Hi Dirk On Mon, 25 Jan 2021 at 16:13, Dirk Eddelbuettel <[email protected]> wrote: > But upstream had that choice too and selected a warning not a stop. It > appears to be a non-fatal issue.
My deviation from upstream here was intentional. The purpose being to fail the autopkgtest, and our users should never ever see the failure, as autopkgtest would prevent this package from ever migrating to testing. On Mon, 25 Jan 2021 at 16:21, Dirk Eddelbuettel <[email protected]> wrote: > Yes. The cleanest might be a test script that just tries library(glmmTMB) > having set options("warn"=2). However, I do agree that your solution is more elegant, and also does not incur the burden of maintaining a patch. When I get a chance tomorrow, I will try your suggestion in an autopkgtest in addition to the pkg-r-autopkgtest that is already run. Thanks again for your input on this issue! Regards Graham
