In a message of Mon, 17 Oct 2016 21:41:08 +0100, Ralph Corderoy writes: >Hi Laura, > >> Just giving them utf-8 even though that wasn't what they asked for has >> fixed a huge number of headaches when running mailing lists around >> here. > >Does that mailing-list software check that what they're sending out is >valid for the encoding they claim, UTF-8? Or when replying to an ISO >8859-13 do they send invalid UTF-8 back? > >-- >Cheers, Ralph. >https://plus.google.com/+RalphCorderoy
It does some checking. I'm pretty sure that I could construct some cases that would break it, but there was no great demand for that. While, on the other hand, mail that claims to be US-ASCII but is really UTF-8 happens all the time. The weekly report from the Python bug tracker, for instance, insists that its mail is US-ASCII no matter how many bug reports that I send about the fact that people are signing their bug reports with their non-ASCII names. Things may be different where you are, but around here, when there is a difference of opinion between what the encoding says, and what the content has inside it, and the encoding is US-ASCII, the encoding is always wrong. "You got an 8-bit char in your mail by mistake" is not a common problem here, and, when it does occur, it's not a problem that people care about, or rather they care about it just as much as they do about any other typo in their mail -- if they didn't run the thing through a spell checker, then it wasn't one of those important pieces of mail where perfection in content matters, but more like this one, where if I make a typo, I will trust that you will suffer through it with no long term ill effects. Laura _______________________________________________ Nmh-workers mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/nmh-workers
