Checking `Reply-To'

Checking `Reply-To'

[Hrvoje Niksic]

I've just read the GNKSA review of RadicalNews, and have seen that the
reviewer seems to require that the `Reply-To' header should be checked
the same way as `From'.  Gnus currently checks `From' only (and
remember, that was the rationale for failing twice!).

Regardless of GNKSA, I think Gnus should perform the same checks for
`Reply-To' as it does for `From'.  What do you think?

-- 
Hrvoje Niksic <hniksic@srce.hr> | Student at FER Zagreb, Croatia
--------------------------------+--------------------------------
WWW:          World-Wide-Waste.  Waste management corporation, which
              handles the billions of tons of garbage generated by just
              about everybody these days.

Re: Checking `Reply-To'

[Stainless Steel Rat]

-----BEGIN PGP SIGNED MESSAGE-----

>>>>> "Hrv" == Hrvoje Niksic <hniksic@srce.hr> writes:

Hrv> Regardless of GNKSA, I think Gnus should perform the same checks for
Hrv> `Reply-To' as it does for `From'.  What do you think?

I think this is a violation of RFC822, actually.  It depends on just what
kinds of checking Gnus performs.  A common use of Reply-To (and one of the
intents of RFC822) is to specify a gatewayed mailbox that will work when
the conventional "user@domain" format will fail.  Such addresses have
potentially infinite chance to fail the tests for an Internet "user@domain"
address because most are completely different, and many are technically
illlegal.

-----BEGIN PGP SIGNATURE-----
Version: 2.6.3
Charset: cp850

iQCVAwUBMw+yA56VRH7BJMxHAQFsVQP/eil4uzxCrsRx2UYbPr/7ZDcMrNmMv1nf
3N2aUr5kaL4ywItM9JtXwuuJhGC5dJdjbW/6Rh9dDqUh3QKMbSPQqBl75qDji0kN
RR4Gbv07MYZc0SdtTAL12iuVGz8Jr3zemnCIAayEoJUi98F3qrMkjnMLjhKCzxoc
iQAuIKEAzZY=
=B1at
-----END PGP SIGNATURE-----
-- 
Rat <ratinox@peorth.gweep.net>    \ When not in use, Happy Fun Ball should be
PGP Key: at a key server near you! \ returned to its special container and
                                    \ kept under refrigeration.

Re: Checking `Reply-To'

[Steven L Baur]

Stainless Steel Rat writes:

>>>>>> "Hrv" == Hrvoje Niksic <hniksic@srce.hr> writes:

Hrv> Regardless of GNKSA, I think Gnus should perform the same checks for
Hrv> `Reply-To' as it does for `From'.  What do you think?

I felt the GNKSA/U review of RadicalNews was mean-spirited and the
reviewer had an axe to grind, but that's besides the point.

> I think this is a violation of RFC822, actually.  It depends on just what
> kinds of checking Gnus performs.  A common use of Reply-To (and one of the
> intents of RFC822) is to specify a gatewayed mailbox that will work when
> the conventional "user@domain" format will fail.  Such addresses have
> potentially infinite chance to fail the tests for an Internet "user@domain"
> address because most are completely different, and many are technically
> illlegal.

Yup.  Consider example A.2.4 from RFC822:

     A.2.4.  Committee activity, with one author

             George is a member of a committee.  He wishes to have any
        replies to his message go to all committee members.

            From:     George Jones <Jones@Host.Net>
            Sender:   Jones@Host
            Reply-To: The Committee: Jones@Host.Net,
                                     Smith@Other.Org,
                                     Doe@Somewhere-Else;

        Note  that  if  George  had  not  included  himself   in   the


     August 13, 1982              - 37 -                      RFC #822

 
     Standard for ARPA Internet Text Messages


        enumeration  of  The  Committee,  he  would not have gotten an
        implicit reply; the presence of the  "Reply-to"  field  SUPER-
        SEDES the sending of a reply to the person named in the "From"
        field.

Example A.2.6 has another example containing an address that would
otherwise be considered invalid.
-- 
steve@miranova.com baur
Unsolicited commercial e-mail will be billed at $250/message.

Re: Checking `Reply-To'

[Lars Magne Ingebrigtsen]

Hrvoje Niksic <hniksic@srce.hr> writes:

> I've just read the GNKSA review of RadicalNews, and have seen that the
> reviewer seems to require that the `Reply-To' header should be checked
> the same way as `From'.

That review was the strongest blow to the authority of GNKSA I've seen
so far.  The reviewer wibbled on and on about how much he didn't like
the user interface and what features he wished that it had -- which is
*totally* irrelevant when it comes to minimal Usenet compliance, which
is what the GNKSA was set up to do something about.

> Regardless of GNKSA, I think Gnus should perform the same checks for
> `Reply-To' as it does for `From'.  What do you think?

Well, people shouldn't include Reply-To headers unless they need to,
and if they need to, they probably know what they are doing.

The GNKSA does not require syntax checking of that header, I think.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@ifi.uio.no * Lars Ingebrigtsen

Re: Checking `Reply-To'

[Hrvoje Niksic]

Lars Magne Ingebrigtsen <larsi@ifi.uio.no> writes:

> That review was the strongest blow to the authority of GNKSA I've seen
> so far.  The reviewer wibbled on and on about how much he didn't like
> the user interface and what features he wished that it had[...]

Yup.  The part about the program not being multi-threaded was
infamous.  After several criticisms in private mail (probably by many
people, not just me), the reviewer promised to separate the GNKSA
requirements from his own opinions.

> > Regardless of GNKSA, I think Gnus should perform the same checks for
> > `Reply-To' as it does for `From'.  What do you think?
> 
> Well, people shouldn't include Reply-To headers unless they need to,
> and if they need to, they probably know what they are doing.

You are probably right.  The point of GNKSA are the default settings.

-- 
Hrvoje Niksic <hniksic@srce.hr> | Student at FER Zagreb, Croatia
--------------------------------+--------------------------------
"Psychos _do not_ explode when sunlight hits them."

Re: Checking `Reply-To'

[Nathan O. Siemers]

> 
> Well, people shouldn't include Reply-To headers unless they need to,
> and if they need to, they probably know what they are doing.
> 

	Sorry to break up the politics with a simple question ( have
been off the mailing list for a while, moving across the country,
etc), but the modern gnusen don't seem to honor mail-default-reply-to
or MAILTO.  How *do* I set it?

Note that replies to me may bounce because of this; mail to
siemers@bms.com is preferred ;). 


nathan

Re: Checking `Reply-To'

[Lars Magne Ingebrigtsen]

siemers@bms.com (Nathan O. Siemers) writes:

> 	Sorry to break up the politics with a simple question ( have
> been off the mailing list for a while, moving across the country,
> etc), but the modern gnusen don't seem to honor mail-default-reply-to
> or MAILTO.  How *do* I set it?

You can use `message-default-headers', but why on Earth do you want to
add a Reply-to header?  Why not just make the From header say the
right thing in the first place?

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@ifi.uio.no * Lars Ingebrigtsen

Re: Checking `Reply-To'

[Stainless Steel Rat]

-----BEGIN PGP SIGNED MESSAGE-----

>>>>> "tGBGnB" == Lars Magne Ingebrigtsen <larsi@ifi.uio.no> writes:

tGBGnB> You can use `message-default-headers', but why on Earth do you want
tGBGnB> to add a Reply-to header?

Read RFC-822 some time :).  There are plenty of good reasons to generate a
Reply-To header.  The gist of it is that one uses a Reply-To header when
the sender wishes to direct replies to a mailbox other than the one in the
- From header.

tGBGnB> Why not just make the From header say the right thing in the first
tGBGnB> place?

Because making the From header say "the right thing" may be incorrect
behaviour as far as the mail and news RFCs are concerned.

-----BEGIN PGP SIGNATURE-----
Version: 2.6.3
Charset: cp850

iQCVAwUBMyNAwJ6VRH7BJMxHAQFvlgP/RUTKf9s+Y6KN5Bm4W+KobzG3G5TehH2R
1l2rLQBZXoOAjgxzeaQQ3aKQMGIHAZv/oxMi/rCPuiM9gMPAOgEGnwnJUBVd6Aip
ZEj0Qr1e4JWU1f5PROPEwTCw1AdEnuhfVari59RtM0cW/GvRN4dygBde2DAjRl+n
6aDN1lazRCc=
=mktp
-----END PGP SIGNATURE-----
-- 
Rat <ratinox@peorth.gweep.net>    \ Do not taunt Happy Fun Ball.
PGP Key: at a key server near you! \ 
                                    \ 

Re: Checking `Reply-To'

[Lars Magne Ingebrigtsen]

Stainless Steel Rat <ratinox@peorth.gweep.net> writes:

> tGBGnB> You can use `message-default-headers', but why on Earth do you want
> tGBGnB> to add a Reply-to header?
> 
> Read RFC-822 some time :).  There are plenty of good reasons to generate a
> Reply-To header.  The gist of it is that one uses a Reply-To header when
> the sender wishes to direct replies to a mailbox other than the one in the
> - From header.

Yes -- but that's only useful in very limited circumstances.  The user
wanted to add a Reply-To header that he seemed to want to have the
same contents as his From header.  Many people misunderstand the
meaning of From/Reply-To, and seem to think that if there is no
Reply-To in the mails, then people will be unable to reply to the
mail.

Most people don't need to use Reply-To, and those who do don't need to
ask about it.  :-)

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@ifi.uio.no * Lars Ingebrigtsen

Re: Checking `Reply-To'

[Stainless Steel Rat]

-----BEGIN PGP SIGNED MESSAGE-----

>>>>> "tGBGnB" == Lars Magne Ingebrigtsen <larsi@ifi.uio.no> writes:

tGBGnB> Yes -- but that's only useful in very limited circumstances.  The
tGBGnB> user wanted to add a Reply-To header that he seemed to want to have
tGBGnB> the same contents as his From header.

I guess I missed that, or I figured that he knew what he was doing.

If so, then the Reply-To is both redundant and incorrect.  Like Sender, it
should only exist when the mailbox is different from the From header.

-----BEGIN PGP SIGNATURE-----
Version: 2.6.3
Charset: cp850

iQCVAwUBMySbdp6VRH7BJMxHAQGFfgQAsXrqFIoD16NA6zABS+IBV1JAwo4KUzbG
nxUVwE1ORddfxvUuShEb6M5x1XH7RFRdljzUtf93nJo5CoNh8zvrHbOf3fYllFdp
bbHcTFIX2pgNlgdZEqr9sH7WQJzZiQo2BjRoa9+p9hMHGJTuIiV1GRTgPOLJzauT
fSfquUAH3go=
=xe3u
-----END PGP SIGNATURE-----
-- 
Rat <ratinox@peorth.gweep.net>    \ Happy Fun Ball contains a liquid core,
PGP Key: at a key server near you! \ which, if exposed due to rupture, should
                                    \ not be touched, inhaled, or looked at.

Re: Checking `Reply-To'

[François Pinard]

Lars Magne Ingebrigtsen <larsi@ifi.uio.no> writes:

| You can use `message-default-headers', but why on Earth do you want to
| add a Reply-to header?  Why not just make the From header say the
| right thing in the first place?

The "right thing"!  How do you define it?

I once did exactly that, manage to get a lie out of the `From:' and spare
the usage of `Reply-To:'.  I later reverted to the decision that telling
lies is not the right thing to do.  In particular, I sometimes needed
to sort out from where a message was *really* coming from.  It's better
when headers are plain dependable.  `Reply-To:' is a good invention,
with a purpose.  Just use it when it is proper to use it.

More it goes, more I believe one should not make stunts to help others
about broken software.  If each one was doing what s/he should do,
things would improve much faster.  It is important that the sufferings
be endured by those running bad software, and not imposed to everybody
by those wanting to be overly helping.  Especially in the area of emailing.

P.S. - Consider this.  Is there anything more irritating in life that a
friend who *wants* to help you? :-)

-- 
François Pinard         ``Vivement GNU!''        pinard@iro.umontreal.ca
Support Programming Freedom, join our League!  Ask lpf@lpf.org for info!

Re: Checking `Reply-To'

[Per Abrahamsen]

pinard@progiciels-bpi.ca (François Pinard) writes:

> The "right thing"!  How do you define it?

From: The prefered mail address of the person who wrote the message.

Sender: The address of the mailbox from where the message originates,
	if this is different from the "From: " address.

Reply-To: The mail address where you want replies to this particular
	  message to go, if not to the "From: " address.

In most cases, "From: " should be enough.  "Reply-To: " should only be
used if you want replies /for this particular message/ to go somewhere
different than your usual mail address.

I.e. if I talk about AUC TeX on the net, I may use "Reply-To: " to
direct answers to the mailbox I use for AUC TeX related stuff.

Re: Checking `Reply-To'

[Stainless Steel Rat]

-----BEGIN PGP SIGNED MESSAGE-----

>>>>> "PA" == Per Abrahamsen <abraham@dina.kvl.dk> writes:

PA> From: The prefered mail address of the person who wrote the message.

No, From is the canonical mailbox of the originator (author) of the
message.  Regardless of his preferences, From should contain a canonical
host name for his host or his domain's mail hub.  If his "prefered" mailbox
is something other than this, then he should use a Reply-To.

PA> Sender: The address of the mailbox from where the message originates,
PA> 	if this is different from the "From: " address.

More correctly, Sender is the mailbox of the agent responsible for
submitting the message to the network if it is someone (or something) other
than the Originator.

PA> Reply-To: The mail address where you want replies to this particular
PA> 	  message to go, if not to the "From: " address.

Correct.

-----BEGIN PGP SIGNATURE-----
Version: 2.6.3
Charset: cp850

iQCVAwUBMyWP2J6VRH7BJMxHAQGxnAP/WRdCYTRMU22IvHWSqJQqusMlyEI+a49K
isrb3c+cIV/OaLFlY8Lyh0PRdRJjA/OFViO33rOsELoGAV9mMBV08TucCTMqkfTR
ZDvd4ZRZ+uFLSIswL3mYdSRAlU1Mr/wjHkOwcQqTdLnAQX+rMjyT6K7qTs+NNexs
m5MDvL31DLA=
=k02m
-----END PGP SIGNATURE-----
-- 
Rat <ratinox@peorth.gweep.net>    \ Happy Fun Ball contains a liquid core,
PGP Key: at a key server near you! \ which, if exposed due to rupture, should
                                    \ not be touched, inhaled, or looked at.

Re: Checking `Reply-To'

[Per Abrahamsen]

Stainless Steel Rat <ratinox@peorth.gweep.net> writes:

> No, From is the canonical mailbox of the originator (author) of the
> message.

I -- like many others -- have many email addresses that are equally
valid from a technical point of view.  I prefer one of them, which is
the only thing that distinguish it from the others.  My preferences
(and these alone) thus makes it my "canonical" address.  So we do not
really disagree.

Re: Checking `Reply-To'

[Sten Drescher]

>>>>> Stainless Steel Rat writes:

Rat> -----BEGIN PGP SIGNED MESSAGE-----
>>>>> "PA" == Per Abrahamsen <abraham@dina.kvl.dk> writes:

PA> From: The prefered mail address of the person who wrote the
PA> message.

Rat> No, From is the canonical mailbox of the originator (author) of
Rat> the message.

	No, From is the identity of the person(s) who wished the
message to be sent.  My 'identity' as stend@sten.org has quite
different associations than my 'identity' as sten.drescher@tivoli.com,
so the From header should include the 'identity' I am operating as.


-- 
+----------------------  Tivoli Customer Support  ----------------------+
|   Sten Drescher                     Tivoli Systems, Inc               |
|   email: sten.drescher@tivoli.com   9442 Capital of Texas Hwy North   |
|   phone: +1 512 794 9070            Arboretum Plaza One, Suite 500    |
|   fax  : +1 512 345 2784            Austin, Texas 78759               |
+-----------------------------------------------------------------------+
It is by caffeine alone I set my mind in motion, it is by the beans of
Java that thoughts acquire speed, the hands acquire shaking, the
shaking becomes a warning, it is by caffeine alone I set my mind in
motion. -- Carlos Nunes-Ueno, 3/29/95

Re: Checking `Reply-To'

[Stainless Steel Rat]

-----BEGIN PGP SIGNED MESSAGE-----

>>>>> "PA" == Per Abrahamsen <abraham@dina.kvl.dk> writes:

PA> I -- like many others -- have many email addresses that are equally
PA> valid from a technical point of view.  I prefer one of them, which is
PA> the only thing that distinguish it from the others.  My preferences
PA> (and these alone) thus makes it my "canonical" address.  So we do not
PA> really disagree.

If the mailbox specified in the From header does not point back to the
originating host or originating site's mailhub, then the address is not
canonical as far as RFC-822 is concerned.

If I send mail from peorth (my laptop) with "ratinox@ccs.neu.edu" as the
- From mailbox, it is wrong even though the mailbox is a valid mailbox,
because "ccs.neu.edu" is not a canonical name for peorth in any way, shape,
or form.  If I want replies to a particular mail to go to the
ratinox@ccs.neu.edu mailbox I should have ratinox@peorth.gweep.net in the
- From header and I would put ratinox@ccs.neu.edu in the Reply-To header.

ratinox@ccs.neu.edu in the From header would be a forged From header.

-----BEGIN PGP SIGNATURE-----
Version: 2.6.3
Charset: cp850

iQCVAwUBMyXfMp6VRH7BJMxHAQE18gQAm15BYyisAB4Yox2KkMICCecDESpFmu+d
GVpSQRLNzg7TxmvpYmhwH3u//2rzPVf/DQRm/6ZWO92rk2BRqtBaQ2pjSfGIN5sq
emr4GqkgkqoUi+T06hd4oIbdVK/Vb1dxhpCIhIs/HzGwhlfSEUsciIjPh70YQpLO
r1g4KnKLp/Q=
=9HCL
-----END PGP SIGNATURE-----
-- 
Rat <ratinox@peorth.gweep.net>    \ Ingredients of Happy Fun Ball include an
PGP Key: at a key server near you! \ unknown glowing substance which fell to
                                    \ Earth, presumably from outer space.

Re: Checking `Reply-To'

[Stainless Steel Rat]

-----BEGIN PGP SIGNED MESSAGE-----

>>>>> "SD" == Sten Drescher <sten.drescher@tivoli.tivoli.com> writes:

SD> 	No, From is the identity of the person(s) who wished the
SD> message to be sent.  My 'identity' as stend@sten.org has quite
SD> different associations than my 'identity' as sten.drescher@tivoli.com,
SD> so the From header should include the 'identity' I am operating as.

If my boss wants me to write up something and send it out, he is not the
originator, I am.  If my boss writes up something that he wants me to send
out, he is the originator and I am the Sender.  Desire has nothing to do
with what goes in the From header; who writes the message determines that.

If you are originating at tivoli.com then your 'identity' is sten.drescher
and your mailbox is sten.drescher@tivoli.com; if you are originating at
sten.org then your 'identity' is sten and your mailbox is stend@sten.org.
Yes, they are two different mailboxes associated with two different
'identities'.  This is one of the reasons for the Reply-To header.  If you
want to send mail from tivoli.com but want replies to go to your sten.org
mailbox, you should have 'sten.drescher@tivoli.com' in your From header and
'stend@sten.org' in your Reply-To header.

That is the easiest way to do it.  Alternately, an equally correct setup
would be to put 'stend@sten.org' in the From header and
'sten.drescher@tivoli.com' in the Sender header, with the former identity
being the originator and the later identity being the one responsible for
submitting it to the network.  However, this may not work if your site's
mail hub is in the habit of rewriting outgoing headers.  It can also break
with some MUAs (Mickey$oft Mail) that honor Sender over From.

-----BEGIN PGP SIGNATURE-----
Version: 2.6.3
Charset: cp850

iQCTAwUBMyXiYp6VRH7BJMxHAQHwnAPw99lhcwaZc4NHUnz2XEiymTqZS6sxE/m/
KU89QV+pxR++1nCkTi7XUe8fMNGYeAy9gttdxwlhobAPgtAvDXV6jI9/VUiFk6qp
LldnOMyp2/t2WvEPAkErjbE11Ky9BvsUCbgF+e9oo4OdmE8+oaGF1EcuW4upjyhH
5qVNCMOK
=COXM
-----END PGP SIGNATURE-----
-- 
Rat <ratinox@peorth.gweep.net>    \ When not in use, Happy Fun Ball should be
PGP Key: at a key server near you! \ returned to its special container and
                                    \ kept under refrigeration.

Re: Checking `Reply-To'

[Sudish Joseph]

Stainless Steel Rat writes:
> If the mailbox specified in the From header does not point back to the
> originating host or originating site's mailhub, then the address is not
> canonical as far as RFC-822 is concerned.

RFC 822 has no such restrictions.  It's only suggestions are
wrt. default entries, entries generated by *machines* when the user
has /no particular preference/.  In this latter case, it says From:
should contain what Sender: would otherwise.

The contents of From: are entirely up to the user.  There is no
concept of "canonical" identity in 822.  

822bis includes a "SHOULD NOT" regarding the gratuitous use of
Reply-To:.  It also snips the "typical uses" statements that are
interpreted by some as law in the description of Reply-To.

More interestingly, 822bis promotes the usage of References: for
threading.  Yay.

-Sudish

Re: Checking `Reply-To'

[Stainless Steel Rat]

-----BEGIN PGP SIGNED MESSAGE-----

>>>>> "SJ" == Sudish Joseph <sj@eng.mindspring.net> writes:

SJ> 822bis includes a "SHOULD NOT" regarding the gratuitous use of
SJ> Reply-To:.

822bis is not an accepted Internet mail standards document.  It proposes
some good things, but it is not the baseline to which MUAs should adhere.

-----BEGIN PGP SIGNATURE-----
Version: 2.6.3
Charset: cp850

iQCVAwUBMybsPJ6VRH7BJMxHAQETigQAn8xsM3ZYjXnKOdAL0Xau4uIHJix39Hb4
qedl2XO6xNbAC7LaHnqaGn4s/kGGjF8NTgB7zlIKo/CwMOayY9czseA8qluOZ+bX
DuEJymahN2FxPvjr22WSO5SemiC6RkuoNQAPSn/oc8isORuZ+2pSqjm6cWICfC8C
8clPv85QLSI=
=ZaBa
-----END PGP SIGNATURE-----
-- 
Rat <ratinox@peorth.gweep.net>    \ Caution: Happy Fun Ball may suddenly
PGP Key: at a key server near you! \ accelerate to dangerous speeds.
                                    \ 

Re: Checking `Reply-To'

[Sudish Joseph]

Stainless Steel Rat writes:
SJ> 822bis includes a "SHOULD NOT" regarding the gratuitous use of
SJ> Reply-To:.

> 822bis is not an accepted Internet mail standards document.  It proposes
> some good things, but it is not the baseline to which MUAs should adhere.

Ah, but that was included just FYI.  The problems lie in your
interpretation of 822 as is stated in the first two paragraphs.

-Sudish