Cloning of a git repo with files that has to be autoconverted to IBM-1047 fails with error:
error: failed to encode 'buckets/sdb/v2s40/plxsrc/plx00314.plx' from UTF-8 to ibm-1047
So issue mentioned is:
the file hasn't been converted, but the tag does seem to have been changed. So now I have a file that has an IBM-1047 tag, but is really UTF-8, so when I open it in vim, I see gibberish.
More info in
https://ibm-systems-z.slack.com/archives/C039TNSH9J5/p1773079327579579
So suggested fix is : avoid setting the tag in the case of an error