Refactorcpp

[stevencarlislewalker]

Getting ready to bump the major version by merging this refactorcpp branch. This branch is pretty much backwards incompatible. There are exceptions, but not many. The word 'refactor' in the branch name is not correct, because a refactor implies that the user-facing experience is unchanged -- not true in this case. We set out to do a refactor but then decided otherwise.

[bbolker]

One of the take-home messages from the BIRS workshop session was that having our current development version on a non-default branch is confusing.

What will it take to merge refactorcpp with main/master, e.g.

• update all vignettes to work with this branch (and revert 432985c). That might be a pretty good test (what is our current test coverage stat?
• document levels of back compatibility (i.e. code should still work, but saved model objects won't) and tag the current HEAD of the main branch for people who want to get back to/stay at that point
• maybe release oor to CRAN in preparation for releasing macpan2 ?
• set up workflow for building binary packages and pushing them to drat

?

[bbolker]

FWIW the action is currently failing in the setup stages, apparently while trying to unpack the withr binary. withr was updated on CRAN yesterday, so it wouldn't surprise me if this were a transient problem ...

[stevencarlislewalker]

FWIW the action is currently failing in the setup stages, apparently while trying to unpack the withr binary. withr was updated on CRAN yesterday, so it wouldn't surprise me if this were a transient problem ...

I had the same thought. I'll wait a bit.