Re: mail-followup-to / mail-copies-to

On Thu, 26 May 2005, Bruce Lilly wrote:

> On Thu May 26 2005 16:26, Paul Jakma wrote:
>
> > However, if I would suggest this "list-reply" functionality should be
> > taken as "reply in a list context"
> One problem is defining precisely what that means, in a way which is
> likely to reach consensus.
Agreed.

I think by looking at what MUAs do, one could define that context.

The fact that already several MUAs (popular ones at that, 
thunderbird, apple mail, and mutt) /do/ do see fit to provide a "list 
context", in addition to the long standing "reply" and "reply to all" 
contexts, suggests there may be value in distinguishing this context 
from others.
> That strikes me as equivalent to saying "screw you, I'm going to 
> completely ignore your stated request as well as the relevant 
> standards and do something different for spite".
Ha! To an extent, I'd agree - and it's definitely caused me problems. 
However, just complaining about it though isn't productive, because 
the Mutt developers (and its users) do seem to feel it solves a 
problem.
And it /does/ for mutt. If I were to use Mutt and configure MFT for 
all the various lists, I'd generally be happy - except when non-Mutt 
users replied to me.
So I'd like to try help fix that, somehow, even if it's no more than 
clarifying best practice in some way. However, note that I /agree/ 
"list context reply" should /not/ consider Reply-To, for if it did, 
that would imply it should also consider "From", which obviously isnt 
the idea.
> > I would disagree that the list-reply functionality of some MUAs 
> > means "reply /only/ to list". One of more popular MUA which 
> > implements list-reply supports this view (as it will take MFT into 
> > consideration).
> The problem with that is that it leads to nasty surprises; an invisible
> (non-standard, not displayed) header field changes the way a function
> works, with no warning to the user.
I don't quite agree with MFT either. Please don't take the fact that 
I have this header in my mails as a sign I agree with it - it's there 
for least-worst reasons.
> Beware of some things that you may read; there are bogus claims 
> that "Reply-To prevents direct responses to an author" -- that is 
> simply untrue.  A moment's thought should convince you that given a 
> message:
>  From: fred(_at_)somewhere(_dot_)example(_dot_)net
>  Reply-To: the-mertzs(_at_)example(_dot_)com
>  Cc: ethel(_at_)another(_dot_)example(_dot_)edu
>
> nothing prevents you from sending a response to any of those 
> mailboxes, or indeed to random-person(_at_)some(_dot_)example(_dot_)org(_dot_)  Now it may 
> be the case that *some* MUAs lack a single-key/click/whatever 
> "reply-to-author(s)" function, but that is an implementation issue 
> with those MUAs -- there certainly are several MUAs that do have 
> such a function -- and in no way results from some putative evil 
> characteristic of the Reply-To field.
Yes, I agree. Pine is quite good in this respect as it asks you what 
to do, it makes you think. Other MUAs unfortunately aren't as 
careful.
However, see above, I do not (at the moment) think it would be right 
to consider From/Reply-To in a "list reply" context. The entire 
reason "list reply" exists is because of a perceived desire that most 
people prefer not to get direct copies of list mail, my limited and 
statistically meaningless research seems to support this. :)
Hence, Reply-To can /not/ be used. For reasons you gave previously 
(what exactly does it indicate? A preference to direct public or 
personal mail? Whatever the rights or wrongs of that grey area, it 
exists - otherwise MUAs would not be offering a "list context" reply 
feature).
> No need. Pine already provides the option for a user to reply to 
> the author(s) (From field) or where the author(s) suggested 
> (Reply-To field) or to "group" and/or to any mailbox in the users 
> address book or which the user can type.
Yes, Pine is fine.

I suspect though I would fail if I instead set out to persuade 
everyone to switch to Pine. Activism here (and even drafting a 
proposal, if needs be) is, I hope, likely to be more productive (in 
the long term, I realise).
> > So is there any answer? Without wishing to have you prejudice your 
> > stance on the question(s), would you be able to suggest which 
> > avenues would the most fruitful to investigate?
> I'd first have to see a definition of that "list context" that is 
> somehow distinctive from any other "context" and which makes some 
> sort of sense given the fact that -- no matter what -- a respondent 
> is going to have to decide where he wants to send responses.
Or rather than "where", in what /context/ they wish to respond. 
Choosing the /context/ is easy for an user.
Those contexts are, in every MUA I know of, 'personal', 'all' and, in 
several of the more popular MUAs, 'list':
- personal

        Typically (it seems to me):
                To: (Reply-To) ? Reply-To : From

        Better MUAs prompt the user:
                prompt: "Use Reply-To or From?"

- all

        Typically:

                To: ((Reply-To) ? Reply-To : From)
                Cc: To, Cc

        (some MUAs may prompt)

- list

        Typically (List-Post from header, or statically configured):

                To: List-Post

        Some:

                To: List-Post, (MFT ? MFT : '')

        Minority:

                To: (MFT ? MFT : List-Post)

(apologies for the ad-hoc selection description language, it's 
intended to be vaguely according to C, as if above were headers, and 
anything left of initial : were variables named according to fields 
of replied-to-email along with the C ? operator. Duplicate and empty 
address/header removal after selection is implicit. I'll try figure 
out how to specify above (and below) in the correct way).
My concern is solely with the 'list' case. I would propose it be (and 
apologies again for not yet being familiar with the /correct/ 
language for expressing these things):
- list
        To: List-Post
        Cc: (Foo-Copies-To ? Foo-Copies-To : '')

The 'Foo-Copies-To' header would be additive, any one MAY add their 
address to it, and transitive, if it were present in a mail being 
replied to MUAs SHOULD retain it and its contents.
(I'd look at language of In-Reply-To and References to ensure it 
would be defined in a consistent and similar manner - apologies, but 
i'm not yet familiar with the correct language for this.).
> Up to you.  Unless you have a very clear idea of the problem and 
> the breadth of possible solutions, and have carefully considered 
> the implications, a draft may be premature.
See the above. Would work I think. Or at least, would /allow/ MUAs to 
work in a way that would keep me happy, wrt "list context replies".
> Sounds like the root of the problem is MUAs that implement the 
> "spiteful" stance noted earlier.
Yes, but they implement it because of a perceived problem. As much I 
would prefer that everyone would hit "reply to all" on list mails, I 
realise it's not going to happen - and MUAs now ever more are 
providing "list context" to make it easy for users to only mail the 
list and anyone who has indicated a preference for copies on list 
context mail - the problem is there no way to indicate that 
preference. :)
> Maybe a draft along the lines of "MUAs that do not respect standard 
> field semantics considered harmful" might raise awareness.
It might, but I'd imagine it would be ignored if there is no solution 
to "list context".
> Whether or not that root problem will be fixed is another matter...
Hehe.

regards,
--
Paul Jakma      paul(_at_)clubi(_dot_)ie        paul(_at_)jakma(_dot_)org       
Key ID: 64A2FF6A
Fortune:
If you think last Tuesday was a drag, wait till you see what happens tomorrow!