[bug] Mark limit score

[bug] Mark limit score

[Frank Schmitt]

[-- Attachment #1: Type: text/plain, Size: 343 bytes --]

There are currently two entries in the menu Gnus->Mark Limit which are
called Score. That's already not very nice, but I think that one or both
of them are broken.

Try to hit /p in the summary buffer. This results for me in an error message
saying "There is no diplay group parameter" (sic!).
First of all we've got to get rid of this typo:


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: Fix typo in gnus-sum.el --]
[-- Type: text/x-patch, Size: 457 bytes --]

--- gnus-sum.el	Wed Oct 17 22:14:16 2001
+++ gnus-sum2.el	Thu Oct 18 09:23:32 2001
@@ -6894,7 +6894,7 @@
   "Limit the summary buffer to the predicated in the `display' group parameter."
   (interactive)
   (unless gnus-newsgroup-display
-    (error "There is no `diplay' group parameter"))
+    (error "There is no `display' group parameter"))
   (let (articles)
     (dolist (number gnus-newsgroup-articles)
       (when (funcall gnus-newsgroup-display)


[-- Attachment #3: Type: text/plain, Size: 439 bytes --]

Then: /v calls gnus-summary-limit-to-score. This function is the one I
want, but it doesn't ask for the score I wish to limit the summary
to. It limits the summary to articles with a score bigger or equal than
gnus-summary-default-score which is quite senseless.

-- 
One Ring to rule them all, One Ring to find them,
One Ring to bring them all and in the darkness bind them
In the Land of Mordor where the Shadows lie.
19. Dezember 2001

Re: [bug] Mark limit score

[Björn Torkelsson]

Frank Schmitt <usereplyto@Frank-Schmitt.net> writes:

> There are currently two entries in the menu Gnus->Mark Limit which are
> called Score. That's already not very nice, but I think that one or both
> of them are broken.

The second one should probably be "Predicate" or something...

/torkel

Re: [bug] Mark limit score

[Simon Josefsson]

Frank Schmitt <usereplyto@Frank-Schmitt.net> writes:

> There are currently two entries in the menu Gnus->Mark Limit which are
> called Score. That's already not very nice, but I think that one or both
> of them are broken.
> 
> Try to hit /p in the summary buffer. This results for me in an error message
> saying "There is no diplay group parameter" (sic!).

Fixed.

> Then: /v calls gnus-summary-limit-to-score. This function is the one I
> want, but it doesn't ask for the score I wish to limit the summary
> to. It limits the summary to articles with a score bigger or equal than
> gnus-summary-default-score which is quite senseless.

Maybe it is possible to use a prefix, e.g. '100 / v'.

Re: [bug] Mark limit score

[Frank Schmitt]

[-- Attachment #1: Type: text/plain, Size: 410 bytes --]

Simon Josefsson <jas@extundo.com> writes:
 
>> Then: /v calls gnus-summary-limit-to-score. This function is the one I
>> want, but it doesn't ask for the score I wish to limit the summary
>> to. It limits the summary to articles with a score bigger or equal than
>> gnus-summary-default-score which is quite senseless.
> 
> Maybe it is possible to use a prefix, e.g. '100 / v'.

What do you think about this:


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: Patch against gnus-sum.el --]
[-- Type: text/x-patch, Size: 726 bytes --]

--- lisp/gnus-sum.el	Thu Oct 18 20:03:02 2001
+++ /cygdrive/c/Programme/XEmacs/xemacs-packages/lisp/gnus/gnus-sum.el	Thu Oct 18 22:26:50 2001
@@ -6958,9 +6958,12 @@
 (defun gnus-summary-limit-to-score (&optional score)
   "Limit to articles with score at or above SCORE."
   (interactive "P")
-  (setq score (if score
-		  (prefix-numeric-value score)
-		(or gnus-summary-default-score 0)))
+  (if (not score)
+      (setq score 
+	    (string-to-number (read-from-minibuffer 
+			       "Limit to articles with a score bigger or equal than: " 
+			       (number-to-string (or gnus-summary-default-score 0)))))
+    (setq score (prefix-numeric-value score)))
   (let ((data gnus-newsgroup-data)
 	articles)
     (while data


[-- Attachment #3: Type: text/plain, Size: 178 bytes --]

 

-- 
One Ring to rule them all, One Ring to find them,
One Ring to bring them all and in the darkness bind them
In the Land of Mordor where the Shadows lie.
19. Dezember 2001

Re: [bug] Mark limit score

[Kai Großjohann]

Frank Schmitt <usereplyto@Frank-Schmitt.net> writes:

> What do you think about this:
> 
> 
> --- lisp/gnus-sum.el	Thu Oct 18 20:03:02 2001
> +++ /cygdrive/c/Programme/XEmacs/xemacs-packages/lisp/gnus/gnus-sum.el	Thu Oct 18 22:26:50 2001
> @@ -6958,9 +6958,12 @@
>  (defun gnus-summary-limit-to-score (&optional score)
>    "Limit to articles with score at or above SCORE."
>    (interactive "P")
> -  (setq score (if score
> -		  (prefix-numeric-value score)
> -		(or gnus-summary-default-score 0)))
> +  (if (not score)
> +      (setq score 
> +	    (string-to-number (read-from-minibuffer 
> +			       "Limit to articles with a score bigger or equal than: " 
> +			       (number-to-string (or gnus-summary-default-score 0)))))
> +    (setq score (prefix-numeric-value score)))
>    (let ((data gnus-newsgroup-data)
>  	articles)
>      (while data

If I understand correctly, this means that calling
gnus-summary-limit-to-score without specifying a score via the prefix
arg means that the user is prompted.

It's not clear to me that this is a good idea: C-d does not prompt you
how many characters to delete, either, it just deletes 1 character
unless you specify otherwise.

Opinions?

kai
-- 
GNU/Linux provides a nice `poweroff' command, but where is `poweron'?

Re: [bug] Mark limit score

[Frank Schmitt]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:
 
>If I understand correctly, this means that calling
>gnus-summary-limit-to-score without specifying a score via the prefix
>arg means that the user is prompted.

Correct. With the value formerly used without asking as default. 

>It's not clear to me that this is a good idea: C-d does not prompt you
>how many characters to delete, either, it just deletes 1 character
>unless you specify otherwise.

Well, but all the other gnus-summary-limit-to-* function ask too. What's
the sense of this function otherwise?

-- 
One Ring to rule them all, One Ring to find them,
One Ring to bring them all and in the darkness bind them
In the Land of Mordor where the Shadows lie.
19. Dezember 2001

Re: [bug] Mark limit score

[Daniel Pittman]

On Fri, 19 Oct 2001, Kai Großjohann wrote:
> Frank Schmitt <usereplyto@Frank-Schmitt.net> writes:
> 
>> What do you think about this:
>> 
>> 
>> --- lisp/gnus-sum.el	Thu Oct 18 20:03:02 2001
>> +++ /cygdrive/c/Programme/XEmacs/xemacs-packages/lisp/gnus/gnus-sum.el

[...]

> If I understand correctly, this means that calling
> gnus-summary-limit-to-score without specifying a score via the prefix
> arg means that the user is prompted.
> 
> It's not clear to me that this is a good idea: C-d does not prompt you
> how many characters to delete, either, it just deletes 1 character
> unless you specify otherwise.

Control-D is operating in an environment where there is only *one*
possible thing that you want it to do: delete the character following
point.

You can *repeat* this operation several times if you wish, getting the
same effect each time.


By contrast, there is a broad range of scores in the average Gnus
buffer[1] that the user may mean when they request a 'narrow to score'.

Also, limit to sender, limit to subject and friends don't have any
default that they use, for precisely the same reason: it's *not* clear
which value the user meant.


Finally, I support this change because every single time I have used
narrow to score I have ended up spitting and cursing at Gnus for being a
brain-dead, stupid thing that has a crazy-ass default on that command.

I suspect that I am not alone in this. :)

        Daniel

Footnotes: 
[1]  Assuming that most Gnus buffers have scoring like mine or more
     advanced.

-- 
Most of the luxuries and many of the so-called comforts of life, are not only
not indispensable, but positive hindrances to the elevation of mankind.
        -- Henry David Thoreau, _Walden_, "Economy" [1854]

Re: [bug] Mark limit score

[Kai Großjohann]

Frank Schmitt <usereplyto@Frank-Schmitt.net> writes:

> Simon Josefsson <jas@extundo.com> writes:
>  
>>> Then: /v calls gnus-summary-limit-to-score. This function is the one I
>>> want, but it doesn't ask for the score I wish to limit the summary
>>> to. It limits the summary to articles with a score bigger or equal than
>>> gnus-summary-default-score which is quite senseless.
>> 
>> Maybe it is possible to use a prefix, e.g. '100 / v'.
> 
> What do you think about this:
> 
> 
> --- lisp/gnus-sum.el	Thu Oct 18 20:03:02 2001
> +++ /cygdrive/c/Programme/XEmacs/xemacs-packages/lisp/gnus/gnus-sum.el	Thu Oct 18 22:26:50 2001
> @@ -6958,9 +6958,12 @@
>  (defun gnus-summary-limit-to-score (&optional score)
>    "Limit to articles with score at or above SCORE."
>    (interactive "P")
> -  (setq score (if score
> -		  (prefix-numeric-value score)
> -		(or gnus-summary-default-score 0)))
> +  (if (not score)
> +      (setq score 
> +	    (string-to-number (read-from-minibuffer 
> +			       "Limit to articles with a score bigger or equal than: " 
> +			       (number-to-string (or gnus-summary-default-score 0)))))
> +    (setq score (prefix-numeric-value score)))
>    (let ((data gnus-newsgroup-data)
>  	articles)
>      (while data

How about the more simple-minded:

(defun gnus-summary-limit-to-score (score)
  "..."
  (interactive "NLimit to articles with score greater or equal to: ")
  ...)

Note how the argument is not optional anymore (callers can supply
gnus-summary-default-score if they like) and how the N spec for
interactive uses the prefix arg, if supplied, but prompts for a number
if no prefix arg is supplied.

Maybe we would like to make the argument optional, so that you don't
have to supply it when calling from Lisp.  But OTOH, there is no call
to this function from Lisp, so maybe this is pointless.

Thoughts?

kai
-- 
Lisp is kinda like tpircstsoP

Re: [bug] Mark limit score

[Frank Schmitt]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:
 
>Note how the argument is not optional anymore (callers can supply
>gnus-summary-default-score if they like) and how the N spec for
>interactive uses the prefix arg, if supplied, but prompts for a number
>if no prefix arg is supplied.

This looks great.

>Maybe we would like to make the argument optional, so that you don't
>have to supply it when calling from Lisp.  But OTOH, there is no call
>to this function from Lisp, so maybe this is pointless.

Yepp, I too don't think that this is needed.

BTW: How about an (additional) enhanced function which limits to
articles with a score within %small and %big (we need to ask the user
for %small and %big of course.)

One more thought: I heard several reports that people don't understand
from the name what Mark limit actually does. I too think that the name
could be improved.

How about "Filter articles" instead of "Mark limit"?

-- 
One Ring to rule them all, One Ring to find them,
One Ring to bring them all and in the darkness bind them
In the Land of Mordor where the Shadows lie.
19. Dezember 2001

Re: [bug] Mark limit score

[Frank Schmitt]

Frank Schmitt <usereplyto@Frank-Schmitt.net> writes:
 
>>Maybe we would like to make the argument optional, so that you don't
>>have to supply it when calling from Lisp.  But OTOH, there is no call
>>to this function from Lisp, so maybe this is pointless.
>
>Yepp, I too don't think that this is needed.

Argh! I wanted to say, that it is not necessary to make the argument optional.

-- 
One Ring to rule them all, One Ring to find them,
One Ring to bring them all and in the darkness bind them
In the Land of Mordor where the Shadows lie.
19. Dezember 2001

Re: [bug] Mark limit score

[Kai Großjohann]

Frank Schmitt <usereplyto@Frank-Schmitt.net> writes:

> Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:
>  
>>Note how the argument is not optional anymore (callers can supply
>>gnus-summary-default-score if they like) and how the N spec for
>>interactive uses the prefix arg, if supplied, but prompts for a number
>>if no prefix arg is supplied.
> 
> This looks great.

Done.

> BTW: How about an (additional) enhanced function which limits to
> articles with a score within %small and %big (we need to ask the user
> for %small and %big of course.)

We accept patches :-)

> One more thought: I heard several reports that people don't understand
> from the name what Mark limit actually does. I too think that the name
> could be improved.
> 
> How about "Filter articles" instead of "Mark limit"?

Well, `filter articles' could mean the same as splitting, or it could
mean the same as limiting.

The `mark' part still needs to be there, for `limit to marks' means to
limit, well, to specific marks that the user can enter.  For example,
the R mark, or the E mark.

Or are you talking about something other than `/ m'?

kai
-- 
Lisp is kinda like tpircstsoP

Re: [bug] Mark limit score

[Frank Schmitt]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

 >> One more thought: I heard several reports that people don't understand
>>from the name what Mark limit actually does. I too think that the name
>>could be improved.
>>
>>How about "Filter articles" instead of "Mark limit"?
>
>Well, `filter articles' could mean the same as splitting, or it could
>mean the same as limiting.
>
>The `mark' part still needs to be there, for `limit to marks' means to
>limit, well, to specific marks that the user can enter.  For example,
>the R mark, or the E mark.
>
>Or are you talking about something other than `/ m'?

Yes, we are. I'm talking about the name of the parent
menu-point. (Gnus->Mark Limit) I really don't see the connection between
"Mark Limit" and "Show me only articles, where
[from|score|subject|date|...] is ..."

In fact I don't really understand what "Mark Limit" does mean after
all. (But the reason for this might be my poor English.)

Nevertheless if it was called "filter articles" or something similar I
think it would be clearer what the entries under this node do.

-- 
One Ring to rule them all, One Ring to find them,
One Ring to bring them all and in the darkness bind them
In the Land of Mordor where the Shadows lie.
19. Dezember 2001

Re: [bug] Mark limit score

[Kai Großjohann]

Frank Schmitt <usereplyto@Frank-Schmitt.net> writes:

> Yes, we are. I'm talking about the name of the parent
> menu-point. (Gnus->Mark Limit) I really don't see the connection between
> "Mark Limit" and "Show me only articles, where
> [from|score|subject|date|...] is ..."

I see.  I guess that `Mark Limit' is a typo and should be `Limit'.
Opinions?

We could make a tooltip that says `Limit to articles according to
various criteria' -- would that help?

Is it possible to put tooltips on menu items which are submenus?

kai
-- 
Lisp is kinda like tpircstsoP

Re: [bug] Mark limit score

[Frank Schmitt]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

>I see.  I guess that `Mark Limit' is a typo and should be `Limit'.
>Opinions?

How about "Limit to ..."

-- 
One Ring to rule them all, One Ring to find them,
One Ring to bring them all and in the darkness bind them
In the Land of Mordor where the Shadows lie.
19. Dezember 2001

Re: [bug] Mark limit score

[Frank Schmitt]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:
 
>>>Note how the argument is not optional anymore (callers can supply
>>>gnus-summary-default-score if they like) and how the N spec for
>>>interactive uses the prefix arg, if supplied, but prompts for a number
>>>if no prefix arg is supplied.
>>
>>This looks great.
>
>Done.

I might have misunderstood you, but why don't Gnus ask me which score
when I do /v? Is it broken, or by design?

-- 
One Ring to rule them all, One Ring to find them,
One Ring to bring them all and in the darkness bind them
In the Land of Mordor where the Shadows lie.
19. Dezember 2001

Re: [bug] Mark limit score

[Kai Großjohann]

Frank Schmitt <usereplyto@Frank-Schmitt.net> writes:

> I might have misunderstood you, but why don't Gnus ask me which score
> when I do /v? Is it broken, or by design?

I needed to re-byte-compile, but now `/ v' asks me, I just tested it.

kai
-- 
Lisp is kinda like tpircstsoP