ports/commit/?id=962f4aa7d8cdfe5ac7244cc806f4d451a0414da0 commit 962f4aa7d8cdfe5ac7244cc806f4d451a0414da0 Author: Wen Heping <wen@FreeBSD.org> AuthorDate: 2025-11-01 17:49:09 +0000 Commit: Matthias Andree <mandree@FreeBSD.org> CommitDate: 2025-11-01 20:11:55 +0000 lang/python314: add new port ----- committer's note by mandree@: This does not enable the optional free-threaded build. PR: 282176 lang/Makefile | 1 + lang/python314/Makefile (new) | 167 + lang/python314/Makefile.version (new) | 7 + lang/python314/distinfo (new) | 3 + lang/python314/files/patch-Makefile.pre.in (new) | 62 + .../files/patch-Misc_python-config.sh.in (new) | 11 + lang/python314/files/pkg-message.in (new) | 12 + lang/python314/pkg-descr (new) | 2 + lang/python314/pkg-plist (new) | 8238 ++++++++++++++++++++ 9 files changed, 8503 insertions(+)
This passes the self-test suite on FreeBSD 13.5 amd64, 14.3 amd64 with and without WITH_DEBUG, and 14.3 amd64 without WITH_DEBUG under openssl35.
(In reply to Matthias Andree from comment #13) I am working to build it with DEBUG option and add a new option WITHOUT_GIL.
(In reply to Wen Heping from comment #14) I tried that way a long time ago when 3.13 was still new upstream, it's not possible. Free-threaded mode is different ABI which results in a completely separate distribution hierarchy (like site-packages). See how lang/python313 and lang/python313t actually landed, because it's not one new flavour, but two.