Ray said:
I have configured message rules every way possible but cannot delete any
messages.
I choose Apply Where the message is on - alt.binaries.images.fun
Where the From line contains - Idol Basics
Delete it
Apply to folder - alt.binaries.images.fun
Close
Nothing changes - the asshole is still there with all his damned yenc
entries
I haven't used WLM is quite awhile (and that was before v15 where
Microsoft screwed it up even worse by removing proper quoting) but I
recall you could define rules that delete messages (from the local
store, not from the server).
Look at the order of your rules. Perhaps you have a rule before this
one which has the stop-clause and is firing. It a rule fires and has
the stop-clause in it then no following rules are exercised on that
article.
Look at the raw source for the article (in Outlook Express, you hit
Ctrl+F3 to see the raw source so it might be the same in WLM). What you
see rendered in the GUI may not be what's actually in the article.
That's because ISO encoding could be used in several headers (From,
Subject, etc). What you see looks like ASCII-7 text but it could be
encoded text. For example, you might see "Subject: Idol Basics" in the
GUI but the raw source shows "Subject: =?<charset>?=...". Microsoft
tests on the actual value, not the rendered value. Your rule is testing
on the ASCII-7 text string you see but the poster is using the old trick
of using encoding so your text rule won't see the same string in the raw
source of the article.
For example, there's a Roger poster in a newsgroup that I don't want to
see his posts (I test more than just on "Roger" in the headers to
identify him and not some other Roger). He fluctuates his nym by using
ASCII-8 and encoded text strings, like Roger, Rôger, Rogêr, Rôgêr, and
=?...?R=F4g=EAr?= (I forget what encoding he uses but then my filter
skips over the encoding spec portion, anyway). So my rule, along with
testing on the e-mail address in the From header (an overview header)
and the injection node in the Path header (a non-overview header).
Encoding can also be used in the Subject header. They use encoding that
looks like ASCII-7 text to you in the rendered version in the GUI but
testing on that string will fail because the actual string in the header
is an encoded string, like "=?<encoding>?<string>?=". Yep, if they use
encoding then you have to test on the actual encoding string, not its
rendered (displayed) string.
If the boob is flooding the newsgroup (like Panzke in the Windows Vista
group and like Bullwinkle in the 24hoursupport.helpdesk group), do you
really want to see ANY of his posts? Once someone behaves like a
hijacker of a group, just filtering out ALL their posts. And don't
bother telling them that you filtered them out (as it could prod them to
nymshift and it makes you look foolish in a vacuous threat). If they
have enough info that's stable in their headers to identify them, filter
them out. Alas, you're using a weak newsreader (WLM) for filtering as
you can only test on the overview headers (what your newsreader gets
when it gets a list of articles) and none of the other non-overview
headers (that are available when you retrieve the body which retrieves
the entire article).
