Yeah, but, that setting affects reception, not transmission. As in, if you disable e-mailed PM notifications, then the PMs you receive won't be duplicated over SMTP, but the PMs you send can still be.

Yup. That's kind of my point: If very few people have a genuine need (or at least, a deep appreciation) for the ability to be out-of-band sent HTML-stripped versions of the PMs they receive, then, I think the forum should take the message content itself out of any e-mailed PM notifications (for all the reasons I mentioned).

If it were up to me, I'd take out the subject, the sender, and the instant-reply link, too, but, I can't see getting a change like that past theymos.

Agreed. And it's something that I've considered doing more than once...

There are three stumbling blocks (that I can see):

(*) It would involve JavaScript and move Bitcointalk even further toward a state of not being able to work without scripting. (This doesn't bother me any, but, I'm aware that there are some no-JS folks out there that really bristle at being forced to enable browser scripting. I don't find their stance to be realistic, but, I can't say that I blame them for feeling the way that they do: Most programmers, and especially web developers, seem to have no problem with relying on a mutating nest of dependencies that they could never have written on their own, and therefore can't fully understand. You shouldn't accept a vouch from someone when it's about something that they don't understand. If you can't program a given thing from scratch, then you don't understand it.)

(*) It would break PM search. (But, I don't see this as a huge problem. When I originally made this topic, I was working on a filter-by-user patch for PMs. That patch slipped through the cracks and I forgot about it, but, I left it in a close-to-finished state, and if I finish it now and manage to convince theymos to merge it, then, I could see a lack of server-side PM search being much less annoying. Eventually, I could implement client-side search based on server-side user-filtering, but, its first-use/uncached bandwidth usage would depend on how many PMs you've sent to or received from that particular user. I've also got some ideas around re-basing the whole PM system on top of a rank-dependent amount of per-user API-accessible storage, and I could make something like that work really efficiently, but, those ideas are too involved to unpack here.)

(*) I forget the third point I was going to make. It was prolly good, though.

Anyway, when something gets complicated enough that I either can't see a way to very safely splice it into the existing software design, or I can see a way to do that but I expect it to be a huge uphill battle to get it merged, then, my energy wanes and I try to turn my attention back to very small improvements that don't leave much room for argument.

I think what a lot of people don't really understand about me is that I'm in a very particular "mode" when I'm on Bitcointalk: I very rarely suggest (or code) the things that I personally want, because I realize that the things I want are radical, and I don't have the energy to argue for them in what I perceive to be a very change-resistant environment (I don't only mean the user base; I'm also referring to theymos, because, ultimately, things come down to, or are at least very affected by, what he personally likes and dislikes). I don't begrudge theymos his iron grip on Bitcointalk, because I understand it, and my own grip would be at least as tight if I were in his position, but, it leaves me in a situation where I know that I'm not going to be able to get things over a certain complexity-limit or even with a certain flavor past him. Unfortunately, I also know that I'm not really built for the kind of work that I get to do for the forum, and so I'm almost certainly going to run out of interest at some point and move on to things that I actually find stimulating (or at least ideologically satisfying). So, I'm stuck with the problem of how to intelligently ration out my dwindling supply of energy so that I can get the most amount of "good" done while I'm still around to affect things (not only that, but, I also have to make my decisions as smartly as I can in the presence of a tech lead that seems to lean very heavily toward inaction, and a community that sometimes makes either the mistake of engaging in far too much wishful thinking given the status quo, or the mistake of encouraging inaction by discussing things to death, instead of just saying: "Yeah, that would be an improvement. +1").

Yup. That's the basis of that argument (not being signed-in). Like you, I have no need to read PMs without also being signed-in to Bitcointalk, but, like I said, I'm playing devil's advocate with all three of my arguments against implementing this change.

Yup. That came up in a private conversation I had a while ago about this. Like most of the diffs I share on the forum, my expectation is that theymos will re-imagine them in terms of his own source tree (as in, I can't see anything besides 1.1.19, so it's often the case that my diffs are "wrong", but, he knows that, and can account for it).