Gnus development mailing list
 help / color / mirror / Atom feed
* Summary scoring commands/a puzzle
@ 1998-03-11 13:32 Harry Putnam
  1998-03-12 21:37 ` Nils Goesche
  0 siblings, 1 reply; 3+ messages in thread
From: Harry Putnam @ 1998-03-11 13:32 UTC (permalink / raw)



I may have committed some sort of 'faux-paux' here by posting this here
and to gnu.emacs.gnu

I believe gnus Info needs some further explanation in this area.  And
would like to offer some of the contents here as a beginning.  Hopefully
some one who is not iliterate like I am, will step forward to do it.
But if not, I will do my damndest with the help of other contributions
and emacs,ccc to get this, and what ever contributions others want to send
into a much briefer and clearer expostition of using the summary scoring
commands.



Lengthy and detailed -- please bear with me.

Having spent ungodly amounts of time trying to understand and utilize
the scoring commands available in Summary mode.  I have learned to use
some of them but am still quite confused about others.

Firstly, some confusing terms used seem to  make it much harder to 
figure out what is expected at the different prompts.

When doing a simple 'I'or 'L' the prompt offered says 
'Increase/Decrease Header: (asbhitxldfT?):'

Clear enough what to do here and if you don't know what the letters mean
there is the handy (?) you can press to find out.(very nice) Now it
begins to get more interesting.

Author and subject, message-ID, refererences  all make
good sense, after all, just like the prompt says, they are 'headers'.
But 'h'=head and 'b'=body do not make sense because they are not
'headers'.  What part of a header is the body or the head.

So at this point you must ignore what the prompt is asking for and
strike out on your own.  It turns out,'h= 'head' means all the headers
instead of the 'head' of a 'header' and b=body means the text body of
message.  Neither of them are called 'headers'.

Using the terms 'all headers' and 'message body'or similar would be a
lot clearer than 'Head' and 'body'.  But maybe only dim wits like yours
truly, stumble over these things, so moving on:

We choose 'h' and then see "Increase header 'head' with match type (zp?):"
[ED z=substring p=regexp string]

So we are being shown our choices so far and asked to make another one.
[ED NOTE:  Those handy question marks (?) appear throughout]

Now you might think that you need to put the string, in which ever format
you pick, but not so.  You just pick which one you want to search with
and Gnus will convert it nicely for you.

We chose 'p':
We now see the prompt 'Increase permanence (tpi?):'
[ED temp perm and immediate]

Either 't' or 'p' will work at this point  'i' will not work, if it is
choosen the scoring will fail.  

NOTE: It would seem that choosing 'i' would let you run the scoring
without putting it in your .SCORE file, but apperantly it does not.


We choose 'p':

Prompt now says: "Match permanent on head, raise:"

This might be taken as a question of how much to raise what you are
going to choose.  But no, it is asking for a 'header string' ie. To:,
From:, or any of the header lines.

We insert:  To: Harry Putnam <reader@newsguy.com>

And sure as water runs downhill, articles with that header are increased
by the default 1000. (assuming there is nothing in score file to the
contrary)

NOTE:  If you want a different amount you must prefix C-u <number>
before pressing the initial 'I'.

Notes and questions:

What role does the 'i' in (tpi?) play?

How can a scoring session be run with out putting it in a score file,
using the summary commmands?

In other words I want to run a scoring session that ignores my SCORE
file.  

I can see how to do this by jerking around the files themselves
but not by simply by-passing.

'V c' and 'V f' will clearly allow using a 'doctored' score file, but is
there no way to just score a one time scoring with out using your
existing score file or any others?


I think the 'V a' command may answer this, but have not figured out how
to use it yet.


Using 'V a'
Header: <h> or <head> <RET>

Match: <string> <RET>

Use regexp match? (y or n) <y>

Add to score file?  (y or n): <y>

NOTE: IF 'n' is chosen here, no scoring is done

Expire Kill? (y or n) <no>

NOTE: Either a 'y' or an 'n'  will have the same result here

Scoring is done immediately and new entry is added to score file.

Questions:

What is supposed to happen at the prompt that says:
"Add to score file?  (y or n):" 

Is 'n' just a way to cancel out?  Don't we have C-g for that?
Either chosing 'n' or doing C-g have identical results.

What is supposed to happen at the prompt that says:
"Expire Kill? (y or n)"
Either choice has identical results.  The scoring is done and thats it.

thanks for hanging in there
-- 

Harry Putnam  reader@newsguy.com


^ permalink raw reply	[flat|nested] 3+ messages in thread
end of thread, other threads:[~1998-03-12 21:37 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
1998-03-11 13:32 Summary scoring commands/a puzzle Harry Putnam
1998-03-12 21:37 ` Nils Goesche
1998-03-12 17:52   ` Harry Putnam

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).