Re: on a buffer performance test on GNU EMACS 28

[Eric Abrahamsen <eric@ericabrahamsen.net>]

On 08/19/20 22:22 PM, Rafi Khan wrote:
> Eric Abrahamsen <eric@ericabrahamsen.net> writes:
>> To my embarrassment, I don't know how it works, either. I use Gnus as a
>> newsreader, and access these emacs.* groups via nntp. But sometimes I
>> followup using "F", and sometimes I reply with "S W". I'm often not even
>> aware of which I've used. In the latter case, I must be sending email to
>> ding@gnus.org, but it all ends up looking the same to me.
>
> I never really understood the difference between "F" and "S W"... could
> you explain?

Gnus can send messages via two methods: "news" (which is done directly
by an nntp server), and "mail" (which requires separate SMTP
configuration). In the Summary buffer, you've got multiple
semi-redundant commands which do the same thing for either news (called
a "followup") or mail (called a "reply"). Because followup is only ever
sent to a newsgroup (not multiple recipients), it only has two flavors
-- with or without quotation -- on "f" and "F". A mail reply can be
"plain", "wide" or "very wide", and also with or without quotation,
leading to a total of (*counts on fingers*) six bound commands. The full
complement is on the "S" prefix key, but "r" and "R" also do the most
obvious.

Where I'm less certain is what happens when you mail reply in an nntp
newsgroup. In the case of gnus.general, the address "ding@gnus.org" is
always in the cc, so if you do any of the (very) wide reply options,
Gnus apparently knows that your message will also go to the group, and
it lets you send the mail reply (as opposed to the news followup). If
you try a non-wide mail reply (which would not go to the group) it
assumes you're making a dumb mistake and asks for confirmation.

I don't know how much of this is standard nntp stuff, how much
particular to Gnus the newsreader, and how much particular to
gnus.general the newsgroup.

Hope that didn't make things worse!

Eric