Date: Mon, 14 Sep 2020 23:09:21 +0200 From: "Steffen Nurpmeso via austin-group-l at The Open Group" <austin-group-l@opengroup.org> Message-ID: <20200914210921.h1m8r%stef...@sdaoden.eu>
| This is of course very primitive handling and does not take care | for RFC 2822, no Sender: handling, etc. That it didn't take 2822 into account, in 199x (for any x) is hardly surprising, as 2822 didn't appear until 2001 (though it was being worked on for a few years before that - not as early as '96 though I don't think). RFC822 was not nearly as clear, and while 2822 wasn't really intended to change anything in this area, just clarify it (unlike other stuff where lots was made obsolete, and a few new things added, to match what people were actually doing with mail at the time) it is understandable that e-mail MUA authors (the very few of them who ever read RFC822, or even knew it existed ... most seemed to code by guesswork) might have not gotten things exactly right. And except in extraordinary circumstances (and Mail having just 2 reply variants doesn't have enough to ever get there) Sender isn't supposed to be used (in almost any way at all) - certainly not in your average reply. A really good MUA might have a "reply to sender" (though it might be phrased differently so people don't just assume "sender" means the address in the From field) as it can occasionally be useful, as in a "Did you really intend to send that message to me?" reply, which is one kind of reply where sending to the addr in the Sender: field (if there isn't one, From: is used) is appropriate, another might be "you typo'd fred's address, you should send another copy to the correct one" (just in case the bounce message didn't get seen). With all of this we're also ignoring anything that we might want to be able to do with the Resent-* fields, which are also not normally used by typical replies, but sometimes can be the right thing to use. But as long as we're confining ourselves to what mailx's 'r' and 'R' commands should do, both Sender and Resent-* are irrelevant. kre ps: Sorry Steffen, I know you're going to see this twice (or perhaps 3 times - but certainly twice) - I forgot this list's obnoxious Reply-To (again!). This time I still had the draft file when I realised though, so this resend includes the list.