If I process mark a couple of articles to followup, so as to quote all the text, the references constructed seems to be based only on the first article. I wonder if the references from all the articles could be included. Or if that's a bit hard then maybe use the last article instead of the first. This would be mostly for following up at the tail of some thread, but also wanting to cite something from an earlier message.
If I process mark a couple of articles to followup, so as to quote all the text, the references constructed seems to be based only on the first article. I wonder if the references from all the articles could be included. Or if that's a bit hard then maybe use the last article instead of the first. This would be mostly for following up at the tail of some thread, but also wanting to cite something from an earlier message.
Kevin Ryde <user42@zip.com.au> writes: > If I process mark a couple of articles to followup, so as to quote all > the text, the references constructed seems to be based only on the > first article. > I wonder if the references from all the articles could be included. > Or if that's a bit hard then maybe use the last article instead of the > first. The problem is that References is, in principle, a strict linear parentage list from head to tail of a thread. Trying to use References with multiple, immediately-referring message-ids creates something of a lattice, abandoning the linear interpretation. Reputedly, the USEFOR draft includes a much-modified concept of In-Reply-To which can contain the message-ids of all messages to which the new message refers. (I confess to not having personally seen USEFOR's draft in an indecently long time.) But I don't believe there's any support for that yet. In any event, complicating matters is the problem of how one would thread a message in a summary display, based on linear References interpretation, that had lattice-like References content.
Kevin Ryde <user42@zip.com.au> writes: > If I process mark a couple of articles to followup, so as to quote all > the text, the references constructed seems to be based only on the > first article. > I wonder if the references from all the articles could be included. > Or if that's a bit hard then maybe use the last article instead of the > first. The problem is that References is, in principle, a strict linear parentage list from head to tail of a thread. Trying to use References with multiple, immediately-referring message-ids creates something of a lattice, abandoning the linear interpretation. Reputedly, the USEFOR draft includes a much-modified concept of In-Reply-To which can contain the message-ids of all messages to which the new message refers. (I confess to not having personally seen USEFOR's draft in an indecently long time.) But I don't believe there's any support for that yet. In any event, complicating matters is the problem of how one would thread a message in a summary display, based on linear References interpretation, that had lattice-like References content.
>>>>> On Wed, 20 Feb 2002 10:22:48 +1000, Kevin Ryde <user42@zip.com.au> said:
Kevin> If I process mark a couple of articles to followup, so as to
Kevin> quote all the text, the references constructed seems to be
Kevin> based only on the first article.
As long as I'm at it, doing followups of multiple articles no longer
includes a merged copy of the to/cc lists and just takes the first or
last one (I forget) (no I don't. I just tried it again, it takes the
first address list one only, but quotes text from them all).
--
"Ninjas aren't dangerous. They're more afraid of you than you are of them."
>>>>> On Wed, 20 Feb 2002 10:22:48 +1000, Kevin Ryde <user42@zip.com.au> said:
Kevin> If I process mark a couple of articles to followup, so as to
Kevin> quote all the text, the references constructed seems to be
Kevin> based only on the first article.
As long as I'm at it, doing followups of multiple articles no longer
includes a merged copy of the to/cc lists and just takes the first or
last one (I forget) (no I don't. I just tried it again, it takes the
first address list one only, but quotes text from them all).
--
"Ninjas aren't dangerous. They're more afraid of you than you are of them."
Karl Kleinpaste <karl@charcoal.com> writes: > > The problem is that References is, in principle, a strict linear > parentage list from head to tail of a thread. Trying to use > References with multiple, immediately-referring message-ids creates > something of a lattice, abandoning the linear interpretation. I'd wondered about that, without knowing what the standards say. Taking the references just from the last message marked would be enough for the common uses I can envisage. > In any event, complicating matters is the problem of how one would > thread a message in a summary display, based on linear References > interpretation, that had lattice-like References content. I guess if you draw together two strands by referring to both then there's a question of under which one that message should come out. Yes, that'd be a bit ugly.
Karl Kleinpaste <karl@charcoal.com> writes: > > The problem is that References is, in principle, a strict linear > parentage list from head to tail of a thread. Trying to use > References with multiple, immediately-referring message-ids creates > something of a lattice, abandoning the linear interpretation. I'd wondered about that, without knowing what the standards say. Taking the references just from the last message marked would be enough for the common uses I can envisage. > In any event, complicating matters is the problem of how one would > thread a message in a summary display, based on linear References > interpretation, that had lattice-like References content. I guess if you draw together two strands by referring to both then there's a question of under which one that message should come out. Yes, that'd be a bit ugly.