The new linux-il - a few tips to get you (re)started

[Nadav Har'El]

Hi Shachar and Oleg, I decided to reply now just so it be known that Oleg's
opinion isn't bizarre or outlandish, but rather an opinion held by many.
I agree with every word that Oleg said on this subject.

On Sat, Jan 31, 2009, Shachar Shemesh wrote about "Re: The new linux-il - a few tips to get you (re)started":
>...
> It is an interesting notion. One I totally and utterly disagree with. 
> When I reply to your email I am replying to YOUR email. It is not the 
> "mailing list", an abstract being with no intelligence, that I am 
> conversing with. It is Oleg Goldshmidt. If you will please read your own 

When I was a child, one of the most important rules of etiquette that I
learned was that there should be "no secrets in company" (loosly translating
from the Hebrew), i.e., if you're conversing with a group of people - be it
at a table eating a meal or in our case in a mailing list - it is impolite
to start private conversations. What is the rationale of this rule? When
people start whispering (in our case, mailing directly), members of the
group excluding from the private discussion feel bad. Moreover, the group
breaks up as people start making every conversation private and never
including everyone. For example, why would anyone want to be part of a
mailing list that only has questions, because all the answers are public?
Or a mailing list where no discussion is every concluded, because the
conclusions are always reached in private discussions? Would you want to
be in a mailing list where a few people get the list's message more quickly
then you do (because they're on the CC - Shachar wants this to happen) so
whenever you read some thread, it is was already hashed to death by the
original poster and one or two others?

I strongly feel that mailing lists should have just one mode of communication:
you read a mail on the mailing list, and reply to it to the entire list.
Only in rare situations will you decide to write a direct mail to the author
of the message and not to the whole list. There is no reason whatsoever
to make a list of a few people who should get another copy of this reply.
There is no reason whatsoever to allow people who are not subscribers to
get this mail. There is no reason whatsoever to cross-post a discussion
(like Oleg said, the situation for an announcement would be different, but
also there, after the initial announcement the discussion should be separate
on each mailing list). There is no reason whatsoever for anyone to get
part of a discussion thread to his main inbox instead of the mailing list
folder (because this way, you are getting only part of the discusion, out
of context, encouraging you to answer without reading other people's
opinions, which is impolite).

To summarize, I consider that "reply-to considered harmful" document, and
Shachar's opinions that are close to it, flat-out wrong.

But frankly, I don't really mind that each mailing list be run the way
its administrator believes in. I will run my own mailing lists the way I
believe, and Shachar and Ely will run theirs the way they like. This is
perfectly fine with me, and I don't have to like every detail of how a
mailing list is run to subscribe to it. But what I REALLY DON'T LIKE is
when people start confusing their opinions with progress. The "no reply-to"
isn't the "standard way" or "modern way" or "better way" as some people
paint it. It is a way, which some people prefer, and some people prefer
the exact opposite.  Let's be honest about it.

> email, it is not addressing my email in the third person. It addresses 
> it in the second person. Even your email is written to Shachar Shemesh, 
> while keeping in mind that the entire list reads it. It is not written 
> to the list, hoping that Shachar will read it. When you write:

I don't think it is so clear-cut. A list which is like you say - two-person
conversations which other people incidentally overhear, will not be a good
mailing list. In a good mailing list, people write postings with everyone
on the list as the intended audience - even those who don't answer or simply
haven't answered yet (someone can join a discussion thread late in the game,
simply because he was asleep when it started, or whatever).

> You do not seriously think that the list has been around when Usenet was 
> common, nor even that the majority of this list's subscribers were. You 
> think that I, Shachar, have been.

About the majority, I don't know. I know that I have, and I believe that when
I first heard about this list (circa 1996 I believe?) Usenet was still quite
popular, albeit declining. I don't know why it matters, though. Does the fact
that we no longer write letters via post, mean that none of the lessons learnt
in that medium has any implications on newer media?

> To me, this pretty much pulls the rug from under your reasoning. If I 
> need to answer you in private (say, because what I have to say is based 
> on my personal knowledge of you, and is too personal for the entire list 
> to know), then that very same use of the second person in the email will 

This should be a very rare exception. If it is not, then something is broken
in this mailing list or its subscriber community. We should care about the
common case, not the rare exception. Try to think yourself what percentage
of your replies to linux-il go to the list (Reply-to-all) and what percentage
goes only to the person (Reply) - I think (or at least would like to hope)
that the percentages will be overwhelming toward the first one.

> Either way, the "forward" button is not a replacement to the "reply" 

I think Oleg over-emphasised the "forward" button. In mutt, for example,
when I press "reply" and there is a "reply-to", it asks me whether I want
to reply to the reply-to address (the list), and the default is "yes". If,
however, I press no, then I get to send a reply to the person. This is how
defaults should be - the default should be the common case, and the non-
default is the rarer case.

> And, like I said in my previous email, if I get it wrong, the failure is 
> catastrophic.

This is the only objection raised by the anti-reply-to crowd which I
ever understood. This is correct, but it has to be weighed against all
the other benefits (personal and social) of having a reply-to.
Personally, I think that even this argument is flawed: if you do have
separate "reply to all" and "reply" buttons (like you advocate), you can
still accidentally click "reply to all" when you didn't intend to.
In fact, I see this happening all the time on the corporate email where
I work (someone writing an announcement to 100 people, and then one
silly person accidentally replies to everyone instead of the announcer).

Nadav.

-- 
Nadav Har'El                        |     Saturday, Jan 31 2009, 7 Shevat 5769
nyh at math.technion.ac.il             |-----------------------------------------
Phone +972-523-790466, ICQ 13349191 |If marriage was illegal, only outlaws
http://nadav.harel.org.il           |would have in-laws.