All Travis CI osx builds will fall due to known upstream issue.

[bokov]

The data.table team has closed the following ticket with a workaround for an issue someone else reported that seems to be the same as the one we are having:

Rdatatable/data.table#3938 (comment)

Please specify whether your issue is about:

• a possible bug
• a question about package functionality
• a suggested code or documentation change, improvement to the code, or feature request

If you are reporting (1) a bug or (2) a question about code, please supply:

• a fully reproducible example using a publicly available dataset (or provide your data)
• if an error is occurring, include the output of traceback() run immediately after the error occurs
• the output of sessionInfo()

Put your code here:

## load package
library("rio")

## code goes here


## session info for your system
sessionInfo()

[bokov]

Made bugfix branch implementing their change to .travis.yml and pushed it to my fork. In a few hours we'll see if it fixes the osx data.table error. If it does, I will submit a PR.

[bokov]

Travis CI again works for MacOSX!! The developer versions of R still failing for Appveyor. I think maybe Appveyor is not finding the binary versions of 22 packages that rio depends on or suggests.

No idea so far what to do about Appveyor, but here's a PR for Travis-CI.

[bokov]

Here is the build that's working on Mac again:

https://travis-ci.org/bokov/rio/builds/597576597

[bokov]

Closed by #243 , thanks!

[bokov]

I am reopening this issue because apparently the pre-built Mac binary for data.table is now available, and builds will go faster without having to install LLVM and compile data.table.

I undid my changes from #243 in my fork, and that problem no longer happens in the builds (a different one does, unrelated and possibly at the Travis CI end, I will submit a separate ticket for that one).

Do you prefer to roll back PR #243, for me to submit a new PR undoing the previous one, to leave it as-is, or something else?