Ksyriums

"Matt O'Toole" <[email hidden]> wrote in message "]news:[email hidden]...

Quoted message said:

"Simon Brooke" <[email hidden]> wrote in message
"]news:[email hidden]...

Quoted message said:

Tim McNamara said:

[email hidden] writes:

> Long live TIN, emacs and UNIX!

You use Emacs with TIN rather than Emacs's built-in newsreader, Gnus? How come? I don't have
the capacity to compile a version of TIN for Mac OS X 10.1.5, so I haven't been able to try
it. I've had the opportunity to try slrn which is pretty good, but currently am vacillating
between Emacs/Gnus and MT-NewWatcher. Emacs/Gnus is far more configurable for personal
preference than any software I have ever seen...

I used to use Emacs for everything. These days I've switched to using a gooey 'integrated
development environment' for writing software, and a whizzywig word processor for writing
letters... but I still read the gnine o'clock gnus, because I've never found a news client to
better
it. For me, Emacs still rules, at least on Usenet.

emacs is awesome, but a real PITA to learn. I've tried several times,

made New

Quoted message said:

Years' resolutions, given up GUIs for Lent, etc., but every time I've

started

Quoted message said:

playing with emacs I've forgotten all the commands from the last time.

~shuddres~

Give me 'vi' anytime! I find it to be much more flexible (for my tastes). I eschew ANYTHING of an
IDE for code development. It is more of a hinderance than a help. Plus, it prevents me from writing
my own makefiles 😄

Joe

Carl Fogel said:

David Damerell <[email hidden]>:

Quoted message said:

Google run a news server that is unusual in two respects; it has a Web interface for posting and
reading, and it does not ordinarily expire old articles.

I win my bet--thanks for the explanation. Posts on the Google news server seem to stretch back to
the crack of doom, which I find delightful. Am I right to infer that other servers lack the space,
patience, or interest to keep the old posts as long and ruthlessly delete them when they're past
some freshness date?

Expire policy is up to an individual site, and often varies from group to group. For instance, the
commercial ISPs generally carry alt.binaries ([censored] and copyright violations), but expire it very
quickly because of the huge volume. Conversely, local hierarchies (like chiark.* here) may not be
expired at all.

Quoted message said:

From what Simon says elsewhere in this thread, I'm guessing that this might have something to do
with differences between trn3 and emacs?

No, it doesn't. Ordinarily cancelling would be a single keystroke. When I say "a local oddity", I
mean an oddity of the system on which I read news.

Quoted message said:

If some servers will cancel, does this mean that posts vanish from some servers, but are still
visible on others?

Yes.

Quoted message said:

Do they vanish without a trace on the obliging servers, or is there a no-longer-available message?

From a user's point of view they disappear without trace, unless an article is posted as a followup
to the cancelled article, in which case it will mention the Message-ID of the cancelled article in
the References line.

Quoted message said:

And what about this supersede business? Is it just a cancel that's contingent on posting a
replacement as part of the process?

Functionally, yes, but it is implemented by having the replacement identify itself as superseding
the original.

Quoted message said:

Or is it an editing of the original?

The mechanism is often presented to users as editing the original article, but in fact it is
transmitted as a complete replacement message.

Brandt often supersedes articles.
--
David Damerell <[email hidden]> Distortion Field!

David Damerell <[email hidden]> wrote in message
news:<[email hidden]>...

[snip]

Dear David,

Marvelous!

I'm twitting you in other threads, but here I can only admire how clearly and carefully you
explained what's going on behind the curtain of the newsgroup's mechanics.

I appreciate the time and trouble that you must have taken to consider and answer my flurry of
questions.

So here's another.

Last night, I was writing a rude follow-up to Chalo--arrogant might be a better word--and he
disappeared for half an hour!

The thread index clearly showed him, I read his post, waxed indignant, scribbled a follow-up, and
was told that "Re:" was not allowed except in a follow-up when I tried to post. His post had
disappeared while I was typing.

Backing up, refreshing, and even re-booting (pathetic, I know) failed to find Chalo's two most
recent posts. I thought at first that perhaps he had cancelled them and that I was finally seeing
what you've been explaining, but after half an hour his posts reappeared.

I've seen other now-you-see-it-now-you-don't behavior on the newsgroup, but nothing quite as bad as
this. (The previous worst-case was a thread in which I found myself reading post 61 and then
noticing that the thread index stopped at 60, even when I refreshed it a few times.)

Am I right in thinking that there are some bizarre hiccups in the indexing, propagation, or server
caching of new posts?

I wish that I had a reply as useful as yours to offer in return.

Thanks,

Carl Fogel

Carl Fogel said:

Am I right in thinking that there are some bizarre hiccups in the indexing, propagation, or server
caching of new posts?

Clearly I cannot comment on the internal mechanics of Google's news server. The propagation
mechanism itself could not cause the behaviour you describe.
--
David Damerell <[email hidden]> flcl?