this seems to work fine, not sure if it's a bad fix. but, i had version httr 1 local, and 1.2.1 on the server, and it worked together ok. also see debugging this logic seems to work:
Coverage decreased (-0.04%) to 30.572% when pulling e89f2e14a11d9d0ccd90bbfb24f4ff03fdbbcde8 on mhines-usgs:master into 64a9fcd174bc78bca86cbec57409ed8728860929 on USGS-R:master.
Coverage remained the same at 30.61% when pulling db6d5e967550a9e9bf4d3c685d30a344032f0eac on mhines-usgs:master into 64a9fcd174bc78bca86cbec57409ed8728860929 on USGS-R:master.
Coverage remained the same at 30.61% when pulling 4fa8f60a76066d5c7b8eea3b243d93322fb4bba6 on mhines-usgs:master into 64a9fcd174bc78bca86cbec57409ed8728860929 on USGS-R:master.
this seems to work fine, not sure if it's a bad fix. but, i had version httr 1 local, and 1.2.1 on the server, and it worked together ok. also see debugging this logic seems to work: