Re: Funny scoring.

Yngvar Folling (yngvar.folling@login.eunet.no)
Mon, 10 Jun 1996 12:56:14 +0200

In article <Xbzuxc6r3DxV091yn@io.org>, cthuang@io.org (Chin Huang) wrote:

> Because of the precedence of the | operator, this regular expression
> matches any line in the article header that contains the word "funny".

I realized eventually that it had to be something like that.

> In the next release, I'm going to search on the header value, not the
> entire line.

It might be an idea, when searching header fields, to regard fields that
span several lines as one single line. That way we could search for
followups to an article by searching for "References: <message-id>"
regardless of whether the Message-ID appeared on the same line as the
text References or not.

That reminds me, I think there could be a simple way to search for the
article whose Message-ID matches the last ID on the References field (to
search for the message to which this is a followup), other messages with
References fields ending with the same Message-ID (to search for other
followups to the same message as this one), and messages with the
current Message-ID in the References field (to search for followups to
the current message). These are all searches I do quite often, but it's
tedious to type the current Message-ID manually.

You could, for instance, map searches like these to Ctrl-arrow
combinations. Ctrl-Up to search for a "parent," Ctrl-Left and
Ctrl-Right for previous and following "siblings," and Ctrl-Down to
search for a "child," or followup.

Another thing I'd very much like, would be a simple search for followups
to articles written by me. I could search for articles referring to
message-IDs ending with my domain, but that catches followups to anyone
else posting from my site as well.

Yngvar