New feature request: "faked" identity, and supercede

New feature request: "faked" identity, and supercede

[Steinar Bang]

Platform: Emacs 19.34, RGnus-0.52, SPARC Solaris 2.4, 

One thing I've done, is to change my identity (my address and
organization) for some mail and newsgroups.

I do it, by putting the following into my ~/.emacs:

(add-hook 'message-header-setup-hook 'custom-organization-and-sender)

The custom-organization-and-sender function can be found at the end of
this message.

In short, what I do is set the variables user-email-addres, and
message-user-organization, when I enter message-mode from certain
news, and mail groups.

This works fine for posting news and sending mail.  The place where it
doesn't work, is when it comes to Supercede and Cancel.

Then I get the "This article is not yours", caused by this bit of
message-cancel-news in message.el:
	...
	;; Make sure that this article was written by the user.
	(unless (string-equal
		 (downcase (cadr (mail-extract-address-components from)))
		 (downcase (message-make-address)))
	  (error "This article is not yours"))
	...

The culprit here, is message-make-address, which returns sb@metis.no,
and I guess there is no hacking around this, short of overriding this
function in Gnus (which I don't much like doing).

So... I would like to see a built-in customization mechanism for the
sender address, and organization.  I might even squeeze in the time to
implement it (no promises, but the motorcycle season is over).

However, I'm unsure about where the proper way to implement it.  The
cleanest place would seem to be to have it in the Group parameters.
But I don't know if this would allow me to set the address for whole
ranges of mail and news groups (like the example below).

So, I dunno...


- Steinar

-----------Custom address code------------------------
(defun custom-organization-and-sender ()
  "Set custom organization, and sender on some news groups"
  (cond ((or 
	  (string-match "nnml:mail.private" gnus-newsgroup-name)
	  (string-match "nnml:folk/.*" gnus-newsgroup-name)
	  (string-match "nnml:hobby/.*" gnus-newsgroup-name)
	  (string-match "nnmh:folk/.*" gnus-newsgroup-name)
	  (string-match "nnmh:hobby/.*" gnus-newsgroup-name)
	  (string-match "no.alt.motorsykler" gnus-newsgroup-name)
	  (string-match "nordunet.rec.motorcycles" gnus-newsgroup-name)
	  (string-match "rec.motorcycles" gnus-newsgroup-name)
	  (string-match "rec.aviation.military" gnus-newsgroup-name)
	  )
	 (progn
	   (make-variable-buffer-local 'message-user-organization)
	   (make-variable-buffer-local 'user-mail-address)
	   (make-variable-buffer-local 'message-syntax-checks)
	   
	   (setq message-user-organization "DoD, Norway Chapter")
	   (setq user-mail-address "steinar@bang.priv.no")
	   (setq message-syntax-checks (list '(sender . disabled)))
	   ))
        ((string-match "no.test" gnus-newsgroup-name)
	 (progn
	   (make-variable-buffer-local 'message-user-organization)
	   (make-variable-buffer-local 'user-mail-address)
	   (setq message-user-organization "Testers Inc.")
	   (setq user-mail-address "steinar@bang.priv.no")
	   ))
	))

Re: New feature request: "faked" identity, and supercede

[Edward J. Sabol]

Excerpts from mail: (29-Oct-96) New feature request: "faked" identity, and supercede by Steinar Bang
> One thing I've done, is to change my identity (my address and organization)
> for some mail and newsgroups.
>
> I do it, by putting the following into my ~/.emacs:
>
> (add-hook 'message-header-setup-hook 'custom-organization-and-sender)
[...]
> So... I would like to see a built-in customization mechanism for the
> sender address, and organization. I might even squeeze in the time
> to implement it (no promises, but the motorcycle season is over).

This sounds a lot like a simplified version of the gnus-posting-styles
variable that was in September Gnus but was removed very late in the
September Gnus development cycle when Lars created message-mode to replace
mail-mode and news-reply-mode. At the time, Lars said he would put
re-implementing gnus-posting-styles on the to-do list for Red Gnus, but it
seems to have fallen through the cracks for this development cycle. :-(

Lars, could you raise the priority on re-implementing gnus-posting-styles for
the next development cycle? I can see disassociating it from Gnus and
renaming the variable to message-posting-styles or message-styles or
something like that. However, I thought the basic idea was a good and useful
one that I'd still very much like to see.

Thanks,
Ed

Re: New feature request: "faked" identity, and supercede

[Lars Magne Ingebrigtsen]

"Edward J. Sabol" <sabol@alderaan.gsfc.nasa.gov> writes:

> Lars, could you raise the priority on re-implementing gnus-posting-styles for
> the next development cycle? I can see disassociating it from Gnus and
> renaming the variable to message-posting-styles or message-styles or
> something like that. However, I thought the basic idea was a good and useful
> one that I'd still very much like to see.

Yup.  This is still on the todo list, so it'll be written, sooner or
later... 

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

Re: New feature request: "faked" identity, and supercede

[Lars Magne Ingebrigtsen]

Steinar Bang <sb@metis.no> writes:

> So... I would like to see a built-in customization mechanism for the
> sender address, and organization. 

Well, you could just add the thingies to `message-default-headers' in
a hook?

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

Re: New feature request: "faked" identity, and supercede

[Steinar Bang]

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

> Steinar Bang <sb@metis.no> writes:
>> So... I would like to see a built-in customization mechanism for the
>> sender address, and organization. 

> Well, you could just add the thingies to `message-default-headers' in
> a hook?

But that still won't allow me to cancel my own messages, no?

Would it be different/simpler than my current solution?

Re: New feature request: "faked" identity, and supercede

[Lars Magne Ingebrigtsen]

Steinar Bang <sb@metis.no> writes:

> >>>>> Lars Magne Ingebrigtsen <larsi@ifi.uio.no>:
> 
> > Steinar Bang <sb@metis.no> writes:
> >> So... I would like to see a built-in customization mechanism for the
> >> sender address, and organization. 
> 
> > Well, you could just add the thingies to `message-default-headers' in
> > a hook?
> 
> But that still won't allow me to cancel my own messages, no?

Well, the Sender and Organization header don't affect Gnus' handling
of cancel messages, do they?

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

Re: New feature request: "faked" identity, and supercede

[Steinar Bang]

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

> Steinar Bang <sb@metis.no> writes:
>>>>>>> Lars Magne Ingebrigtsen <larsi@ifi.uio.no>:

>>> Steinar Bang <sb@metis.no> writes:
>>>> So... I would like to see a built-in customization mechanism for the
>>>> sender address, and organization. 

>>> Well, you could just add the thingies to `message-default-headers' in
>>> a hook?

>> But that still won't allow me to cancel my own messages, no?

> Well, the Sender and Organization header don't affect Gnus' handling
> of cancel messages, do they?

Sender, no.  From, yes.

See
	<URL:news:whvibt3brz.fsf@norne.metis.no>

For a clarification.

Re: New feature request: "faked" identity, and supercede

[Sten Drescher]

>>>>> Steinar Bang writes:

>>>>> Lars Magne Ingebrigtsen <larsi@ifi.uio.no>:
>> Steinar Bang <sb@metis.no> writes:
>>> So... I would like to see a built-in customization mechanism for
>>> the sender address, and organization.

>> Well, you could just add the thingies to `message-default-headers'
>> in a hook?

SB> But that still won't allow me to cancel my own messages, no?

	I use the following, both at home and at work, and I can
cancel or supercede my posted messages from either location.

(defadvice message-user-mail-address (before generate-user-mail-address activate compile)
  "Change the value of the user mail address based on the group."
  (setq user-mail-address 
	(cond
	 ((message-news-p) (concat 
			    "stend+"
			    (gnus-short-group-name gnus-newsgroup-name)
			    "@grendel.texas.net"))
	 (t "sten.drescher@tivoli.com"))))

SB> Would it be different/simpler than my current solution?

-- 
+----------------------  Tivoli Customer Support  ----------------------+
|   Sten Drescher                     Tivoli Systems, Inc               |
|   email: sten.drescher@tivoli.com   9442 Capital of Texas Hwy North   |
|   phone: (512) 794-9070             Arboretum Plaza One, Suite 500    |
|   fax  : (512) 345-2784             Austin, Texas 78759               |
+-----------------------------------------------------------------------+

Re: New feature request: "faked" identity, and supercede

[Lars Magne Ingebrigtsen]

Steinar Bang <sb@metis.no> writes:

> Sender, no.  From, yes.

Well, this is a problem.  Perhaps there should be a variable to allow
people to cancel/supersede articles that have differing From headers?
If the From header matches a certain regexp, then the cancel/supersede
message is sent out with the same From header as the article to be
canceled/superseded?

On the other hand, it sounds very, uhm, naughty.

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

Re: New feature request: "faked" identity, and supercede

[Steinar Bang]

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

> Steinar Bang <sb@metis.no> writes:
>> Sender, no.  From, yes.

> Well, this is a problem.  Perhaps there should be a variable to allow
> people to cancel/supersede articles that have differing From headers?
> If the From header matches a certain regexp, then the cancel/supersede
> message is sent out with the same From header as the article to be
> canceled/superseded?

> On the other hand, it sounds very, uhm, naughty.

That's why the posting styles mentioned earlier, might be a very good
idea.  This could be handled quitely, and deep down in the lisp code.

Remember that all "user friendly" news readers and MUAs on systems
like the Mac and MSWin already lets you set the sender freely, so it
wouldn't raise any eyebrows if Gnus allowed the same thing.

The fact that this would also affect cancelling/superseeding, could be
swept quietly under the carpet.

IMHO, of course...!


- Steinar

Re: New feature request: "faked" identity, and supercede

[Steinar Bang]

>>>>> Sten Drescher <sten.drescher@tivoli.com>:

> 	I use the following, both at home and at work, and I can
> cancel or supercede my posted messages from either location.

> (defadvice message-user-mail-address (before generate-user-mail-address activate compile)
>   "Change the value of the user mail address based on the group."
>   (setq user-mail-address 
> 	(cond
> 	 ((message-news-p) (concat 
> 			    "stend+"
> 			    (gnus-short-group-name gnus-newsgroup-name)
> 			    "@grendel.texas.net"))
> 	 (t "sten.drescher@tivoli.com"))))

Interesting... but what is "defadvice"?  I've never seen it before,
and 
	C-h f defadvice RET
came up with nothing.  Neither did I find anything with index search
in Gnus, Emacs, or Elisp.

Re: New feature request: "faked" identity, and supercede

[Steven L Baur]

>>>>> "Steinar" == Steinar Bang <sb@metis.no> writes:

Steinar> Interesting... but what is "defadvice"?  I've never seen it before,

Do a M-x locate-library advice to find the source and read the
comments there.  That's the only documentation.  It's got a steep
learning curve (IMO), but is well worth the effort to learn how to
use.

-- 
steve@miranova.com baur
Unsolicited commercial e-mail will be billed at $250/message.
What are the last two letters of "doesn't" and "can't"?
Coincidence?  I think not.

Re: New feature request: "faked" identity, and supercede

[Lars Magne Ingebrigtsen]

Steinar Bang <sb@metis.no> writes:

> That's why the posting styles mentioned earlier, might be a very good
> idea.  This could be handled quitely, and deep down in the lisp code.
> 
> Remember that all "user friendly" news readers and MUAs on systems
> like the Mac and MSWin already lets you set the sender freely, so it
> wouldn't raise any eyebrows if Gnus allowed the same thing.

You mean the From header?  :-)  (The Sender header is something else,
you know.)

> The fact that this would also affect cancelling/superseeding, could be
> swept quietly under the carpet.

That's a good point.  Well, in any case, it'll have to wait until
Mamey Sapote Gnus.

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

Re: New feature request: "faked" identity, and supercede

[Steinar Bang]

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

> Steinar Bang <sb@metis.no> writes:
>> That's why the posting styles mentioned earlier, might be a very good
>> idea.  This could be handled quitely, and deep down in the lisp code.

>> Remember that all "user friendly" news readers and MUAs on systems
>> like the Mac and MSWin already lets you set the sender freely, so it
>> wouldn't raise any eyebrows if Gnus allowed the same thing.

> You mean the From header?  :-) 

Nah! ...and yes... by "sender" I meant the actual name of the supposed
sender (real name, and email address), ie. what ends up in the from
field. 

> (The Sender header is something else, you know.)

>> The fact that this would also affect cancelling/superseeding, could be
>> swept quietly under the carpet.

> That's a good point.  Well, in any case, it'll have to wait until
> Mamey Sapote Gnus.

You can add another thing to this list:
 it might be a good idea to quietly override a custom email address
 for a domain that doesn't have an MX record, or an A record, in DNS

Leave this fact out of the documentation, and out of the elisp
comments (maybe...).  If someone wants to override this they'll have
to work at it, which would redirect email forger wannabess to Netscape
or something.

Re: New feature request: "faked" identity, and supercede

[David Moore]

Steinar Bang <sb@metis.no> writes:
> You can add another thing to this list:
>  it might be a good idea to quietly override a custom email address
>  for a domain that doesn't have an MX record, or an A record, in DNS
> 
> Leave this fact out of the documentation, and out of the elisp
> comments (maybe...).  If someone wants to override this they'll have
> to work at it, which would redirect email forger wannabess to Netscape
> or something.

	This is a good plan, we can also suggest that all email forgers
use 'metis.no' as a hostname when forging. :)  The current trend of
spammers is to use valid hostnames, except ones they don't have accounts
on.  So it wouldn't help much there.  Now it might be useful to check
something like this just so it can _tell_ the user that they have their
return address configured incorrectly.

	I guess I could see someone wanting their mail reader to check
the addresses of incoming messages for host validity, something called
similar to mc-verify.  You might want to go further and have your
'forgery-verify' routine look at the insertion points (received from, or
path headers) compared with the addresses.  But this seems like a lot of
work, and the heuristics might be wrong often?

-- 
David Moore <dmoore@ucsd.edu>       | Computer Systems Lab      __o
UCSD Dept. Computer Science - 0114  | Work: (619) 534-8604    _ \<,_
La Jolla, CA 92093-0114             | Fax:  (619) 534-1445   (_)/ (_)
<URL:http://oj.egbt.org/dmoore/>    | Solo Furnace Creek 508 -- 1996!

Re: New feature request: "faked" identity, and supercede

[Steinar Bang]

>>>>> David Moore <dmoore@UCSD.EDU>:

> This is a good plan, we can also suggest that all email forgers
> use 'metis.no' as a hostname when forging. :)  The current trend of
> spammers is to use valid hostnames, except ones they don't have accounts
> on.  So it wouldn't help much there.  

I wasn't thinking of stopping spammers.  I was just trying to put some
obstacles in the way of "humorous" cow-orkers sending email to
all@metis.no as elvis@heaven.com or somesuch.

"Let them eat Netscape"
	-- Marie Antoinette

> I guess I could see someone wanting their mail reader to check the
> addresses of incoming messages for host validity, something called
> similar to mc-verify.  You might want to go further and have your
> 'forgery-verify' routine look at the insertion points (received
> from, or path headers) compared with the addresses.  But this seems
> like a lot of work, and the heuristics might be wrong often?

Nah!  Overkill.

Re: New feature request: "faked" identity, and supercede

[Hrvoje Niksic]

Steinar Bang (sb@metis.no) wrote:
> I wasn't thinking of stopping spammers.  I was just trying to put some
> obstacles in the way of "humorous" cow-orkers sending email to
> all@metis.no as elvis@heaven.com or somesuch.

I don't think such obstacles are a good idea.  Do you mean that when I
post from public.srce.hr with `From: hniksic@fly.cc.fer.hr' (which
*is* my valid internet address), that I will be hampered?  Or did I
misread the messages?

-- 
WWW:          World-Wide-Waste.  Waste management corporation, which
              handles the billions of tons of garbage generated by just
              about everybody these days.
You owe the Oracle a good book.  In HyperText, please.

Re: New feature request: "faked" identity, and supercede

[Mark Eichin]

defadvice is *scary*.  It's how debug-on-entry is implemented (at
least in emacs 19.recent, 3x and maybe late 2x, and xemacs) and I've
recently used it as a way to "let" environment variables...

Re: New feature request: "faked" identity, and supercede

[Steven L Baur]

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

Hrv> I don't think such obstacles are a good idea.  Do you mean that
Hrv> when I post from public.srce.hr with `From:
Hrv> hniksic@fly.cc.fer.hr' (which *is* my valid internet address),
Hrv> that I will be hampered?  Or did I misread the messages?

There isn't any sure-fire method of distinguishing spam from the email
headers alone, so my guess is they intend you to be hampered.

Cut it out guys.  Security through obscurity doesn't work.

Doing checks based on DNS isn't going to give you any protection,
since there is no current way for Gnus & Emacs to check the validity
of the DNS data it receives.  (As one example consider a machine with
a tunnelled connection to the internet, running a standalone root DNS
server).

Nearly all of the spam I receive is from stupid Microsoft Windows
losers using Pegasus or Eudora anyway.  Or consider public enemy
CyberPromo and how they doctor their sendmail.cf to pretend to be from
some other domain.  Making Gnus harder to use for legitimate purposes
isn't going to stop any of that.  So why bother to pretend?
-- 
steve@miranova.com baur
Unsolicited commercial e-mail will be billed at $250/message.
What are the last two letters of "doesn't" and "can't"?
Coincidence?  I think not.