Package Details: python-qiskit 1:1.2.4-1

Git Clone URL: https://aur.archlinux.org/python-qiskit.git (read-only, click to copy)
Package Base: python-qiskit
Description: An open-source SDK for working with (IBM) quantum computers
Upstream URL: https://github.com/Qiskit/qiskit
Licenses: Apache-2.0
Conflicts: python-qiskit-terra
Submitter: iyanmv
Maintainer: iyanmv
Last Packager: iyanmv
Votes: 4
Popularity: 0.001264
First Submitted: 2020-12-06 14:52 (UTC)
Last Updated: 2024-10-06 18:14 (UTC)

Latest Comments

1 2 Next › Last »

iyanmv commented on 2024-07-18 09:46 (UTC)

msrd0: can you check 1:1.1.1-2? Does it make a difference to have CARGO_TARGET_DIR or not? Also, most of the guidelines for Rust I think don't apply to python packages that use python-setuptools-rust.

My point is that I'm not an expert in Rust, and you seem to know way more than me about Rust. I'm totally open to suggestions and ideas to improve the PKGBUILD, but please justify them clearly and, ideally, provide a patch. Simply saying remove x because it doesn't work with my custom setup y will not do it.

iyanmv commented on 2024-07-18 08:28 (UTC)

msrd0: In the Rust packaging guidelines they don't mention anything about what you are saying. Actually, the do mention to include export RUSTUP_TOOLCHAIN=stable in prepare() if upstream does not provide a rust-toolchain or rust-toolchain.toml file.

If you have a particular setup you should modify the PKGBUILD for your needs or build it in a clean chroot env. Regarding the issue with your custom RUSTUP_HOME, perhaps I can improve the PKGBUILD exporting the variable CARGO_TARGET_DIR to use the same directory as your build directory.

Please let me know if this makes sense to you. If I'm totally wrong here, please send a patch for the PKGBUILD so I can have a look.

msrd0 commented on 2024-07-16 17:30 (UTC)

Please remove all rust-toolchain* files before compilation. Rust is stable enough and I really don't need the exact version they want installed on my system. Even worse, given that my RUSTUP_HOME is intentionally not writeable by my user account, this package fails to compile because it allegedly cannot find a rust compiler, when in reality it refuses to use the existing good latest stable rust compiler because it thinks it would be intelligent. This should not be the default behaviour of rustup IMHO, but given that I can't fix that, please just remove all rust-toolchain* files from the repo before compilation, so this stupid pseudo-intelligent behaviour does not get triggered, and the package compiles.

iyanmv commented on 2023-08-27 14:48 (UTC)

This is no longer a "metapackage" and it is what it used to be called qiskit-terra. Please let me know if you have any issues with the last update. I had to add epoch=1 because the version went from 0.44.0 to 0.25.1.

iyanmv commented on 2022-09-28 08:43 (UTC) (edited on 2022-09-28 08:44 (UTC) by iyanmv)

I changed my mind and I will keep all the packages as they are right now. At least, until things are changed upstream.

petronny commented on 2022-09-28 08:39 (UTC)

I plan to rewrite all these python-qiskit PKGBUILDS soon.

Hi, will these changes to be applied soon?

iyanmv commented on 2022-02-10 10:36 (UTC)

I plan to rewrite all these python-qiskit PKGBUILDS soon. Instead of so many packages python-qiskit-*, I will only keep two: python-qiskit (terra, aer, and ibqm) and python-qiskit-extra (optimization, finance, nature and experiments).

iyanmv commented on 2021-12-13 11:38 (UTC)

See this: https://aur.archlinux.org/packages/python-qiskit-terra/#pinned-840637

iyanmv commented on 2021-12-02 12:30 (UTC)

micwoj92: I will update the packages within this week. Hopefully, it doesn't take too much time, but in the past, things didn't work as smooth as they should. I think qiskit modules are a bit "chaotic" at the moment, and the moment you leave the "official" way of doing things (pip install qiskit), there are a few issues.

1 2 Next › Last »