I'll include this if it gets merged upstream.
Search Criteria
Package Details: python-future 1.0.0-6
Package Actions
| Git Clone URL: | https://aur.archlinux.org/python-future.git (read-only, click to copy) |
|---|---|
| Package Base: | python-future |
| Description: | Clean single-source support for Python 3 and 2 |
| Upstream URL: | https://python-future.org/ |
| Licenses: | MIT |
| Provides: | futurize, pasteurize |
| Submitter: | carsme |
| Maintainer: | micwoj92 |
| Last Packager: | micwoj92 |
| Votes: | 9 |
| Popularity: | 0.105238 |
| First Submitted: | 2024-12-22 16:15 (UTC) |
| Last Updated: | 2026-02-16 22:23 (UTC) |
Dependencies (10)
- python
- python-fissix
- python-sndhdr (python-standard-sndhdr)
- python-uu (python-standard-uu)
- python-build (make)
- python-installer (make)
- python-setuptools (make)
- python-pytest (check)
- python-requests (check)
- python-setuptools (optional) – futurize and pasteurize scripts
Required by (9)
Sources (1)
micwoj92 commented on 2026-02-16 22:23 (UTC)
mirh commented on 2026-02-16 12:37 (UTC)
You can presumably drop the uu dependency if you include https://github.com/PythonCharmers/python-future/pull/644
micwoj92 commented on 2026-02-01 07:19 (UTC)
test_future still fails and I couldn''t figure it out.
micwoj92 commented on 2025-11-27 01:41 (UTC)
Thanks, that's similar to what I've been doing. First try to build, check with namcap. If it says that Dependency included, but may not be needed ('python-future') there are big chances that it can just be removed. Not foolproof, but worked for me in >95% cases for this particular issue.
That is not entirely true about __future__. There were 2 additions in 3.x series: https://docs.python.org/3/library/__future__.html#module-contents
Also because of the "nameclash" I've seen few cases like this where import __future__ was added and at the same time needlessly added future to install_requires.
gwuensch commented on 2025-11-27 01:14 (UTC)
If someone cared then these would have been fixed already.
You're probably right, thanks for the answer. Nonetheless, if anyone still wants to bother checking which packages actually require python-future, I've updated my previous comment with some info on what to look for.
micwoj92 commented on 2025-11-27 00:55 (UTC)
I did, more or less.
For passhole I submited orphan request. But if package was broken for many years, dead upstreams, no aur pkg activity then I submit deletion requests. If someone cared then these would have been fixed already. @dreieck did a great job with sending reminders comments on packages that still required python-future and it was almost a year ago. So people should not act surprised when something is deleted after being broken* for 1 year.
From that round of requests:
Deletion: yandex-tank splatnet2statink python-postpic python-perspective python-pyevmasm python-conda-verify python-conda-build nemesyst-git mavlink ardupilot-copter-sitl
Orphan: rmtoo python-kicadmodtree-git python-gogs-client python-coq-jupyter python-bayesicfitting passhole greatfet-git funky-git
- with broken I mean that this very package fails to build without intervention. Another way to send a "reminder".
gwuensch commented on 2025-11-27 00:41 (UTC) (edited on 2025-11-27 01:19 (UTC) by gwuensch)
Before filing deletion requests, have you checked these packages are actually “irreparably broken” and not just incorrectly packaged? Two packages I was using wrongly specified python-future as a dependency, even though it was never required when using Python 3.x.
I actually just randomly picked a package relying on python-future (passhole), and it's the same thing. I wouldn't be surprised if more packagers (or upstream maintainers!) did the same mistake.
General rule: If a package does not use the past package (which is included with future), it very likely doesn't actually require python-future on Python 3.x. Any __future__ statements which would be needed on Python 2.x are simply ignored by Python 3.x.
micwoj92 commented on 2025-11-26 19:46 (UTC)
Required by 27 now with few deletion requests pending.
micwoj92 commented on 2025-11-25 17:21 (UTC)
No, rather drop/update packages that depend of it 28 left. For python 3.13 this works, but 3.14 it might not.
markf commented on 2025-11-25 15:25 (UTC) (edited on 2025-11-25 15:25 (UTC) by markf)
@micwoj92 if you're still interested in updating this to work, here's a patch I just used locally:
diff --git a/PKGBUILD b/PKGBUILD
index 9a65199d..55069fe5 100644
--- a/PKGBUILD
+++ b/PKGBUILD
@@ -11,7 +11,7 @@ pkgdesc="Clean single-source support for Python 3 and 2"
url="https://python-future.org/"
arch=('any')
license=('MIT')
-depends=('python')
+depends=('python' 'python-fissix')
provides=('futurize' 'pasteurize')
checkdepends=('python-requests' 'python-pytest')
makedepends=('python-build' 'python-installer' 'python-setuptools' 'python-wheel')
@@ -40,3 +40,11 @@ package() {
install -D -m644 LICENSE.txt \
"$pkgdir"/usr/share/licenses/$pkgname/LICENSE.txt
}
+
+prepare() {
+ cd "$srcdir"/future-$pkgver
+ # Only touch likely text/code files to reduce noisy warnings
+ find . -type f \( -name "*.py" -o -name "*.pyi" -o -name "*.txt" -o -name "*.rst" -o -name "*.cfg" \) \
+ -exec sed -i 's|lib2to3|fissix|g' {} +
+}
Pinned Comments
micwoj92 commented on 2025-01-02 21:47 (UTC)
Do not post check errors here. Python-future package is not compatible with Python 3.13. If you use packages that depend on it, please contact maintainers and/or upstream of these packages and try to work out whether this dependency is really needed.