Quansight-Labs / quansight-labs-site

💻 Development site and blog for Quansight Labs
https://labs.quansight.org
24 stars 44 forks source link

IPython 8 #289

Closed Carreau closed 2 years ago

Carreau commented 2 years ago

Text styling

Non-text contents

netlify[bot] commented 2 years ago

✔️ Deploy Preview for admiring-northcutt-a5be8c ready!

🔨 Explore the source changes: daf8bdfba53fd19c8f73df71a1de35066844608a

🔍 Inspect the deploy log: https://app.netlify.com/sites/admiring-northcutt-a5be8c/deploys/61ded543c61f3f0008717ccc

😎 Browse the preview: https://deploy-preview-289--admiring-northcutt-a5be8c.netlify.app

Carreau commented 2 years ago

Thanks @thomasjpfan , I've updated the sections you commented on.

ccordoba12 commented 2 years ago

I don't understand: why this one was merged before the final IPython 8 release?

Carreau commented 2 years ago

Because it was uploading while I merged.

https://pypi.org/project/ipython/8.0.0/

ccordoba12 commented 2 years ago

Ups, sorry about that. Since I saw a placeholder link for a release announcement at the Jupyter blog, I thought this could be a mistake.

Carreau commented 2 years ago

I also needed the URL to be live to update https://blog.jupyter.org/release-of-ipython-8-0-6e034ff122ef and link to it, this other blog post was published at the "same" time.

rgommers commented 2 years ago

Nice post Matthias!

Carreau commented 2 years ago

Ups, sorry about that. Since I saw a placeholder link for a release announcement at the Jupyter blog, I thought this could be a mistake.

Yeah it's a bit of a chicken and egg problem, I need to publish and update after, and it takes time to deploy. I think I should be able to guess the final URL but I was a bit lazy. It should be fixed in the source now, but will take time to hit CDN.

Nice post Matthias!

Thanks ! Any chance for numpy to grow a version_info tuple ?

rgommers commented 2 years ago

Thanks ! Any chance for numpy to grow a version_info tuple ?

Depends, for what? There's a bunch of flavors in np.version, but that's a weird thing. The only thing that was meant to be public was np.__version__.

There's no standard for a tuple version; the string can adhere to PEP 440. So is a tuple useful?

Carreau commented 2 years ago

So is a tuple useful

It's much easier to compare, especially since tuple don't need to be the same length:

>>> (3,) < (7,3,'7rc')
True

Using a version string is tricky as I pointed in the post as splitting on dots and converting to int may fail. For me at least a tuple that has major/minor as numeric values.

The following would be cool for me

version_info = packaging.version.parse(numpy.__version__).release

Also numpy is taken as an example by many other project in the scientific ecosystem. Or at least if not a tuple, expose a version object that properly compare itself to tuples, so that dependees don't roll their own parsing logic.

rgommers commented 2 years ago

Okay, we can add it I think. As a plain tuple of 3 integers. I'd prefer it be called version_tuple rather than version_info. Wanna submit a PR?

Carreau commented 2 years ago

Okay, we can add it I think. As a plain tuple of 3 integers. I'd prefer it be called version_tuple rather than version_info. Wanna submit a PR?

https://github.com/numpy/numpy/pull/20803