On Saturday 26 April 2003 10:57 am, Graham Murray wrote:
Vince Littler <suse@archipelago.eclipse.co.uk> writes:
Similarly with threading. If there was a _published_ standard [with broad agreement and enough wrinkles ironed out to be broadly beneficial], I would want a client to support that standard. I would want SuSE to provide clients which complied [presuming on the broadness of SuSE's shoulders!]
There *is* a published standard - RFC2822. The following extract expands
Found it at http://rfc.net/rfc2822.html and your extract is 3.6.4. - Identification fields. Excellent, this is what I think the whole threading issue needs [and looking back, I see that rex also pointed out this one]. And [at the risk of repeating myself] a client is required which supports this _fully_. As Carlos pointed out, both James Mohr and myself use K Mail. James thinks it is threading compliant, I don't - or rather I think the support is partial. All the complaints over broken threading come from the receiving end where the client visibly supports threading. There are NO complaints about "I cannot control the thread I am sending my message on" from anyone at the sending end, but this is where all of the complaints originate. Andre Truter said:
So, if we see 'hijacking' a conversation as rude and improper, then how can we justify 'hijacking' a thread? It comes down to the same thing.
Jim Norton said:
Some people will simply reply to a mail so they don't have to worry about knowing the email address to post a message to a list. They change the subject and throw in the text according to what they are writing about. <snip> I don't think people are neccessarily being rude, more like uninformed.
I think the reason there are no complaints is firstly that there are 3 conceptual candidates for threading [at least]: 1] Subject line 2] rfc2822 support 3] "normalizing the subject" as used for archiving this list And secondly when someone composes a message, they are only confronted with any choice over 1], which looks very much the same as the threading which they see at the receiving end. Suggestion: =*=*=*=*=*= So what is required is [talking KMail here] in the window for composing a reply, besides the fields for [sending] Identity:, From:, Reply To:, To:, CC:, and Subject:, another field Thread References:. This field would be non editable and would have a tick box beside it saying "Start New Thread", which would blank and restore the Thread References field as it was ticked and unticked. And when the message is sent, the Thread References would be incorporated or not as appropriate. At a stroke this should resolve the thread hijacking issue, because at the sending end you will actually have a visible handle on the real data structure which controls the thread. So are there any takers for this suggestion? Let me know by private email and I will enter it on the KDE wish list. regards Vince Littler