RLS: 0.18.1

[jreback]

if we can finish up remaining issues, I'd like to release this next sunday/monday (may 1 or 2nd).

Planning on an 0.18.2, so if something is not in then we can address then.

[jreback]

https://github.com/pydata/pandas/releases/tag/v0.18.1

conda install pandas -c conda-forge now does automated builds for the releases! (of many projects)

(will be in the main channel soon as well)

[jreback]

announcement

its not on main conda yet, but

conda install pandas -c conda-forge works nicely

[jorisvandenbossche]

@jreback Regarding the planning of a 0.18.2 release, I would want to come back to the discussion we had on the mailing list: https://mail.python.org/pipermail/pandas-dev/2016-March/000490.html

The milestones now have a timeline of

• 0.18.0: released March 13
• 0.18.1: released May 5
• 0.18.2: slated for June 14
• 0.19.0: slated for August 31

Of course that is only provisional, but lets say that is what we aim for. That means that the actual development time for 0.19 (the time that PRs can be merged into master) is only 2.5 months, while we took more time for the bug fix releases.

Some changes (eg deprecating Panels?) can use enough time in master, and also some other things like the core.common refactor don't need to be in a bugfix release.

So coming back to the mailing list discussion, we could try it out to have a separate branch for 0.18.x and master can already be for 0.19.0 ? The we don't need to hold up PRs that we deem unsuited for a bugfix release.

Side note: eventually, I think, we will have to switch to a model with branches for bugfix releases, I don't think we will be able to avoid that, certainly when aiming at a 1.0 release. So I think it would be good to try this out before we really need it, so we have the time to test which approach suits pandas/us best.

[jreback]

@jorisvandenbossche the 0.19.0 was an old milestone date. hadn't moved it in a while. I suspect we will need about 4 months (as is typical) for that one.

I don't think having a separate branch is particularly helpful at this time. Its simply more work for very little gain.

Further refactoring non-user code has to happen. I don't think its relevant what release it happens in. Since its touching a lot of code (and is going to be needed for other fixes), it makes sense to put in now.