scoring bugs, problems, and suggestions

scoring bugs, problems, and suggestions

[Greg Stark]

September Gnus v0.51; nntp 4.0; nnml 1.0; nnmh 1.0; nneething 1.0; nndoc 1.0; nndraft 1.0; nnfolder 1.0
GNU Emacs 19.30.1 (mips-dec-ultrix4.2, X toolkit) of Mon Jan  8 1996 on yaz-pistachio

I recently added a `head' score rule for my mail to mark duplicated messages
as read and Gnus started messaging for each article it scored.  Based on my
experience with this a series of suggestions have come to mind, some of these
i consider bugs, some will unfortunately have to wait for Red Gnus:

1) put the Gnus-Warning header in mail overview files
   in fact, allow me to put whatever headers i want in my overview files
2) allow me to save the scores of some groups but not others
   since i only tend to read articles more than once in mail groups.
   a token in the score file sounds better than a regexp imho.
3) sort numbers in numeric, not alphabetic order, duh.
4) I would actually expect number of articles scored and 
   total number of articles needing scoring would be more 
   useful than actual article numbers because of gaps and threads.
5) get a much better guess of how many articles need to be scored
   you have a lot more information at this stage than in the Group buffer
   and you seem to be counting only new articles regardless of the
   number of articles being displayed and regardless of gnus-save-score
	

*Messages* buffer:
================================================================
>  Scoring...
>  Scoring on article 1 of 1...
>  Scoring on article 10 of 1...
>  Scoring on article 11 of 1...
>  Scoring on article 12 of 1...
>  Scoring on article 13 of 1...
>  Scoring on article 14 of 1...
>  Scoring on article 15 of 1...
>  Scoring on article 16 of 1...
>  Scoring on article 17 of 1...
>  Scoring on article 18 of 1...
>  Scoring on article 19 of 1...
>  Scoring on article 2 of 1...
>  Scoring on article 20 of 1...
>  Scoring on article 21 of 1...
>  Scoring on article 22 of 1...
>  Scoring on article 23 of 1...
>  Scoring on article 24 of 1...
>  Scoring on article 25 of 1...
>  Scoring on article 26 of 1...
>  Scoring on article 27 of 1...
>  Scoring on article 28 of 1...
>  Scoring on article 29 of 1...
>  Scoring on article 3 of 1...
>  Scoring on article 30 of 1...
>  Scoring on article 31 of 1...
>  Scoring on article 32 of 1...
>  Scoring on article 33 of 1...
>  Scoring on article 34 of 1...
>  Scoring on article 4 of 1...
>  Scoring on article 5 of 1...
>  Scoring on article 6 of 1...
>  Scoring on article 7 of 1...
>  Scoring on article 8 of 1...
>  Scoring on article 9 of 1...
>  Scoring...done
================================================================

Re: scoring bugs, problems, and suggestions

[Lars Magne Ingebrigtsen]

gsstark@MIT.EDU (Greg Stark) writes:

> 1) put the Gnus-Warning header in mail overview files
>    in fact, allow me to put whatever headers i want in my overview files

This is on the Red Gnus todo list.

> 2) allow me to save the scores of some groups but not others
>    since i only tend to read articles more than once in mail groups.
>    a token in the score file sounds better than a regexp imho.

I think you can control this by setting `gnus-save-score' in an
appropriate hook...

> 3) sort numbers in numeric, not alphabetic order, duh.

It doesn't really matter what order articles are scored, does it?

> 4) I would actually expect number of articles scored and 
>    total number of articles needing scoring would be more 
>    useful than actual article numbers because of gaps and threads.
> 5) get a much better guess of how many articles need to be scored
>    you have a lot more information at this stage than in the Group buffer
>    and you seem to be counting only new articles regardless of the
>    number of articles being displayed and regardless of gnus-save-score

I'm not sure what you mean here, exactly...

-- 
  "Yes.  The journey through the human heart 
     would have to wait until some other time."

Re: scoring bugs, problems, and suggestions

[Greg Stark]

>>>>> Lars Magne Ingebrigtsen writes:

>> 3) sort numbers in numeric, not alphabetic order, duh.
> It doesn't really matter what order articles are scored, does it? [...]

>> 4) I would actually expect number of articles scored and total number of
>> articles needing scoring would be more useful than actual article numbers
>> because of gaps and threads. 5) get a much better guess of how many
>> articles need to be scored you have a lot more information at this stage
>> than in the Group buffer and you seem to be counting only new articles
>> regardless of the number of articles being displayed and regardless of
>> gnus-save-score
> I'm not sure what you mean here, exactly...

What i meant is that these messages are pretty useless:

Scoring...
Scoring on article 335 of 335...
Scoring on article 339 of 335...
Scoring on article 168 of 335...
Scoring on article 197 of 335...
Scoring on article 309 of 335...
Scoring on article 322 of 335...
Scoring on article 324 of 335...
Scoring on article 329 of 335...
Scoring on article 363 of 335...
Scoring on article 378 of 335...
Scoring on article 450 of 335...
Scoring on article 549 of 335...
Scoring on article 623 of 335...
Scoring on article 780 of 335...
Scoring on article 783 of 335...
Scoring on article 803 of 335...
Scoring on article 288 of 335...
Scoring on article 89 of 335...
Scoring on article 748 of 335...
Scoring on article 750 of 335...
Scoring on article 752 of 335...
Scoring on article 772 of 335...
Scoring on article 774 of 335...
Scoring on article 779 of 335...
Scoring on article 826 of 335...
Scoring...done
Scoring...
Scoring...done

They don't give me an idea of how many articles have been done, how many
remain, how many have to be done total, or any other useful information.

The only bit of information they give is the article number it's currently on,
which doesn't really mean much.  it doesn't even give the user an idea where
it is in the process until the user looks at the *Messages* buffer and figures
out how it's being sorted.

I just think messages like:
Scoring on article 1 of 25
Scoring on article 2 of 25
Scoring on article 3 of 25
Scoring on article 4 of 25
Scoring on article 5 of 25
Scoring on article 6 of 25
Scoring on article 7 of 25
Scoring on article 8 of 25
Scoring on article 9 of 25
Scoring on article 10 of 25
Scoring on article 11 of 25
Scoring on article 12 of 25
Scoring on article 13 of 25
Scoring on article 14 of 25
Scoring on article 15 of 25
Scoring on article 16 of 25
Scoring on article 17 of 25
Scoring on article 18 of 25
Scoring on article 19 of 25
Scoring on article 20 of 25
Scoring on article 21 of 25
Scoring on article 22 of 25
Scoring on article 23 of 25
Scoring on article 24 of 25
Scoring on article 25 of 25

would be far more useful, possibly with an (article number 335) at the end of
the line if you insisted.


greg

Re: scoring bugs, problems, and suggestions

[Lars Magne Ingebrigtsen]

gsstark@mit.edu (Greg Stark) writes:

> What i meant is that these messages are pretty useless:
> 
> Scoring...
> Scoring on article 779 of 335...
> Scoring on article 826 of 335...

Heh.  I see what you mean.  Fix in 0.55.

-- 
  "Yes.  The journey through the human heart 
     would have to wait until some other time."