conda-forge / jupyter_core-feedstock

A conda-smithy repository for jupyter_core.
BSD 3-Clause "New" or "Revised" License
2 stars 21 forks source link

Release 4.4.0 #11

Closed takluyver closed 7 years ago

conda-forge-linter commented 7 years ago

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

jakirkham commented 7 years ago

@conda-forge-admin, please rerender.

jakirkham commented 7 years ago

Also does noarch: python make sense here?

takluyver commented 7 years ago

Probably yes.

takluyver commented 7 years ago

@conda-forge-admin, please rerender.

takluyver commented 7 years ago

Looks happy now. Anything we should check before merging it?

minrk commented 7 years ago

:shipit:

jasongrout commented 7 years ago

How long should we give people (including people in restricted sys-admin administered environments) to upgrade from conda 4.2? FYI, rendering this package as noarch makes the package break for conda 4.2.

I switched ipywidgets back to not being noarch because conda 4.3 has been out less than a year, and I consider a year about the right amount of time to give people to upgrade to 4.3.

minrk commented 7 years ago

@jasongrout good question. conda fails with:

Installing conda-forge::jupyter_core-4.4.0-py_0 requires a minimum conda version of 4.3.

But it certainly would be nicer if the solver had pip's 'Requires-Python' behavior, where it would pick the latest-supported version, rather than picking a version it can't install and then balking. If it still installed but didn't get the latest version, I would have no reservations about requiring 4.3.

I'm not sure how much we can do about it, since noarch: python is already starting to spread throughout conda-forge. At some point, one of our dependencies is going to pick up noarch, and at that point there isn't any benefit left to us avoiding it, I think.

One year is only in a month, so I'm okay holding off. I'm also okay following conda-forge as saying "conda forge is now adopting noarch, that means conda 4.3", but this is only implicit now because it is available.

@jakirkham thoughts on when conda-forge packages should bump the minimum required conda? Should this be a per-package decision, or is conda-forge officially dropping support for conda 4.2, now that noarch: python is available?

jasongrout commented 7 years ago

Installing conda-forge::jupyter_core-4.4.0-py_0 requires a minimum conda version of 4.3.

Huh, I didn't see that in a test environment where this package was being installed as a dependency - it just failed the installation when it couldn't import jupyter_core in the test phase of the installation, with no such error. Perhaps I was testing in an older version of conda 4.2?

@jakirkham - an official decision and announcement would be very helpful to make sure people are aware of the ramifications of moving to noarch.

jasongrout commented 7 years ago

If it still installed but didn't get the latest version, I would have no reservations about requiring 4.3.

Agreed.