Closed gavinsimpson closed 9 years ago
Yep, I think MIT for license
huh, i wasn't aware that we should add LICENSE
to .Rbuildignore
- I don't think check complains about that does it?
@sckott I thought it did if it was in the top level of the sources as R expects arbitrary files to be in inst/
.
ah, sounds good
@gavinsimpson hmm, check throws an error about path to license https://travis-ci.org/ropensci/pangaear/builds/34618941#L685 - how to avoid this when LICENSE
file is in inst/
- this error doesn't happen I don't think when the file is in root dir.
Ahh, my bad @sckott; seems i either misremembered this or things have changed somewhat since I last had to do this. Writing R Extensions states that LICENCE
should be in the top level directory of the sorces, not inst/LICENCE
as we discussed. However, it also talks about having LICENCE
in both the top level "source" and "installation" directories. The only way to achieve the latter is via inst/LICENCE
. So I wonder if we shouldn't put the full MIT licence in pangaear/LICENCE
and the short two-line version in inst/LICENCE
? That way people on Github see clearly what the licence is without needing to know the R ships with generic copies.
I'll have to look at the other commits to see if we need to undo or modify things to make this change.
Okay, thanks @gavinsimpson
R CMD check
warns about an unstated licence as one is not specified inDESCRIPTION
. Also we need to explicitly state what licence the code base is under via an explicitLICENCE
file in the top level.To do:
Licence:
field toDESCRIPTION
LICENCE
file to the top level./LICENCE
to.Rbuildignore