Skip to main content
news

Re: nospam still not admitt...

Sandman
SubjectRe: nospam still not admitting to an error (was: The closest we'll get to nospam admitting to an err
FromSandman
Date01/27/2016 07:40 (01/27/2016 07:40)
Message-ID<sandman-6ce376e3bb2d81f9ce74d2e5e29dccf8@individual.net>
Client
Newsgroupsrec.photo.digital
FollowsEric Stevens
FollowupsEric Stevens (14h & 33m) > Sandman

In article <8ragablbq9o64gd9up987vuvshe6qo9v10@4ax.com>, Eric Stevens wrote:

Eric Stevens
Quote my text you deleted and see if you can still justify your bluster.

Sandman
Here is your irrelevant verbiage re-insterted:

Eric Stevens
See https://tools.ietf.org/html/rfc5256#ref-THREADING

Sandman
This is an irrelevant RFC, hence it was snipped. It pertains mainly to a IMAP sorting extension (command) called "THREAD", which is a version of the SEARCH command, where the IMAP server can respond to a THREAD command and give the mail program messages sorted as they pertain to each other in a conversation.

Eric Stevens
Exactly the same thing happens in news groups.

Incorrect.

Also, the same libraries handle IMAP, NNTP, POP3 etc.

The extension outlined in the RFC above does not, I repeat; DOES NOT exist for NNTP or POP3. You have now made yet another incorrect statement.

"The ORDEREDSUBJECT threading algorithm is also referred to as "poor man's threading". The searched messages are sorted by base subject and then by the sent date. The messages are then split into separate threads, with each thread containing messages with the same base subject text. Finally, the threads are sorted by the sent date of the first message in the thread."

Sandman
The above quote, again, is only relevant to a parameter to the THREAD command to a IMAP server, and has exactly nothing to do with NNTP.

Eric Stevens
Threading order has nothing to do with NNTP either.

Of course not. No one has claimed it does.

How threading is handled is determined by the author of the software.

You're the only one that has claimed otherwise:

Eric Stevens 01/23/2016 <mig5abdjtuoep0eln74jdkmik4jr0v30l6@4ax.com>

"As far as the reader is concerned, you have changed the subject and it's now a different thread"

Sandman
I.e., in short, an IMAP client can issue a THREAD command as such:

THREAD <algorithm><charset><search phrase>

So, issuing it as such:

THREAD ORDEREDSUBJECT UTF-8 Eric has problems with RFC's

Would search the mailbox on the server for everything that contains "Eric has problems with RFC's" and return the search result in subject order and then date order.

Again, this has NOTHING to do with NNTP (usenet) where messages are fetched from the server using either the XOVER or XHDR commands.

Eric Stevens
Exactly.

You just didn't understand anything of the above, right?

Your software will not enable you to construct threads like this and for that reason you seem to think it is wrong.

Sandman
The above is irrelevant to your explicit claim:

Eric Stevens 01/23/2016 <mig5abdjtuoep0eln74jdkmik4jr0v30l6@4ax.com>

"As far as the reader is concerned, you have changed the subject and it's now a different thread"

I have never said that something is "wrong". I have correctly pointed out that your news client of choice displays threads in a non-standard way, where it not only displays a change of subject as a new thread, it also deletes the references in any followups to it, thereby breaking the thread.

Eric Stevens
It doesn't you know. It makes a new thread but keeps the previous references. I can always change the display mode and go back to your preferred way of doing things.

It has nothing to do with me - it's the *standard* way of displaying threading, shared by all threading news readers except one; yours.

Agent says of this:

"Check this box to have Agent include messages with the same subject under the same thread, even if the follow-up message doesn't contain a valid reference to the original. People frequently mis-post responses and this helps match follow-ups with the message to which they are responses.

Checking this box will also cause Agent to thread your email messages."

As you say, not many news reader will enable this but my first quote shows that this does not mean that Agent is wrong to do so. Threading on subject is permitted and can be a good idea.

Sandman
Agent is doing it in a non-standard way, which I have shown. Also, it is in violation of the RFC, which I have also shown. You, on the other hand, have failed to support your *explicit* claims about what "the reader" thinks is a new thread or not.

Eric Stevens
The wording of the RFC means that it is not compulsory.

No matter how much weaseling you want to put in your reading of the RFC, you still made incorrect explicit claims that have have proven wrong. And that was *two years ago* and here you are again bringing it up again.

Sandman
Since it has been shown that the majority of news clients shows posts as threaded by the References header, then it is clear that "the reader" would most certainly see the post as part of the ongoing thread, regardless of subject.

So, just stop making explicit claims about what a thread is based on how your non-standard news client present it to you. Usenet has been around for decades and what is and what isn't a "thread" is already firmly established

Eric Stevens
I suggest you study the wording of the RFCs and consider what they are actually saying. Pay particular attention to the distinction between 'should' and 'must'.

The word is "may", as in "the References field may be used to identify a "thread" of conversation".

Again, the wording of the RFC is 100% irrelevant. You are the one that two years ago started to ask for RFC's, not I. This "argument" didn't start because I said the RFC claims anything about what a news client "must" be doing. You're the one that said this:

"You can write software to do anything you like. That doesn't mean that it conforms with a standard."

"Standard" is the normal way things is done. A RFC can be considered a way to standardize things, but nowhere in any RFC does it say that threading must be based on the Subject header. The only time threading is ever mention in any RFC that pertains to NNTP or internet messaging as a whole, it explicitly mentions the References header.

Also, "standard" can refer to how something normally is done, and I have shown without a shadow of a doubt that the standard way to visually thread a usenet conversation is by the References header.

I'm not taking this argument any further.

You shouldn't have brought it up again to begin with. You had four chances to bow out where I did nothing but ask you why you were bringing it up again. You responded by insisting that you wanted to have this argument yet again, and I have handed you your ass yet again. So yes, you need to step out and rethink your goals. Good choice.

-- Sandman

Eric Stevens (14h & 33m) > Sandman