Bind 'w' to a more useful function

Bind 'w' to a more useful function

[Leo]

Dear all,

In summary buffer, 'w' is bound to a function that is barely used. I
wonder would it make more sense to bind it to
`gnus-article-fill-cited-article'.

HTH,
-- 
Leo <sdl.web AT gmail.com>                         (GPG Key: 9283AA3F)

Re: Bind 'w' to a more useful function

[Reiner Steib]

On Mon, Jul 02 2007, Leo wrote:

> In summary buffer, 'w' is bound to a function that is barely used. I
> wonder would it make more sense to bind it to
> `gnus-article-fill-cited-article'.

Whether `gnus-summary-stop-page-breaking' deserves a single lowercase
key binding is discussable[1].  But `W w' is simple enough to type as
well.  If we'd remove `gnus-summary-stop-page-breaking' from `w', we
should reserve it for a more important function.

In the thread [1], Lars suggested:

,----[ http://article.gmane.org/gmane.emacs.gnus.general/62839 ]
| Here's an idea: How about if we unbind a couple of the blatantly silly
| single-letter bindings in the summary buffer now (`i' and `w' being
| excellent candidates) so that we can bind them at a later point (say,
| a few years) to more exciting commands?
`----

I won't object if someone would go ahead and [submit a patch to]
remove(s) the `i' and `w' from the code an the documentation in No
Gnus.

Bye, Reiner.

[1] In face I suggested `w' as a candidate for removing it's single
    letter binding:
    http://thread.gmane.org/gmane.emacs.gnus.general/61766/focus=62838
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: Bind 'w' to a more useful function

[Leo]

On 02/07/2007, Reiner Steib wrote:
> On Mon, Jul 02 2007, Leo wrote:
>
>> In summary buffer, 'w' is bound to a function that is barely used. I
>> wonder would it make more sense to bind it to
>> `gnus-article-fill-cited-article'.
>
> Whether `gnus-summary-stop-page-breaking' deserves a single lowercase
> key binding is discussable[1].  But `W w' is simple enough to type as
> well.  If we'd remove `gnus-summary-stop-page-breaking' from `w', we
> should reserve it for a more important function.

Do you have any candidate on your mind? altho W w is easy enough to
type, it is still painful in a group that most of the articles are
ugly. Those groups usually have users using outlook.

-- 
Leo <sdl.web AT gmail.com>                         (GPG Key: 9283AA3F)

Re: Bind 'w' to a more useful function

[Reiner Steib]

On Mon, Jul 02 2007, Leo wrote:

> On 02/07/2007, Reiner Steib wrote:
>> If we'd remove `gnus-summary-stop-page-breaking' from `w', we
>> should reserve it for a more important function.
>
> Do you have any candidate on your mind? 

No, we'd just reserve it for some future spiffy new feature as
discussed in the cited thread.  (Changing it's meaning at once might
be too abrupt, so a transition phase is a good idea so that previous
users off `w' aren't surprised to much.)

> altho W w is easy enough to type, it is still painful in a group
> that most of the articles are ugly. Those groups usually have users
> using outlook.

Maybe setting the corresponding treat variable
(`gnus-treat-fill-article'?) for this group is an option for you.  Or
just steal `w' if this function is important for you.  Or use `v' or
`v v' (`v ...' is reserved for the user).

Bye, Reiner.
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: Bind 'w' to a more useful function

[Lars Magne Ingebrigtsen]

Reiner Steib <reinersteib+gmane@imap.cc> writes:

> In the thread [1], Lars suggested:
>
> ,----[ http://article.gmane.org/gmane.emacs.gnus.general/62839 ]
> | Here's an idea: How about if we unbind a couple of the blatantly silly
> | single-letter bindings in the summary buffer now (`i' and `w' being
> | excellent candidates) so that we can bind them at a later point (say,
> | a few years) to more exciting commands?
> `----
>
> I won't object if someone would go ahead and [submit a patch to]
> remove(s) the `i' and `w' from the code an the documentation in No
> Gnus.

It was still in the code, so I removed it now.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Bind 'w' to a more useful function

[Leo]

On 2010-08-31 00:01 +0100, Lars Magne Ingebrigtsen wrote:
> It was still in the code, so I removed it now.

I bound w to gnus-article-fill-cited-article since these days more and
more people write one line per paragraph.

Leo

Re: Bind 'w' to a more useful function

[Lars Magne Ingebrigtsen]

Leo <sdl.web@gmail.com> writes:

> I bound w to gnus-article-fill-cited-article since these days more and
> more people write one line per paragraph.

Yes...  on the other hand, isn't it time that Gnus did this
automatically by default?  I mean, not
`gnus-article-fill-cited-article', but something that guesses whether
you're reading an unfilled article, and then run
`gnus-article-fill-cited-article' automatically if that seems to be the
case.

Although it's not altogether clear what the heuristics of that should
be...

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Bind 'w' to a more useful function

[Leo]

On 2010-08-31 14:00 +0100, Lars Magne Ingebrigtsen wrote:
>> I bound w to gnus-article-fill-cited-article since these days more and
>> more people write one line per paragraph.
>
> Yes...  on the other hand, isn't it time that Gnus did this
> automatically by default?  I mean, not
> `gnus-article-fill-cited-article', but something that guesses whether
> you're reading an unfilled article, and then run
> `gnus-article-fill-cited-article' automatically if that seems to be the
> case.
>
> Although it's not altogether clear what the heuristics of that should
> be...

I guess it is hard to do it right all the time. For example, I have read
emails with long lines mixed with code snippets. I like the long lines
to be filled but that usually screws up the code.

Leo

Re: Bind 'w' to a more useful function

[Lars Magne Ingebrigtsen]

Leo <sdl.web@gmail.com> writes:

> I guess it is hard to do it right all the time. For example, I have read
> emails with long lines mixed with code snippets. I like the long lines
> to be filled but that usually screws up the code.

Yes.  So we need a code snippet that's able to look at a (long) line and guess whether it's natural text or something else.

One heuristic would be if there's a lot of short word-constituent elements, and not too much space.

I mean, the following paragraph has a different signature in that dept than

                                                      (process (start-process
                                                                "images" nil "curl"
                                                                "-s" "--create-dirs"

to take a random example from a buffer.  (This message isn't filled so that I can use it as a test case.)

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Bind 'w' to a more useful function

[Leo]

On 2010-08-31 14:15 +0100, Lars Magne Ingebrigtsen wrote:
> I mean, the following paragraph has a different signature in that dept than
>
>                                                       (process (start-process
>                                                                 "images" nil "curl"
>                                                                 "-s" "--create-dirs"
>
> to take a random example from a buffer.  (This message isn't filled so that I can use it as a test case.)

That sounds nice.

Leo

Re: Bind 'w' to a more useful function

[Ted Zlatanov]

On Tue, 31 Aug 2010 15:15:00 +0200 Lars Magne Ingebrigtsen <larsi@gnus.org> wrote: 

LMI> Leo <sdl.web@gmail.com> writes:
>> I guess it is hard to do it right all the time. For example, I have read
>> emails with long lines mixed with code snippets. I like the long lines
>> to be filled but that usually screws up the code.

LMI> Yes.  So we need a code snippet that's able to look at a (long) line and guess whether it's natural text or something else.

LMI> One heuristic would be if there's a lot of short word-constituent elements, and not too much space.

LMI> I mean, the following paragraph has a different signature in that dept than

LMI>                                                       (process (start-process
LMI>                                                                 "images" nil "curl"
LMI>                                                                 "-s" "--create-dirs"

LMI> to take a random example from a buffer.  (This message isn't filled so that I can use it as a test case.)

I'm not optimistic.  Maybe Lisp+English will work, but natural text is
really hard to parse and source code too.

Ted

Re: Bind 'w' to a more useful function

[Lars Magne Ingebrigtsen]

Ted Zlatanov <tzz@lifelogs.com> writes:

> I'm not optimistic.  Maybe Lisp+English will work, but natural text is
> really hard to parse and source code too.

It is.  I've been googling around to try to see whether anybody has
established a "best effort" consensus anywhere on this point, and I
haven't found anything.

And I haven't been able to come up with anything really useful on my
own. 

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Bind 'w' to a more useful function

[Ted Zlatanov]

On Tue, 31 Aug 2010 20:33:09 +0200 Lars Magne Ingebrigtsen <larsi@gnus.org> wrote: 

LMI> Ted Zlatanov <tzz@lifelogs.com> writes:
>> I'm not optimistic.  Maybe Lisp+English will work, but natural text is
>> really hard to parse and source code too.

LMI> It is.  I've been googling around to try to see whether anybody has
LMI> established a "best effort" consensus anywhere on this point, and I
LMI> haven't found anything.

LMI> And I haven't been able to come up with anything really useful on my
LMI> own. 

You could try Markdown's logic:
http://daringfireball.net/projects/markdown/syntax#precode

"To produce a code block in Markdown, simply indent every line of the
block by at least 4 spaces or 1 tab"

So the only exception would be the leading line of a paragraph.

Ted

Re: Bind 'w' to a more useful function

[Lars Magne Ingebrigtsen]

Ted Zlatanov <tzz@lifelogs.com> writes:

> You could try Markdown's logic:
> http://daringfireball.net/projects/markdown/syntax#precode
>
> "To produce a code block in Markdown, simply indent every line of the
> block by at least 4 spaces or 1 tab"
>
> So the only exception would be the leading line of a paragraph.

    But people write stuff like this.  And it's not code...

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Bind 'w' to a more useful function

[Ted Zlatanov]

On Tue, 31 Aug 2010 21:03:57 +0200 Lars Magne Ingebrigtsen <larsi@gnus.org> wrote: 

LMI> Ted Zlatanov <tzz@lifelogs.com> writes:
>> You could try Markdown's logic:
>> http://daringfireball.net/projects/markdown/syntax#precode
>> 
>> "To produce a code block in Markdown, simply indent every line of the
>> block by at least 4 spaces or 1 tab"
>> 
>> So the only exception would be the leading line of a paragraph.

LMI>     But people write stuff like this.  And it's not code...

Generally (in English usage) a fixed indent denotes a quotation, so I
don't think it's a bad heuristic (assuming we're only talking about long
lines, not detecting code in general).

Ted

Re: Bind 'w' to a more useful function

[Steinar Bang]

>>>>> Ted Zlatanov <tzz@lifelogs.com>:

> I'm not optimistic.  Maybe Lisp+English will work, but natural text is
> really hard to parse and source code too.

For source code just add a dependency to the semantic bovinator.

(For the record: that was a joke!)