Release 5.4

[blink1073]

Fixes #861

[blink1073]

@minrk, good to go?

[MSeal]

Awesome! A lot of people who talked to me at JupyterCon will be happy this is releasing

[MSeal]

We hopefully can get into faster cadence for this moving forward. I'll try to get a few small items pushed for a later release (not close enough to ready for this one)

[Carreau]

setup.py does not seem to have the python_requires metadata.

[Carreau]

Any objection to give @MSeal push right on PyPI/conda-forge ?

[blink1073]

No objections from me!

[blink1073]

I'm afk, so couldn't make the python_requires change until later.

[Carreau]

I'm afk, so couldn't make the python_requires change until later.

See #871

[Carreau]

In which I also add the question of whether we want to drop 2.7....

[MSeal]

Maybe with the major release so we can keep hotfix paths open in an organized manner? I think there's a lot of python 2 users still based on issues / conversations I've seen passed through papermill

[Carreau]

Maybe with the major release so we can keep hotfix paths open in an organized manner

Seem like a good plan. Hopefully with the python_requires python 2 users should still get a version of nbconvert that works on their machine. We also need to keep in mind that kernels can still by python 2 as well.

[blink1073]

Ready to roll. Happy to release tomorrow or defer to @MSeal.

[MSeal]

Either way -- though I'm not on the maintainers list yet.
I'm assuming the project releases without any fancy setup calls (contributing.md doesn't specify if anything beyond bdist + sdist + twine is needed).

[Carreau]

It's always a good thing to have Someone new going through the release to see if documentation is imprecise/missing :-P

[mpacer]

@MSeal I can be around tomorrow online to help work through these steps. Agreed with @Carreau that it's always good to have someone new going through the release process.

I'm going to give MSeal rights on pyPI since it sounds like everyone else was on board with that.

[blink1073]

Sounds good. I also created conda-forge/nbconvert-feedstock#20.

[Carreau]

May want to also set description and long_description in setup.py otherwise the new PyPI page looks a bit blank.

[MSeal]

Ahh good point -- I can make that change.

Had to have M help with #872 so I didn't get to pair on releasing today

[Carreau]

Had to have M help with #872 so I didn't get to pair on releasing today

No problem and no pressure. Just trying to make high quality software, and as good wine it needs to mature :-)

[SylvainCorlay]

Is there any other blocker for a release?

[MSeal]

I need to add the description and long_description, was going to tackle it tomorrow (today is a holiday here) and see if M is around to pair on releasing.

[MSeal]

We're putting together the changelogs -- adding some documentation notes for the next new person to run through this process.

[blink1073]

This PR includes change logs

[MSeal]

There are some late additions missing

[blink1073]

Ah, just making sure you didn't go through the same pain I already did. 😉

[MSeal]

Got caught up in work tasks today. M was helping me describe the major changes (there was a lot over the past year to categorize). Will endeavour to actually release with completed release notes tomorrow.

[Carreau]

Will endeavour to actually release with completed release notes tomorrow.

It's already tomorrow. Go to sleep.

[blink1073]

Ha, what Matthias said. Thanks M and Matthew!

[Carreau]

Apparently setuptools and PyPI do support

project_urls={
            "Bug Tracker": "https://bugs.example.com/HelloWorld/",
            "Documentation": "https://docs.example.com/HelloWorld/",
            "Source Code": "https://code.example.com/HelloWorld/",
        }

https://github.com/pypa/warehouse/issues/1419, https://github.com/pypa/setuptools/pull/1210

[Carreau]

I pushed a short-ish long description plus a couple of other things.
Please see the latest additions.

[SylvainCorlay]

The test failure comes from long_description having been added to the master branch as well.

@blink1073 @Carreau would you like to fix this? I am eager to get this in...

[Carreau]

Feel free to push a commit removing long description. Thanks for looking into this I was baffeled. Currently In a car so won't do it before a couple of hours.

…

On Wed, Sep 5, 2018, 14:41 Sylvain Corlay ***@***.***> wrote: The test failure comes from long_description having been added to the master branch as well. @blink1073 <https://github.com/blink1073> @Carreau <https://github.com/Carreau> would you like to fix this? I am eager to get this in... — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#869 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAUezyahPHaf8b0M-cb-KjEgLbusWQliks5uX8aCgaJpZM4WSLle> .

[blink1073]

I also won't be able to get back to this until later today.

[SylvainCorlay]

Are you fine with me merging and fixing in master?

[SylvainCorlay]

OK, going ahead with it :)

[SylvainCorlay]

Leaving it to M to tag 5.4!

[mpacer]

Matt and I will take another attempt at getting this out today. We got really far for the amount of time we had to work on it yesterday. Now, it’s mostly crafting the notes about the major user-facing feature and convention changes so we can send out the announcement email.

[mpacer]

In the future I’d prefer not to merge changes that break our tests on master, regardless of how urgently any of us want a feature to be released.

[SylvainCorlay]

Matt and I will take another attempt at getting this out today. We got really far for the amount of time we had to work on it yesterday. Now, it’s mostly crafting the notes about the major user-facing feature and convention changes so we can send out the announcement email.

Thanks! Looking forward to it.

In the future I’d prefer not to merge changes that break our tests on master, regardless of how urgently any of us want a feature to be released.

I get it, although the required change was trivial, tested locally, and allowed us not to wait for the original PR author to rebase his PR, as he was not available.

[mpacer]

@SylvainCorlay You could have pushed onto this PR rather than pushing onto master. That I would have been totally ok with :).

That's the beauty of the allow edits from maintainers option from github!

[SylvainCorlay]

Yeah I wanted to do that, but the build error came from merging the branch into master where the same extra argument had already been added at a different line, so I could not have fixed it by changing this branch.