OFFTOPIC Re: [Mailman-Users] Archive URL in postings (2.1b3)

[Chuq Von Rospach]

On Tuesday, October 29, 2002, at 11:33  AM, John Buttery wrote:

>   "Lie" is kinda an interesting way to put it, but I think we all know
> what he means.  How can you say that readdressing a mail (which is what
> you're doing here) isn't a "lie" the way he's referring to it?  It's in
> plain black and white.

we aren't re-addressing. We're standardizing the way the existing 
addresses are presented.

>> um, interesting philosophy. I don't buy it. Not for a moment.
>
>   Well, OK then, how would you differentiate this new behaviour from a
> hypothetical message I might have sent pre-upgrade that _did_ have your
> address in the To: and the list in the CC:?  Hint: you can't.

sure I can. If it went through the list server, it has a List-ID 
attached. So it's clear which one came from where.

> (Of
> course, I acknowledge that sending a copy directly to a person who is
> already subscribed to the list is a bit silly/redundant,

See, I don't agree with that, either. I've explained how I take 
advantage of this in my filtering to continue discussions I'm involved 
in without having to paw through all my list mail...

>   Saying you "don't buy" something that someone has stated is their
> opinion/feeling is indefensible.  :)

yeah, but when a situation is arguably a subjective decision in the 
first place, sometimes it's all you get.

>   When two mails are sent out, I'll get one filtered into my folder for
> the list, and the other one sent to my main inbox; that's the way the
> sender intended it, and that's the way I want it.

and if you filter on list-id, that's how you'll get it no matter who 
does what to the to/cc/bcc. If you DON'T filter on list-id, you depend 
on the end users doing what you expect them to do. That's a filtering 
method guaranteed to fail randomly.

filtering list mail based on to/cc is a flawed approach. to argue that 
we can't break what's broken by definition is a flawed argument. IMHO.

>   True...but should be unecessary.

but it is, because it's the ONLY reliable filtering tool you have. the 
others depend on people doing what you consider "right", so they fail 
when those users don't. If anything, this new header format 
rationalizes that so decisions by the end users can't break the 
filtering, if you absolutely MUST use to/from/cc for some reason.

> Colorization of lines (based on whether your exact address is in the
> headers somewhere) in a message index is another, and List-ID won't 
> help
> you on that one. :)

and if you'll notice, I haven't refuted that one. it's a point that 
needs consideration.


>   But really, this is all getting away from the original philosophical
> point that an address that a message was not sent to should not be 
> faked
> into the headers.

I'd argue whether that's true.

I guess I'm arguing that it's the end of the days of "bulk" mail, and 
the beginning of the days of "mass" mail. There are significant 
semantical differences and advantages. To use paper-mail analogies, 
it's time to move from sending out everything to "Resident", and start 
adding enough intelligence to the system so that it actually shows up 
with your name on it.

>


-- 
Chuq Von Rospach, Architech
chuqui at plaidworks.com -- http://www.plaidworks.com/chuqui/blog/

Yes, I am an agent of Satan, but my duties
are largely ceremonial.