search

r-lib / rcmdcheck

Run R CMD check from R and collect the results
https://rcmdcheck.r-lib.org
Other
115 stars 27 forks source link

False negatives due to timestamps in CRAN checks #128

Closed lionel- closed 3 years ago

lionel- commented 4 years ago

Comparing against CRAN checks causes false failures because they include timestamps:

[[1]]
── R CMD check comparison ───────────────────────────── HyperbolicDist 0.6-2 / 0.6-2 ────
Status: BROKEN

── Fixed

✔ checking R code for possible problems ... [21s/24s] NOTE

── Newly failing

✖ checking R code for possible problems ... NOTE

[[2]]
── R CMD check comparison ───────────────────────────────── GlobalDeviance 0.4 / 0.4 ────
Status: BROKEN

── Fixed

✔ checking R code for possible problems ... [6s/8s] NOTE

── Still failing

✖ checking DESCRIPTION meta-information ... NOTE

── Newly failing

✖ checking R code for possible problems ... NOTE

Perhaps these could be safely stripped with a simple regexp?

gaborcsardi commented 4 years ago

Yeah, or we could have an option to not include them at all. I am actually surprised that they are included.

lionel- commented 4 years ago

This seems to solve it: https://github.com/r-lib/rcmdcheck/commit/b5f1fbd7746f675bf475e3a163c55a9042871922

gaborcsardi commented 4 years ago

Right, we should probably remove it from the local checks as well, in case they are requested via an option.