simple editor required

simple editor required

[Paul Edwards]

I'm wanting to use emacs as a substitute for vi.

I'm after the following features:

1. When I press the "a" key I want an "a" to appear on the screen
(that rules out "vi").

2. When I press the "delete" key I expect the current character
to be deleted, not the previous character (that rules out Emacs 20.7.1).

3. When I hit the enter key I expect the next line to begin underneath
the last line (that rules out Xemacs 20.4).

4. When I attempt to mark a block of text, I don't expect a find or
a goto to lose my marker (that rules out Xemacs 20.4).

5. As per (3), when I have a blank line, I don't want it to insert
spaces.  I don't mind it inserting spaces so much as I want to be
able to delete them with ctrl-k.  But when I do that it deletes my
blank line.  :-(  So I have to delete one character at a time instead.

Now due to normal beauracacy, it is difficult for me to get a
different version of emacs installed, and on some machines I
only have 20.4 and other machines only 20.7.1, so I'd prefer
both to work.  Is there any workaround?

I can't seem to stop it getting into C mode when I edit a ".c"
file.  Basically I don't want any junk loaded, I'm just after a
basic editor, ie "text mode".  I thought inhibit-default-init
would stop it from loading junk, but no such luck.

Here is my .emacs:

(setq-default tab-width 4)

(global-set-key [?\C-h] 'delete-backward-char)

(setq make-backup-files nil)

(setq inhibit-default-init 't)

(global-set-key [insertchar] 'overwritemode)

(setq auto-save-default nil)

(setq-default indent-tabs-mode nil)

;(setq-default indent-according-to-mode nil)

(setq-default default-major-mode 'text-mode)

;(global-set-key "\C-m" 'newline-and-indent)
(global-set-key [?\C-m] 'newline-and-indent)

(global-set-key "\M-g" 'goto-line)

(global-set-key "\M-r" 'replace-string)

(global-unset-key [delete])
(global-set-key [delete] 'delete-char)

; potential alternatives (nothing works on 20.7.1)
;(global-unset-key [deletechar])
;(global-set-key [deletechar] 'delete-char)
;(global-unset-key (kbd "<delete>"))
;(global-set-key (kbd "<delete>") 'delete-char)


Thanks.  Paul.

Re: simple editor required

[Kevin Dziulko]

You can do something like:

(global-set-key "\e[4~" 'delete-char)

to get the delete key to work as you'd like. (\e[4~ would need to be 
subsituted by the escape sequence of your delete key).  To find your 
delete keys escape sequence, go into the *scratch* buffer, hit C-q, then 
press your delete key.

use something like:
(setq auto-mode-alist (cons '("\\.c$" . text-mode) auto-mode-alist))

so that visiting *.c files will land you in text-mode.

Good Luck!


On Tue, 3 Jun 2003, Paul Edwards wrote:

> I'm wanting to use emacs as a substitute for vi.
> 
> I'm after the following features:
> 
> 1. When I press the "a" key I want an "a" to appear on the screen
> (that rules out "vi").
> 
> 2. When I press the "delete" key I expect the current character
> to be deleted, not the previous character (that rules out Emacs 20.7.1).
> 
> 3. When I hit the enter key I expect the next line to begin underneath
> the last line (that rules out Xemacs 20.4).
> 
> 4. When I attempt to mark a block of text, I don't expect a find or
> a goto to lose my marker (that rules out Xemacs 20.4).
> 
> 5. As per (3), when I have a blank line, I don't want it to insert
> spaces.  I don't mind it inserting spaces so much as I want to be
> able to delete them with ctrl-k.  But when I do that it deletes my
> blank line.  :-(  So I have to delete one character at a time instead.
> 
> Now due to normal beauracacy, it is difficult for me to get a
> different version of emacs installed, and on some machines I
> only have 20.4 and other machines only 20.7.1, so I'd prefer
> both to work.  Is there any workaround?
> 
> I can't seem to stop it getting into C mode when I edit a ".c"
> file.  Basically I don't want any junk loaded, I'm just after a
> basic editor, ie "text mode".  I thought inhibit-default-init
> would stop it from loading junk, but no such luck.
> 
> Here is my .emacs:
> 
> (setq-default tab-width 4)
> 
> (global-set-key [?\C-h] 'delete-backward-char)
> 
> (setq make-backup-files nil)
> 
> (setq inhibit-default-init 't)
> 
> (global-set-key [insertchar] 'overwritemode)
> 
> (setq auto-save-default nil)
> 
> (setq-default indent-tabs-mode nil)
> 
> ;(setq-default indent-according-to-mode nil)
> 
> (setq-default default-major-mode 'text-mode)
> 
> ;(global-set-key "\C-m" 'newline-and-indent)
> (global-set-key [?\C-m] 'newline-and-indent)
> 
> (global-set-key "\M-g" 'goto-line)
> 
> (global-set-key "\M-r" 'replace-string)
> 
> (global-unset-key [delete])
> (global-set-key [delete] 'delete-char)
> 
> ; potential alternatives (nothing works on 20.7.1)
> ;(global-unset-key [deletechar])
> ;(global-set-key [deletechar] 'delete-char)
> ;(global-unset-key (kbd "<delete>"))
> ;(global-set-key (kbd "<delete>") 'delete-char)
> 
> 
> Thanks.  Paul.
> 
> 
> _______________________________________________
> Help-gnu-emacs mailing list
> Help-gnu-emacs@gnu.org
> http://mail.gnu.org/mailman/listinfo/help-gnu-emacs
> 

Re: simple editor required

[Bruce Ingalls]

Paul Edwards wrote:
> I'm wanting to use emacs as a substitute for vi.
> 
> I'm after the following features:
> 
> 1. When I press the "a" key I want an "a" to appear on the screen
> (that rules out "vi").

Add http://cream.sourceforge.net/ to your vi.

http://emacro.sf.net/ is the equivalent for emacs. However, you may
find yourself turning off features.

> 4. When I attempt to mark a block of text, I don't expect a find or
> a goto to lose my marker (that rules out Xemacs 20.4).

Rules out most editors.
I would expect to copy or cut selected text, before moving on to another
command.
How do you expect to unselect text?

Emacs is designed to be powerful, not simple.
Try nedit or pico for less featureful editors.

Re: simple editor required

[Paul Edwards]

"Bruce Ingalls" <bingalls@CUT-this-SPAM-BLOCK.fit-zones.com> wrote in message news:6V2Da.96131$h42.53153@twister.nyc.rr.com...
> Paul Edwards wrote:
> > I'm wanting to use emacs as a substitute for vi.
> >
> > I'm after the following features:
> >
> > 1. When I press the "a" key I want an "a" to appear on the screen
> > (that rules out "vi").
>
> Add http://cream.sourceforge.net/ to your vi.
>
> http://emacro.sf.net/ is the equivalent for emacs. However, you may
> find yourself turning off features.
>
> > 4. When I attempt to mark a block of text, I don't expect a find or
> > a goto to lose my marker (that rules out Xemacs 20.4).
>
> Rules out most editors.
> I would expect to copy or cut selected text, before moving on to another
> command.

emacs 20.7.1 behaves as I expect.  I do ctrl-escape to start
marking, and then I can find or goto to get to the spot where
I want the block to end, and then I go ctrl-w to cut.

xemacs loses my marker.

> How do you expect to unselect text?

ctrl-g.

> Emacs is designed to be powerful, not simple.
> Try nedit or pico for less featureful editors.

I'm already familiar with the emacs commands, it's just that I
seem to get different results with every new emacs I encounter.

It's like they keep changing the defaults just to annoy me, and I
need to find some arcane thing to put it back to normal.

BFN.  Paul.

Re: simple editor required

[Bruce Ingalls]

Paul Edwards wrote:
> "Bruce Ingalls" <bingalls@CUT-this-SPAM-BLOCK.fit-zones.com> wrote in message news:6V2Da.96131$h42.53153@twister.nyc.rr.com...
...
> emacs 20.7.1 behaves as I expect.  I do ctrl-escape to start
> marking, and then I can find or goto to get to the spot where
> I want the block to end, and then I go ctrl-w to cut.
> 
> xemacs loses my marker.
> 
>>How do you expect to unselect text?
> 
> ctrl-g.
> 
>>Emacs is designed to be powerful, not simple.
>>Try nedit or pico for less featureful editors.
> 
> I'm already familiar with the emacs commands, it's just that I
> seem to get different results with every new emacs I encounter.
> 
> It's like they keep changing the defaults just to annoy me, and I
> need to find some arcane thing to put it back to normal.

Since you are an accomplished Emacs user, your post is starting to take
on "troll"-like elements.

You ask for a featureless, simple editor, then ask for a feature that no
editor provides. The good news is that Emacs makes more things possible,
than with other editors. The bad news, is
1) this is what makes Emacs powerful, and not simple
2) Emacs is run by the community.

You are somehow looking at the above added/optional feature, and 
wondering why it disappears from some releases, which perhaps stick to 
the core, standard features.

Emacs could mention that some features are enhanced options. If you know 
of an effective way to do so, such that this does not unnecessarily get 
users' hopes up, let us know.

More bad news:
For a short term solution, you will either have to hire a consultant, or 
write a .emacs file.
The good news, is that free software makes this possible, and the Emacs 
community is helpful, in particular. "Free" does not mean "Free beer", 
unless that is what you use to compensate others for giving up their 
"Free" time.

Here is the really good news:
Once you get this elisp working, you release it into the free software 
domain. Others *may* make it available, and it might even make its way 
into future releases of Emacs and XEmacs, making your feature standard.
Or, it might get picked up, and tweaked.

So here is a big hint:
Grab the code, from the version of X/Emacs, which works the way you 
want. See if you can not finish adapting it to the version you are 
working with, within a reasonable effort.
If there is a small, as in well defined problem, which is keeping you 
completing your dream feature, come back to the community with this, and 
I'm confident that you will reach your goals, else learn why they are 
infeasible.

Good Luck!

Re: simple editor required

[Paul Edwards]

"Bruce Ingalls" <bingalls@CUT-this-SPAM-BLOCK.fit-zones.com> wrote in message news:X0mDa.1570$Fu.866812@twister.nyc.rr.com...
> > emacs 20.7.1 behaves as I expect.  I do ctrl-escape to start
> > marking, and then I can find or goto to get to the spot where
> > I want the block to end, and then I go ctrl-w to cut.
> >
> > xemacs loses my marker.
> >
> >>How do you expect to unselect text?
> >
> > ctrl-g.
> >
> >>Emacs is designed to be powerful, not simple.
> >>Try nedit or pico for less featureful editors.
> >
> > I'm already familiar with the emacs commands, it's just that I
> > seem to get different results with every new emacs I encounter.
> >
> > It's like they keep changing the defaults just to annoy me, and I
> > need to find some arcane thing to put it back to normal.
>
> Since you are an accomplished Emacs user,

Not sure about "accomplished".  I know the basics of
micro-emacs, and thus also the basics of full-emacs.

> your post is starting to take on "troll"-like elements.

[speech deleted].

> You ask for a featureless, simple editor, then ask for a feature that no

No, I asked how to stop emacs setting up a whole lot of crap that I
don't want, by default, when all I am after is making a powerful
editor get the basics right, given that it doesn't work as a default.

> editor provides. The good news is that Emacs makes more things possible,
> than with other editors. The bad news, is
> 1) this is what makes Emacs powerful, and not simple
> 2) Emacs is run by the community.
>
> You are somehow looking at the above added/optional feature, and
> wondering why it disappears from some releases, which perhaps stick to
> the core, standard features.

I'm wondering why the basics keep changing and aren't
consistent.  Like even marking a block of text.

> Emacs could mention that some features are enhanced options. If you know
> of an effective way to do so, such that this does not unnecessarily get
> users' hopes up, let us know.

Not sure what you're talking about there.

> More bad news:
> For a short term solution, you will either have to hire a consultant, or
> write a .emacs file.

I have a .emacs file, and it behaves differently to each other, plus
different to previous versions, even by default.  Fundamentally
changing the basics.  I only use the basics, and every single company
I go to the .emacs needs to be changed.  Even to get the delete key
to work.  If it is even possible.  That is very strange and has nothing
to do with advanced features.  This is fundamental basic editting
operations.

BFN.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> emacs 20.7.1 behaves as I expect.  I do ctrl-escape to start
> marking, and then I can find or goto to get to the spot where
> I want the block to end, and then I go ctrl-w to cut.
>
> xemacs loses my marker.

Try C-x C-x before C-w.  Also see the variable zmacs-regions (or similar).
-- 
This line is not blank.

RE: simple editor required

[Dmitri.Minaev]

vim -c ':help easy'

--
With best regards,
Dmitri Minaev


> -----Original Message-----
> From: Paul Edwards [mailto:kerravon@nosppaam.w3.to]
> Sent: Tuesday, June 03, 2003 2:47 PM
> To: help-gnu-emacs@gnu.org
> Subject: simple editor required
> 
> 
> I'm wanting to use emacs as a substitute for vi.
> 
> I'm after the following features:
> 
> 1. When I press the "a" key I want an "a" to appear on the screen
> (that rules out "vi").
> 
> 2. When I press the "delete" key I expect the current character
> to be deleted, not the previous character (that rules out 
> Emacs 20.7.1).
> 
> 3. When I hit the enter key I expect the next line to begin underneath
> the last line (that rules out Xemacs 20.4).
> 
> 4. When I attempt to mark a block of text, I don't expect a find or
> a goto to lose my marker (that rules out Xemacs 20.4).
> 
> 5. As per (3), when I have a blank line, I don't want it to insert
> spaces.  I don't mind it inserting spaces so much as I want to be
> able to delete them with ctrl-k.  But when I do that it deletes my
> blank line.  :-(  So I have to delete one character at a time instead.
> 
> Now due to normal beauracacy, it is difficult for me to get a
> different version of emacs installed, and on some machines I
> only have 20.4 and other machines only 20.7.1, so I'd prefer
> both to work.  Is there any workaround?
> 
> I can't seem to stop it getting into C mode when I edit a ".c"
> file.  Basically I don't want any junk loaded, I'm just after a
> basic editor, ie "text mode".  I thought inhibit-default-init
> would stop it from loading junk, but no such luck.
> 
> Here is my .emacs:
> 
> (setq-default tab-width 4)
> 
> (global-set-key [?\C-h] 'delete-backward-char)
> 
> (setq make-backup-files nil)
> 
> (setq inhibit-default-init 't)
> 
> (global-set-key [insertchar] 'overwritemode)
> 
> (setq auto-save-default nil)
> 
> (setq-default indent-tabs-mode nil)
> 
> ;(setq-default indent-according-to-mode nil)
> 
> (setq-default default-major-mode 'text-mode)
> 
> ;(global-set-key "\C-m" 'newline-and-indent)
> (global-set-key [?\C-m] 'newline-and-indent)
> 
> (global-set-key "\M-g" 'goto-line)
> 
> (global-set-key "\M-r" 'replace-string)
> 
> (global-unset-key [delete])
> (global-set-key [delete] 'delete-char)
> 
> ; potential alternatives (nothing works on 20.7.1)
> ;(global-unset-key [deletechar])
> ;(global-set-key [deletechar] 'delete-char)
> ;(global-unset-key (kbd "<delete>"))
> ;(global-set-key (kbd "<delete>") 'delete-char)
> 
> 
> Thanks.  Paul.
> 
> 
> _______________________________________________
> Help-gnu-emacs mailing list
> Help-gnu-emacs@gnu.org
> http://mail.gnu.org/mailman/listinfo/help-gnu-emacs
> 

Re: simple editor required

[Paul Edwards]

"Kevin Dziulko" <dziulko@klaatu.canisius.edu> wrote in message news:mailman.7200.1054643007.21513.help-gnu-emacs@gnu.org...
> You can do something like:
>
> (global-set-key "\e[4~" 'delete-char)
>
> to get the delete key to work as you'd like. (\e[4~ would need to be
> subsituted by the escape sequence of your delete key).  To find your
> delete keys escape sequence, go into the *scratch* buffer, hit C-q, then
> press your delete key.

Thanks.  I'll give that a go.

> use something like:
> (setq auto-mode-alist (cons '("\\.c$" . text-mode) auto-mode-alist))
>
> so that visiting *.c files will land you in text-mode.

Ok, I'll try that.  Can you tell me why these...

> > (setq inhibit-default-init 't)
> > (setq-default default-major-mode 'text-mode)

... didn't stop this auto-file-association from happening?  I basically
don't want emacs to do any junk like that, I'm not interested in its
assumptions, I'm after basic editting functions, rather than by
default being launched into its weird ideas of what C code should
look like (which is probably a style used by less than 10% of
C programmers).

BFN.  Paul.

Re: simple editor required

[Stefan Monnier]

>> (global-set-key "\e[4~" 'delete-char)

This might work sometimes but is not The Right Way.  Of course in Emacs-21,
all this is unnecessary since it works the way you want by default.
I'd recommend something like

    (define-key function-key-map [delete] [?\C-d])

and probably also:

    (define-key function-key-map "\e[4~" [?\C-d])

(the first should work for X11 and W32, while the second is for use on
a text-terminal or `xterm' kind of thing).

>> use something like:
>> (setq auto-mode-alist (cons '("\\.c$" . text-mode) auto-mode-alist))

Should be "\\.c\\'", although it only makes a difference if there's
a newline in your file name.

>> > (setq inhibit-default-init 't)

This says not to load the local `default.el' initialization file that
your sysadmin might have written.  It doesn't change anything to Emacs's
own defaults.

>> > (setq-default default-major-mode 'text-mode)

This just tells Emacs to use text-mode rather than fundamental-mode for
files for which no other mode was found.  I.e. it's rarely used.
And BTW, if you want "a plain editor", you might prefer fundamental-mode
rather than text-mode.

As for your previous requests, try:

   (global-set-key [?\C-m] 'newline-and-indent)


-- Stefan

RE: simple editor required

[Ben Key]

Stefan Monnier wrote:
> Should be "\\.c\\'", although it only makes a difference if there's a
newline in your file name.

I did not know it was possible to have a new line character in a file
name.  At least I do not think it is possible in DOS or MS Windows.  Is
it possible to have a new line character in a file name in Gnu Linux /
Unix?

Re: simple editor required

[Eli Zaretskii]

> From: "Ben Key" <bkey1@tampabay.rr.com>
> Date: Tue, 10 Jun 2003 21:53:21 -0400
> 
> I did not know it was possible to have a new line character in a file
> name.  At least I do not think it is possible in DOS or MS Windows.  Is
> it possible to have a new line character in a file name in Gnu Linux /
> Unix?

It's true that Microsoft's filesystems don't allow newlines (and
quite a few other characters, like `*' and `>') in file anmes, but
Unix and GNU/Linux systems do allow them.

Re: simple editor required

[Peter Lee]

>>>> On Tue, 03 Jun 2003 13:51:57 GMT, "Paul Edwards" <kerravon@nosppaam.w3.to> said:

    Paul> ... didn't stop this auto-file-association from happening?
    Paul> I basically don't want emacs to do any junk like that, I'm
    Paul> not interested in its assumptions, I'm after basic editting
    Paul> functions, rather than by default being launched into its
    Paul> weird ideas of what C code should look like (which is
    Paul> probably a style used by less than 10% of C programmers).

There are several styles provided by emacs (see c-default-style and
c-set-style).  You can also create your own style, or you can simply
change an existing style to suit your needs (which is what I do).

(defun my-c-mode-common-hook ()
  (turn-on-font-lock)
  (c-set-offset 'substatement-open 0)
  (c-set-offset 'arglist-intro 1)
  (c-set-offset 'defun-open 4)
  (c-set-offset 'defun-block-intro 0)
  (c-set-offset 'statement 0)
  (c-set-offset 'substatement-open 4)
  (c-set-offset 'statement-block-intro 0)
  (c-set-offset 'substatement 4)
  (c-set-offset 'statement-cont 4)
  (c-set-offset 'label -4)
  (c-set-offset 'topmost-intro-cont 4)
  (c-set-offset 'brace-list-intro 0)
  (c-set-offset 'class-open 4)
  (c-set-offset 'class-close 4)
  (c-set-offset 'topmost-intro '-)
  (c-set-offset 'brace-list-open 4)
  (c-set-offset 'access-label '--)
  (c-set-offset 'func-decl-cont 0)
  (c-set-offset 'case-label 0)
  (c-set-offset 'statement-case-intro 4)
  (c-set-offset 'statement-case-open 4)
  (c-set-offset 'inline-open '+)
  (c-toggle-auto-hungry-state 1)
  (local-set-key [?\C-c ?\C-i] 'c-includes-current-file)
  (define-abbrev local-abbrev-table "ifx" "" 'my-skeleton-c-if)
  (define-abbrev local-abbrev-table "forx" "" 'my-skeleton-c-for)
  (define-abbrev local-abbrev-table "whilex" "" 'my-skeleton-c-while)
  (define-abbrev local-abbrev-table "switchx" "" 'my-skeleton-c-switch))

(add-hook 'c-mode-common-hook 'my-c-mode-common-hook)


If you really do want text mode for c files then specify it with the
auto-mode-alist:

(setq auto-mode-alist
      (append '(("\\.c$"    . text-mode)
                ("\\.cc$"   . text-mode)
                ("\\.C$"    . text-mode)
                ("\\.CC$"   . text-mode)
                ("\\.h$"    . text-mode)
                ("\\.H$"    . text-mode))
      auto-mode-alist))

Re: simple editor required

[Kevin Rodgers]

Paul Edwards wrote:

> "Kevin Dziulko" <dziulko@klaatu.canisius.edu> wrote in message news:mailman.7200.1054643007.21513.help-gnu-emacs@gnu.org...
>>use something like:
>>(setq auto-mode-alist (cons '("\\.c$" . text-mode) auto-mode-alist))
>>
>>so that visiting *.c files will land you in text-mode.
> 
> Ok, I'll try that.  Can you tell me why these...
> 
>>>(setq inhibit-default-init 't)
>>>(setq-default default-major-mode 'text-mode)
>>>
> 
> ... didn't stop this auto-file-association from happening?  I basically
> don't want emacs to do any junk like that, I'm not interested in its
> assumptions, I'm after basic editting functions,


Leave default-major-mode as is (fundamental-mode).  Instead:

(setq auto-mode-alist '()
       interpreter-mode-alist '())

> rather than by
> default being launched into its weird ideas of what C code should
> look like (which is probably a style used by less than 10% of
> C programmers).

Weird is in the eye of the beholder.


-- 
<a href="mailto:&lt;kevin.rodgers&#64;ihs.com&gt;">Kevin Rodgers</a>

Re: simple editor required

[kgold]

Probably won't convince you, but ...

I consider the programming modes one of the 2 most "must have"
features.  

Actually, emacs' "weird ideas of what C code should look
like" is whatever you want.  There are about 1/2 dozen basic modes,
with different bracket and indent styles.  From there, it's simple to
customize to whatever you think emacs should look like.

And you also get 
- source code coloring
- indexes of functions
- automatic indentation, however you like it
- expansion of macros
- keystrokes to comment and uncomment code

With your "basic text editor", how would you even run a spell checker
on your code without getting false hits on every variable name?

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:
> I'm after basic editting functions, rather than by default being
> launched into its weird ideas of what C code should look like (which
> is probably a style used by less than 10% of C programmers).

-- 
-- 
Ken Goldman   kgold@watson.ibm.com   914-784-7646

Re: simple editor required

[Jon Kåre Hellan]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> "Kevin Dziulko" <dziulko@klaatu.canisius.edu> wrote in message news:mailman.7200.1054643007.21513.help-gnu-emacs@gnu.org...
> ... didn't stop this auto-file-association from happening?  I basically
> don't want emacs to do any junk like that, I'm not interested in its
> assumptions, I'm after basic editting functions, rather than by
> default being launched into its weird ideas of what C code should
> look like (which is probably a style used by less than 10% of
> C programmers).

You should certainly do it your own way. But for me, the point isn't
whether or not I like or dislike the way emacs formats the code. The
point is that when it formats it in a way that surprises me, it's most
often because the code isn't what I thought it was, i.e. it caught me
making a mistake.

Good luck

Jon

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> ... didn't stop this auto-file-association from happening?  I basically
> don't want emacs to do any junk like that, I'm not interested in its
> assumptions, I'm after basic editting functions, rather than by
> default being launched into its weird ideas of what C code should
> look like (which is probably a style used by less than 10% of
> C programmers).

Please try typing C-c . in a C file, then use tab completion to find
the different alternatives.  Try them.  Maybe you find one which is
close to what you like.  When you have it, it can be tweaked some
more to get closer.

It's not always possible to get the *exact* indentation you specify,
but I find that syntax-driven indentation helps me so much with my
work that I just adapt my coding style to the way Emacs likes it.
For example, the indentation can tell me right away that a brace is
missing or too many.
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2525 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:84of1c78ka.fsf@lucy.is.informatik.uni-duisburg.de...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > ... didn't stop this auto-file-association from happening?  I basically
> > don't want emacs to do any junk like that, I'm not interested in its
> > assumptions, I'm after basic editting functions, rather than by
> > default being launched into its weird ideas of what C code should
> > look like (which is probably a style used by less than 10% of
> > C programmers).
>
> Please try typing C-c . in a C file, then use tab completion to find
> the different alternatives.  Try them.  Maybe you find one which is
> close to what you like.  When you have it, it can be tweaked some
> more to get closer.
>
> It's not always possible to get the *exact* indentation you specify,
> but I find that syntax-driven indentation helps me so much with my
> work that I just adapt my coding style to the way Emacs likes it.
> For example, the indentation can tell me right away that a brace is
> missing or too many.

The thing is, the code that I am changing is not just one style,
I have to match the code that I am editting.  So I would rather
just have very basic editting working by default.  ie I don't
mind have to press space 4 times after typing a "{", but I do
mind having to type 20 spaces after a "{".

Officially I am expected to use vi "the standard editor", but
I don't know the commands for it, and I hate it.  Every place I
go to has had emacs installed, often unofficially, so I have
been able to use that, and get most of the commands I am
familiar with from micro-emacs.  But I can't complain to the
sysadmins etc that I need a new version, and I'm not meant to
be spending time mucking around with that either.

All I want is for the basic editting functions to work, at least
consistently within each emacs version, if not identical to
micro-emacs.  I'll worry about fancy features like C indentation
another day.  I have been given a few things to try to see if I can
get closer to that goal.

Particularly frustrating is when something like the "del" key
doesn't do what I expect (not just what I expect, but what other
versions of emacs do when the appropriate command is
given).  As I turn up to a new job, don't know how to use vi,
and it appears that I don't even know how to delete a character.
If I have micro-emacs (or emacs was consistent between
versions so my .emacs did the right thing), it wouldn't be a
problem.

BFN.  Paul.

Re: simple editor required

[Eli Zaretskii]

> From: "Paul Edwards" <kerravon@nosppaam.w3.to>
> Newsgroups: gnu.emacs.help
> Date: Fri, 06 Jun 2003 06:15:19 GMT
> 
> The thing is, the code that I am changing is not just one style,
> I have to match the code that I am editting.  So I would rather
> just have very basic editting working by default.  ie I don't
> mind have to press space 4 times after typing a "{", but I do
> mind having to type 20 spaces after a "{".

Does it help to type "M-x fundamental-mode RET" befiore editing the
code?  Does that do what you want?

Re: simple editor required

[Harry Putnam]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> Particularly frustrating is when something like the "del" key
> doesn't do what I expect (not just what I expect, but what other
> versions of emacs do when the appropriate command is
> given).  As I turn up to a new job, don't know how to use vi,
> and it appears that I don't even know how to delete a character.
> If I have micro-emacs (or emacs was consistent between
> versions so my .emacs did the right thing), it wouldn't be a
> problem.

Sounds like you need to bone up on vi before you go much further with
your employment quests.  An employer expecting you to know vi isn't
asking a terrible lot after all.

Or expecting an employee to know about basic OS stuff like setting
the erase character in shell.. really isn't asking a terrible lot
either.

Both are very basic tools of the trade.  If you can't or won't learn
them you probably need a different trade.  Not to be snotty but your
post comes off like a cry baby who can't do his job.

A couple of evenings spent with a vi tutorial will get you lined out.
Then start worrying about customizing emacs or changing defaults.

Many basic unix books have vi sections.

Re: simple editor required

[Peter Lee]

>>>> Harry Putnam writes:
    Harry> Sounds like you need to bone up on vi before you go much
    Harry> further with your employment quests.  An employer expecting
    Harry> you to know vi isn't asking a terrible lot after all.

Why should the employer care what tool you use to get a particular job
done with?  I mean unless your job is to create vi add-ons (or
something actually targetting vi) what difference does it make how the
text is generated?  If I ever found myself faced with such a
brain-dead employer I'd have to move along.

    Harry> Or expecting an employee to know about basic OS stuff like
    Harry> setting the erase character in shell.. really isn't asking
    Harry> a terrible lot either.

Not sure how you draw a correlation between expecting a developer to
know the basics of the OS he's developing on/for and dictating which
editor he will use.  The former is both obvious and common sense while
the latter is just silly.

Re: simple editor required

[Harry Putnam]

Peter Lee <pete_lee@swbell.net> writes:

>>>>> Harry Putnam writes:
>     Harry> Sounds like you need to bone up on vi before you go much
>     Harry> further with your employment quests.  An employer expecting
>     Harry> you to know vi isn't asking a terrible lot after all.
>
> Why should the employer care what tool you use to get a particular job
> done with?  I mean unless your job is to create vi add-ons (or
> something actually targetting vi) what difference does it make how the
> text is generated?  If I ever found myself faced with such a
> brain-dead employer I'd have to move along.

Not unreasonable for an employer to expect competency with available
tools.  I spent my life in heavy construction.  I never had the
oportuninty to say ' I don't know how to use Miller welding machines.
I only use lincolns.  Could you get a lincoln for me even though you
already have millers?' 

I wouldn't have lasted long..  

>     Harry> Or expecting an employee to know about basic OS stuff like
>     Harry> setting the erase character in shell.. really isn't asking
>     Harry> a terrible lot either.

> Not sure how you draw a correlation between expecting a developer to
> know the basics of the OS he's developing on/for and dictating which
> editor he will use.  The former is both obvious and common sense while
> the latter is just silly.

You are giving the OP bad advice here.  Its only basic self protection
to know how to use the tools you work with.  Knowing how to set the
erase behavior in your shell or other minor adjustments.

I don't know what bee got up your bonnet but what your saying is
really a bit ridiculous.

Re: simple editor required

[Peter Lee]

>>>> Harry Putnam writes:

    Harry> Not unreasonable for an employer to expect competency with
    Harry> available tools.  I spent my life in heavy construction.  I
    Harry> never had the oportuninty to say ' I don't know how to use
    Harry> Miller welding machines.  I only use lincolns.  Could you
    Harry> get a lincoln for me even though you already have millers?'

The difference is the tools we're talking about are both freely
available at no cost.  It doesn't cost the employer anything one way
or the other.  It's like me walking into an interview with a craftsman
hammer and having the employer say, 'sorry we only use brandX hammers
here'.

    Harry> You are giving the OP bad advice here.  Its only basic self
    Harry> protection to know how to use the tools you work with.

I never said he shouldn't know how to use a hammer... only that the
employer shouldn't care which kind he uses.

    Harry> I don't know what bee got up your bonnet but what your
    Harry> saying is really a bit ridiculous.

Long day sorry.

Re: simple editor required

[David Masterson]

>>>>> Peter Lee writes:

>>>>> Harry Putnam writes:

> Why should the employer care what tool you use to get a particular
> job done with?  

Often, the reason (however misguided) is security.  The employer has
qualified some tools as being both useful and secure in their
environment and are unwilling to let anything else.

Another reason is that many companies do not invest in *FULL* backup
capabilities, so, if you spend a lot of time setting up a special,
unapproved environment, you may lose it at the first system crash.

-- 
David Masterson                David DOT Masterson AT synopsys DOT com
Sr. R&D Engineer               Synopsys, Inc.
Software Engineering           Sunnyvale, CA

Re: simple editor required

[kgold]

While I'm as emacs-fanatic as anyone here, it's still good to know
some vi.  vi comes standard with Unix and emacs sometimes does not.
Some day you'll be at a customer site, and vi will be all they have.

Peter Lee <pete_lee@swbell.net> writes:
> 
> Why should the employer care what tool you use to get a particular job
> done with?  I mean unless your job is to create vi add-ons (or
> something actually targetting vi) what difference does it make how the
> text is generated?  If I ever found myself faced with such a
> brain-dead employer I'd have to move along.

-- 
Ken Goldman   kgold@watson.ibm.com   914-784-7646

Re: simple editor required

[Eli Zaretskii]

> From: kgold@watson.ibm.com (kgold)
> Newsgroups: gnu.emacs.help
> Date: 11 Jun 2003 21:03:05 GMT
> 
> Some day you'll be at a customer site, and vi will be all they have.

Perish the thought ;-)

Anyway, I find the time spent downloading and building Emacs on such a
site a better investment than the time spent pressing the ESC key in
vi and cursing when I forget to do that.

Re: simple editor required

[Benjamin Rutt]

"Eli Zaretskii" <eliz@elta.co.il> writes:

> Anyway, I find the time spent downloading and building Emacs on such a
> site a better investment than the time spent pressing the ESC key in
> vi and cursing when I forget to do that.

My favorite is when you are in vi insert mode and you decide to use
the arrow keys to move the cursor.  Gee, I really love having A, B, C
or D inserted when I hit the keys which should take me up, down left
or right.  Way to go, vi!
-- 
Benjamin

Re: simple editor required

[Kin Cho]

Benjamin Rutt <rutt+news@cis.ohio-state.edu> writes:

> "Eli Zaretskii" <eliz@elta.co.il> writes:
> 
> > Anyway, I find the time spent downloading and building Emacs on such a
> > site a better investment than the time spent pressing the ESC key in
> > vi and cursing when I forget to do that.
> 
> My favorite is when you are in vi insert mode and you decide to use
> the arrow keys to move the cursor.  Gee, I really love having A, B, C
> or D inserted when I hit the keys which should take me up, down left
> or right.  Way to go, vi!
> -- 
> Benjamin

However, If you'd haven't forgotten to press ESC, you'd be
rewarded with the ability to use the much quicker kjhl keys to
move around.

-kin

Re: simple editor required

[Benjamin Rutt]

Kin Cho <kin@techie.com> writes:

> However, If you'd haven't forgotten to press ESC, you'd be
> rewarded with the ability to use the much quicker kjhl keys to
> move around.

Sorry, you already lost so much time lifting your hand off the
keyboard to hit the ESC key and re-orient that hand back on the home
row that any efficiency arguments are lost.
-- 
Benjamin

Re: simple editor required

[Kin Cho]

Benjamin Rutt <rutt+news@cis.ohio-state.edu> writes:

> Sorry, you already lost so much time lifting your hand off the
> keyboard to hit the ESC key and re-orient that hand back on the home
> row that any efficiency arguments are lost.

I suppose the arrow keys are in the home row on your keyboard?

I'm right handed, so my right alt, ctrl, and shift keys are no
use to me.  A long time ago I remapped right shift to be escape
and viper becomes even more pleasant to use.

-kin

Re: simple editor required

[Peter Lee]

>>>> Kin Cho writes:
    Kin> I'm right handed, so my right alt, ctrl, and shift keys are
    Kin> no use to me.  A long time ago I remapped right shift to be
    Kin> escape and viper becomes even more pleasant to use.

You could just as easily remap the alt, ctrl, and shift keys just like
you did the escape.

Re: simple editor required

[Benjamin Rutt]

Kin Cho <kin@techie.com> writes:

> I suppose the arrow keys are in the home row on your keyboard?

No, they're not.  Also, in the sane applications I use, they don't
emit A, B, C or D when I press them.
-- 
Benjamin

Re: simple editor required

[Galen Boyer]

On 11 Jun 2003, kgold@watson.ibm.com wrote:
> 
> Some day you'll be at a customer site, and vi will be all they have.

I've been on many customer sites, and I've yet to see this.  I'm
starting to think this is one of those myths that continues to survive,
for some reason.
-- 
Galen deForest Boyer
Sweet dreams and flying machines in pieces on the ground.

Re: simple editor required

[Kin Cho]

Galen Boyer <galenboyer@hotpop.com> writes:

> On 11 Jun 2003, kgold@watson.ibm.com wrote:
> > 
> > Some day you'll be at a customer site, and vi will be all they have.
> 
> I've been on many customer sites, and I've yet to see this.  I'm
> starting to think this is one of those myths that continues to survive,
> for some reason.

Today I have to look into a problem our QA reported in one of
their test Solaris box.  No emacs, not even vim, just plain vi.

-kin

Re: simple editor required

[Paul Edwards]

"Harry Putnam" <reader@newsguy.com> wrote in message news:mailman.7492.1054922593.21513.help-gnu-emacs@gnu.org...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > Particularly frustrating is when something like the "del" key
> > doesn't do what I expect (not just what I expect, but what other
> > versions of emacs do when the appropriate command is
> > given).  As I turn up to a new job, don't know how to use vi,
> > and it appears that I don't even know how to delete a character.
> > If I have micro-emacs (or emacs was consistent between
> > versions so my .emacs did the right thing), it wouldn't be a
> > problem.
>
> Sounds like you need to bone up on vi before you go much further with
> your employment quests.  An employer expecting you to know vi isn't
> asking a terrible lot after all.

I can touch-type and I can write code with blinding productivity,
almost automatically.  I do not like having my prowess limited
by stupid editors that are incapable of doing basic things like
displaying an "a" if I press the "a" key.  Thus I hate "vi" with a
passion.  microemacs is what I use at home, and emacs comes
close enough to emulating that.

> Or expecting an employee to know about basic OS stuff like setting
> the erase character in shell.. really isn't asking a terrible lot
> either.

Wrong delete key.  The "DEL" key is being interpreted by
emacs as erase-backward-character instead of erasing the
current character.  At least on one of my emacs.  On the other
emacs, the .emacs file causes emacs to behave as expected.

> Both are very basic tools of the trade.  If you can't or won't learn
> them you probably need a different trade.

I know enough of vi to hold my nose for long enough to add
emacs to my .profile and switch to an editor that displays an
"a" when I press the "a" key.

> Not to be snotty but your
> post comes off like a cry baby who can't do his job.

:-)  If you have any concerns about my ability to do my job,
extremely productively, it is only because you've never been
my employer.

> A couple of evenings spent with a vi tutorial will get you lined out.
> Then start worrying about customizing emacs or changing defaults.

Do you actually touch-type?  Ever seen a skilled programmer
who can touch-type at 60wpm in action?  The keystrokes are
very important, they are memorized, they are automatic.  I
expect ctrl-e to erase the contents of the line, not the entire
line, and thus I automatically expect two ctrl-es to be required
to erase the line.  I need to concentrate on the work I am doing,
the fingers do the keystrokes by themselves.  Anything else
damages my productivity.  I sell my productivity.

> Many basic unix books have vi sections.

So does the man page.  I already know it sux.  I use emacs
because it is widely-installed.  I don't actually care what the
keystrokes are, so long as I only have to remember one.  Just
the same as I don't care if it is a QWERTY or (can't remember
alternative) keyboard, so long as I only have to know one.

The choice of keyboard was made a long time ago.  The choice
of a flavour of emacs was made circa 1985, on someone else's
recommendation (widely-available, multi-platform).  I just can't
believe the basics keep changing on me, by default.  I've struggled
with emacs "bugs"/"design flaws" for probably a decade, there's
always something that pops up to interfere with the basics, e.g.
having to press ctrl-d instead of DEL.  It breaks the rhythm and
really sux.  Doesn't suck as much as vi though.

BFN.  Paul.

Re: simple editor required

[Stein A. Stromme]

[Paul Edwards]

| expect ctrl-e to erase the contents of the line, not the entire
| line, and thus I automatically expect two ctrl-es to be required
| to erase the line. 

(setq kill-whole-line nil) 

-- 
Stein Arild Strømme            +47 55584825, +47 95801887
Universitetet i Bergen                  Fax: +47 55589672     
Matematisk institutt                www.mi.uib.no/stromme         
Johs Brunsg 12, N-5008 BERGEN           stromme@mi.uib.no

Re: simple editor required

[Paul Edwards]

"Eli Zaretskii" <eliz@elta.co.il> wrote in message news:mailman.7449.1054900407.21513.help-gnu-emacs@gnu.org...
> > From: "Paul Edwards" <kerravon@nosppaam.w3.to>
> > Newsgroups: gnu.emacs.help
> > Date: Fri, 06 Jun 2003 06:15:19 GMT
> >
> > The thing is, the code that I am changing is not just one style,
> > I have to match the code that I am editting.  So I would rather
> > just have very basic editting working by default.  ie I don't
> > mind have to press space 4 times after typing a "{", but I do
> > mind having to type 20 spaces after a "{".
>
> Does it help to type "M-x fundamental-mode RET" befiore editing the
> code?  Does that do what you want?

Fundamental mode unfortunately doesn't stay at the right indent
level.

e.g. If I go

aaaa
  bbb
  ccc
dddd

I expect that after hitting "enter" after typing "bbb" that I can
immediately start typing "ccc".  On older versions of emacs,
I used to have "enter" defined as "newline-and-indent" and it
was all working fine.  But on the newer versions I have been
using, that functionality has been changed.

I looked at the help in emacs and it said that it is (now?) dependent
on the mode, so I tried switching from fundamental to text mode,
which did make that work, but introduced different problems (in
inserting lines full of blanks, and in addition, ctrl-e deletes not
just the blanks, but the entire line, so I have to instead backspace
over the blanks).

And even setting the default mode to text only worked on one
of the emacs versions, the other one wouldn't accept my default.

I've still got several things to try out from this thread though,
next time I have an opportunity at work.

BFN.  Paul.

Re: simple editor required

[Eli Zaretskii]

> From: "Paul Edwards" <kerravon@nosppaam.w3.to>
> Newsgroups: gnu.emacs.help
> Date: Sat, 07 Jun 2003 03:49:33 GMT
> 
> Fundamental mode unfortunately doesn't stay at the right indent
> level.
> 
> e.g. If I go
> 
> aaaa
>   bbb
>   ccc
> dddd
> 
> I expect that after hitting "enter" after typing "bbb" that I can
> immediately start typing "ccc".  On older versions of emacs,
> I used to have "enter" defined as "newline-and-indent" and it
> was all working fine.  But on the newer versions I have been
> using, that functionality has been changed.

Then use indented-text-mode instead of fundamental-mode.  You will
still need to type a TAB every now and then to align the beginning of
a new line, but in most cases it should be fine.

(I suggested fundamental-mode because I thought you wanted to control
the indentation yourself, but now it sounds like you do want some of
the Emacs features I thought were getting in your way.  "Simple
editors" don't have auto-indentation, btw.)

> I looked at the help in emacs and it said that it is (now?) dependent
> on the mode, so I tried switching from fundamental to text mode,
> which did make that work, but introduced different problems (in
> inserting lines full of blanks

What problems?  Details, please.

> And even setting the default mode to text only worked on one
> of the emacs versions, the other one wouldn't accept my default.

Try indented-text-mode instead.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> The thing is, the code that I am changing is not just one style,
> I have to match the code that I am editting.  So I would rather
> just have very basic editting working by default.  ie I don't
> mind have to press space 4 times after typing a "{", but I do
> mind having to type 20 spaces after a "{".

Okay, okay.

Please use C-h v c-syntactic-indentation RET to find out if the
variable is defined.  If it is, you can do the following:

(setq c-syntactic-indentation nil)

If that doesn't do the trick, then maybe the following works:

(defun pe-c-setup ()
  (setq c-syntactic-indentation nil))
(add-hook 'c-mode-hook 'pe-c-setup)

If the variable doesn't exist, then you need to do something more
elaborate:

(defun pe-c-setup ()
  (local-set-key (kbd "TAB") 'indent-relative)
  (local-set-key (kbd "{") 'self-insert-command))
(add-hook 'c-mode-hook 'pe-c-setup)

Maybe you have to add more local-set-key statements, similar to the
second one, for other characters that don't just insert themselves.

Also, the above was for C mode, use add-hook lines with c++-mode-hook
or java-mode-hook instead of c-mode-hook to get the effect for the
corresponding modes.


I don't understand your comment out the delete key, though.
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

"Eli Zaretskii" <eliz@elta.co.il> wrote in message news:mailman.7519.1054976930.21513.help-gnu-emacs@gnu.org...
> > From: "Paul Edwards" <kerravon@nosppaam.w3.to>
> > Newsgroups: gnu.emacs.help
> > Date: Sat, 07 Jun 2003 03:49:33 GMT
> >
> > Fundamental mode unfortunately doesn't stay at the right indent
> > level.
> >
> > e.g. If I go
> >
> > aaaa
> >   bbb
> >   ccc
> > dddd
> >
> > I expect that after hitting "enter" after typing "bbb" that I can
> > immediately start typing "ccc".  On older versions of emacs,
> > I used to have "enter" defined as "newline-and-indent" and it
> > was all working fine.  But on the newer versions I have been
> > using, that functionality has been changed.
>
> Then use indented-text-mode instead of fundamental-mode.  You will
> still need to type a TAB every now and then to align the beginning of
> a new line, but in most cases it should be fine.

Yeah, I don't mind that.  I just need the continuation from one line
with text to the next.

> (I suggested fundamental-mode because I thought you wanted to control
> the indentation yourself, but now it sounds like you do want some of

This is not language-dependent indentation, this is basic text
alignment.

> the Emacs features I thought were getting in your way.  "Simple
> editors" don't have auto-indentation, btw.)

I don't know that that is true.  An editor that doesn't allow
you to stay at a particular level is not an editor, it is not
suitable for programming in any language.  Alignment of text
under text is very very basic.  I don't mind if the behaviour
changes when I break the pattern, but I do mind if the
behaviour changes line by line.  No programmer has to hit
space/tab 20 times at the beginning of every line in order to
line up code.

> > I looked at the help in emacs and it said that it is (now?) dependent
> > on the mode, so I tried switching from fundamental to text mode,
> > which did make that work, but introduced different problems (in
> > inserting lines full of blanks
>
> What problems?  Details, please.

Well the first problem is that only one of the emacses responded
to my .emacs requesting the default mode to be text mode.  The
other remained as fundamental.  I would be far more comfortable
if either both or neither worked.  At the moment I have to conclude
that fundamental behaviour is being changed *by default* between
versions.  I don't care if 1000 new optional things are added
between releases, so long as they aren't activated by default.

The second problem I encountered was that in the version where
text mode actually took effect, was that it inserted spaces instead
of leaving the line empty, when I inserted a blank line.  I haven't
seen emacs do that before, even when it used to work on an older
version.  But this is perhaps a case of faulty memory.

The third problem I encountered was when trying to manually
clean up the blanks, ctrl-k would not simply delete the blanks on
the line, it would delete the entire line.

Someone has suggested putting:
(setq kill-whole-line nil)

to get around this problem.  I didn't have to do that in other
versions of emacs, it was the basic standard consistent
behaviour.

> > And even setting the default mode to text only worked on one
> > of the emacs versions, the other one wouldn't accept my default.
>
> Try indented-text-mode instead.

Ok.  Thanks.  I have several things to try.  With a bit of luck that
mode will just work straight away.  In summary, what I am looking
for is:

1. text to be lined up, line by line, suitable for any programming
language, no matter what file extension I am using.  IIRC, all I needed
to do on old emacs versions was inhibit the site file, either via a
command-line option or .emacs file, can't remember, either is
acceptable, and the junk wouldn't get loaded.

2. Block-marking not to be destroyed by ctrl-f and esc-g.  Xemacs
20.4 has a you-beaut highlighting of the block, which is great, but
I'd trade it any day for the ability to do a find/goto to position to
end of block.  Like emacs 20.7.1.  And like the last time I used a
version of xemacs, IIRC.  I don't know why some people think it
is appropriate to forget the marker, seems pretty basic to me, but
besides, other emacs (like 20.7.1) don't forget the marker.

3. DEL to delete.  Just like xemacs 20.4 with appropriate .emacs
entry.

I consider these to be extremely basic things that should be provided
out-of-the-box with all emacs versions without having to create an
enormous .emacs file, tailored for every version of emacs, to suppress
every quirk that emacs introduces *by default* in every incarnation.
I am shocked that the basics don't remain consistent.

BFN.  Paul.

Re: simple editor required

[Eli Zaretskii]

> From: "Paul Edwards" <kerravon@nosppaam.w3.to>
> Newsgroups: gnu.emacs.help
> Date: Sat, 07 Jun 2003 10:19:34 GMT
> 
> Well the first problem is that only one of the emacses responded
> to my .emacs requesting the default mode to be text mode.  The
> other remained as fundamental.

Please show the fragment of your .emacs that didn't work in one of
these versions.  Also, please tell what Emacs versions are you using.

> At the moment I have to conclude that fundamental behaviour is being
> changed *by default* between versions.

Some defaults indeed change whenever the maintainers find the old
defaults to be contrary to expectations of many users.  However, these
changes are rare; most defaults stay put for many releases.

> The second problem I encountered was that in the version where
> text mode actually took effect, was that it inserted spaces instead
> of leaving the line empty, when I inserted a blank line.

How did you insert a blank line?  What command did you use?

> The third problem I encountered was when trying to manually
> clean up the blanks, ctrl-k would not simply delete the blanks on
> the line, it would delete the entire line.

This is a feature: Emacs removes an empty line together with its
newline terminator.  Use M-\ to remove the blanks alone.

Re: simple editor required

[lawrence mitchell]

Paul Edwards wrote:

[...]

> I don't know that that is true.  An editor that doesn't allow
> you to stay at a particular level is not an editor, it is not
> suitable for programming in any language.
> Alignment of text under text is very very basic.  I don't mind
> if the behaviour changes when I break the pattern, but I do
> mind if the behaviour changes line by line.  No programmer has
> to hit space/tab 20 times at the beginning of every line in
> order to line up code.

I still don't quite understand this.  When I use Emacs, in, say,
C Mode, typing normally lines things up.  Pressing return adds a
newline and indents the current line, and the only thing that
changes indentation is if I forget a close ";", or add an open
or close brace ("{" or "}").  But perhaps I misunderstand your
needs.

[...] Summary of wants.

> 1. text to be lined up, line by line, suitable for any programming
> language, no matter what file extension I am using.  IIRC, all I needed
> to do on old emacs versions was inhibit the site file, either via a
> command-line option or .emacs file, can't remember, either is
> acceptable, and the junk wouldn't get loaded.

> 2. Block-marking not to be destroyed by ctrl-f and esc-g.  Xemacs
> 20.4 has a you-beaut highlighting of the block, which is great, but
> I'd trade it any day for the ability to do a find/goto to position to
> end of block.  Like emacs 20.7.1.  And like the last time I used a
> version of xemacs, IIRC.  I don't know why some people think it
> is appropriate to forget the marker, seems pretty basic to me, but
> besides, other emacs (like 20.7.1) don't forget the marker.

So, you don't want transient-mark-mode?  I don't believe that it
is turned on by default.  Certainly, without any kind of
customisations (emacs -q --no-site-file), when setting the mark,
and then moving away, the mark is not lost.  However, in this
mode, you cannot see what part of the text is marked.

Note that if you use transient-mark-mode (or zmacs-regions in
XEmacs), you can swap point and mark around by doing C-x C-x.


> 3. DEL to delete.  Just like xemacs 20.4 with appropriate .emacs
> entry.

If I understand your needs correctly, this .emacs is probably
close to what you want:

;; turn off all major-mode associations
(setq auto-mode-alist nil)
(setq interpreter-mode-alist nil)

;; Make return indent to the current level
(global-set-key (kbd "RET") #'newline-and-indent)
;; Make tab insert a tab, always
;; If you want tabs to be converted into spaces, you
;; will need to untabify files on saving.  But many people don't
;; like this.
(global-set-key (kbd "TAB") #'self-insert-command)

;; Make the DEL key work the same as C-d (and not like backspace)
(global-set-key (kbd "<delete>") #'delete-char)

;;; end

[...]

> Someone has suggested putting:
> (setq kill-whole-line nil)

> to get around this problem.  I didn't have to do that in other
> versions of emacs, it was the basic standard consistent
> behaviour.

kill-whole-line is nil by default, and has been since it was
first introduced in Emacs in 1993.  So, if it isn't for you, the
site file must be changing it.
The command line option to inhibit site-file loading is
"--no-site-file".

[...]

> I consider these to be extremely basic things that should be provided
> out-of-the-box with all emacs versions without having to create an
> enormous .emacs file, tailored for every version of emacs, to suppress
> every quirk that emacs introduces *by default* in every incarnation.
> I am shocked that the basics don't remain consistent.

Is there any particular need that you need a new Emacs version.
If, like you say, Emacs 20.7 works fine for you, and you have no
need of the new features in Emacs 21, why not stick with Emacs
20?  Or, use micro-emacs everywhere, after all, it's not an
overly large install.

-- 
lawrence mitchell <wence@gmx.li>

Re: simple editor required

[Paul Edwards]

"lawrence mitchell" <wence@gmx.li> wrote in message news:?fnord?87k7by3zjj.fsf@ID-97657.usr.dfncis.de...
> > I don't know that that is true.  An editor that doesn't allow
> > you to stay at a particular level is not an editor, it is not
> > suitable for programming in any language.
> > Alignment of text under text is very very basic.  I don't mind
> > if the behaviour changes when I break the pattern, but I do
> > mind if the behaviour changes line by line.  No programmer has
> > to hit space/tab 20 times at the beginning of every line in
> > order to line up code.
>
> I still don't quite understand this.  When I use Emacs, in, say,
> C Mode, typing normally lines things up.  Pressing return adds a
> newline and indents the current line, and the only thing that
> changes indentation is if I forget a close ";", or add an open
> or close brace ("{" or "}").  But perhaps I misunderstand your
> needs.

Yeah, but I don't require it to be that sophisticated, I just need
it to do indentation even of text.  No matter what file I am
editting, I expect

aaaa
  bbb
  ccc

the "ccc" to automatically start under the first "b".  All programming
editors are able to do this, its a very basic thing.  I'll worry about
fancy stuff at a later date.

I'm sure emacs can do it, I was just saying that it is a basic
function in any editor.

> [...] Summary of wants.
>
> > 1. text to be lined up, line by line, suitable for any programming
> > language, no matter what file extension I am using.  IIRC, all I needed
> > to do on old emacs versions was inhibit the site file, either via a
> > command-line option or .emacs file, can't remember, either is
> > acceptable, and the junk wouldn't get loaded.
>
> > 2. Block-marking not to be destroyed by ctrl-f and esc-g.  Xemacs
> > 20.4 has a you-beaut highlighting of the block, which is great, but
> > I'd trade it any day for the ability to do a find/goto to position to
> > end of block.  Like emacs 20.7.1.  And like the last time I used a
> > version of xemacs, IIRC.  I don't know why some people think it
> > is appropriate to forget the marker, seems pretty basic to me, but
> > besides, other emacs (like 20.7.1) don't forget the marker.
>
> So, you don't want transient-mark-mode?  I don't believe that it
> is turned on by default.  Certainly, without any kind of
> customisations (emacs -q --no-site-file), when setting the mark,
> and then moving away, the mark is not lost.  However, in this
> mode, you cannot see what part of the text is marked.

Ok, I'd better retest and answer this again, I'll stick to one thing
at a time.  See if I can get each feature working consistently
between versions, one at a time.  :-)

> > 3. DEL to delete.  Just like xemacs 20.4 with appropriate .emacs
> > entry.
>
> If I understand your needs correctly, this .emacs is probably
> close to what you want:
>
> ;; turn off all major-mode associations
> (setq auto-mode-alist nil)
> (setq interpreter-mode-alist nil)

Sounds great.

> ;; Make return indent to the current level
> (global-set-key (kbd "RET") #'newline-and-indent)

I believe that in fundamental mode, that doesn't work, but
let me try again.

> ;; Make tab insert a tab, always
> ;; If you want tabs to be converted into spaces, you
> ;; will need to untabify files on saving.  But many people don't
> ;; like this.
> (global-set-key (kbd "TAB") #'self-insert-command)

Actually I want spaces inserted, not tabs, but didn't mention that
because that feature is already working.

> ;; Make the DEL key work the same as C-d (and not like backspace)
> (global-set-key (kbd "<delete>") #'delete-char)

Yeah, I believe this doesn't work on emacs 20.7.1, but let me retest.

> > Someone has suggested putting:
> > (setq kill-whole-line nil)
>
> > to get around this problem.  I didn't have to do that in other
> > versions of emacs, it was the basic standard consistent
> > behaviour.
>
> kill-whole-line is nil by default, and has been since it was
> first introduced in Emacs in 1993.

That's what I thought.

> So, if it isn't for you, the
> site file must be changing it.
> The command line option to inhibit site-file loading is
> "--no-site-file".

Ok, I'll retest.  I only noticed it when I went into text mode in
xemacs.

> > I consider these to be extremely basic things that should be provided
> > out-of-the-box with all emacs versions without having to create an
> > enormous .emacs file, tailored for every version of emacs, to suppress
> > every quirk that emacs introduces *by default* in every incarnation.
> > I am shocked that the basics don't remain consistent.
>
> Is there any particular need that you need a new Emacs version.

No, I just need to use whatever the company has installed.  Every
place I've been to always has *a* version of Emacs installed.

> If, like you say, Emacs 20.7 works fine for you, and you have no

It doesn't work fine.  DEL doesn't work.

> need of the new features in Emacs 21, why not stick with Emacs

I don't have 21 available.

> 20?  Or, use micro-emacs everywhere, after all, it's not an
> overly large install.

I can't tell the company what to install.  I'm grateful that I have
an alternative to vi.  I'm not grateful that emacs (seems to)
change fundamental behaviour between releases.

Anyway, let me retest these things one at a time, before I end
up chasing my own tail around.  :-)

Thanks.  Paul.

Re: simple editor required

[Paul Edwards]

"lawrence mitchell" <wence@gmx.li> wrote in message news:?fnord?87k7by3zjj.fsf@ID-97657.usr.dfncis.de...
> If I understand your needs correctly, this .emacs is probably
> close to what you want:
>
> ;; turn off all major-mode associations
> (setq auto-mode-alist nil)
> (setq interpreter-mode-alist nil)

This worked great, I now get text mode for everything, which is
close to what I want.  Thanks!

> ;; Make the DEL key work the same as C-d (and not like backspace)
> (global-set-key (kbd "<delete>") #'delete-char)

Unfortunately this command doesn't work.  I get the same old
same old...

DEL runs the command backward-delete-char-untabify

But in xemacs it works.

I'll settle for something like this, except how do I put DEL?

(global-set-key "\e[4~" 'delete-char)

>>>to get the delete key to work as you'd like. (\e[4~ would need to be
>>>subsituted by the escape sequence of your delete key).  To find your
>>>delete keys escape sequence, go into the *scratch* buffer, hit C-q, then
>>>press your delete key.

DEL.

>>>   (define-key function-key-map [delete] [?\C-d])

This didn't do anything either.

Any more suggestions?  I compared xemacs and emacs again,
and emacs has less things causing me trouble, but xemacs has
the DEL key working, and that is an important keystroke, the
other stuff isn't, hence the reason I originally switched to xemacs
when I found it installed.

Thanks.  Paul.

P.S. I tried starting emacs with:
emacs --no-windows --no-site-file .emacs
but that didn't change anything.

P.P.S. Here is my latest .emacs.

; Switch off any crap the installation has
(setq inhibit-default-init 't)

; Don't let emacs guess what file types I am editting
(setq auto-mode-alist '() interpreter-mode-alist '())

; Always assume text mode
(setq-default default-major-mode 'text-mode)

; Make backspace backspace
(global-set-key [?\C-h] 'delete-backward-char)

; Set default tab width
(setq-default tab-width 4)

; Don't create junk backup files
(setq make-backup-files nil)
(setq auto-save-default nil)

; make insert toggle overwrite mode
(global-set-key [insertchar] 'overwritemode)

; Don't use tabs to indent
(setq-default indent-tabs-mode nil)

; Indent whenever you hit enter
(global-set-key [?\C-m] 'newline-and-indent)

; micro-emacs lookalikes
(global-set-key "\M-g" 'goto-line)
(global-set-key "\M-r" 'replace-string)

(global-unset-key [delete])
(global-set-key (kbd "<delete>") #'delete-char)
;(global-set-key [delete] 'delete-char)

; potential alternatives
;(global-unset-key [deletechar])
;(global-set-key [deletechar] 'delete-char)
;(global-unset-key (kbd "<delete>"))
;(global-set-key (kbd "<delete>") 'delete-char)

Re: simple editor required

[Stefan Monnier]

>> (global-set-key (kbd "<delete>") #'delete-char)
> Unfortunately this command doesn't work.  I get the same old

Am I talking to a wall or what ?
Have you tried (define-key function-key-map [delete] "\C-d") ?


        Stefan

Re: simple editor required

[Paul Edwards]

"Stefan Monnier" <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> wrote in message news:5l7k7uc8x1.fsf@rum.cs.yale.edu...
> >> (global-set-key (kbd "<delete>") #'delete-char)
> > Unfortunately this command doesn't work.  I get the same old
>
> Am I talking to a wall or what ?
> Have you tried (define-key function-key-map [delete] "\C-d") ?

In the message you responded to I also showed that this didn't
work (or a slight variation, anyway).  I actually tried both
that and the slight variation yesterday, neither worked.  But just
to be sure, I tried the above again today, and still no effect.

I know that I am getting a DEL go through to emacs, because
emacs tells me so.  I know that emacs is binding DEL to
delete-backward-char, because emacs tells me so.

What I can't do is coax emacs into shifting the binding of DEL
from delete-backward-char to delete-char.

And I know the other commands that have been recommended
have been typed in correctly, because xemacs accepts them
and behaves as expected.  Unfortunately xemacs has different
problems, which means it isn't so simple to just use xemacs,
and I would ideally like the emacs I use to work as expected.

Actually, I have an idea.  Maybe I can hack the executable and
change the default from delete-backward-char to delete-char?

BFN.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> Unfortunately this command doesn't work.  I get the same old
> same old...
>
> DEL runs the command backward-delete-char-untabify

Please type x, then hit the delete key, then C-h l.  Then you can see
what Emacs sees when you hit the delete key.  Suppose you see 

    ... x frumple gargle C-h l

in the lossage buffer.  Then do

(global-set-key (kbd "<frumple> <gargle>") 'delete-char)

Alas, it is not obvious when do you have to add the angle brackets
and when you should not add them.  As a rule of thumb, if it looks
like a normal character, possibly with modifiers, then do NOT use
<...>.  When it looks like a function key, DO use <...>.
Additionally, there are some special keys, that look like function
keys, but aren't.  You can recognize them from the upper case.  So
RET, TAB, SPC, LFD, DEL are *not* function keys whereas prior, next,
f3, delete, tab *are* function keys.

When you see C-f3, put C-<f3> into the string (<C-f3> will also
work), including the <...>.

-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 746 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:8465nduadj.fsf@lucy.is.informatik.uni-duisburg.de...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > Unfortunately this command doesn't work.  I get the same old
> > same old...
> >
> > DEL runs the command backward-delete-char-untabify
>
> Please type x, then hit the delete key, then C-h l.  Then you can see
> what Emacs sees when you hit the delete key.  Suppose you see
>
>     ... x frumple gargle C-h l
>
> in the lossage buffer.  Then do
>
> (global-set-key (kbd "<frumple> <gargle>") 'delete-char)

ESC O B ESC O B ESC O B ESC O B ESC O B ESC O B ESC
O B ESC O B ESC O B ESC O B ESC O B ESC O B ESC O B
x DEL C-h l

No funny characters.  Just a DEL.

BFN.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> ESC O B ESC O B ESC O B ESC O B ESC O B ESC O B ESC
> O B ESC O B ESC O B ESC O B ESC O B ESC O B ESC O B
> x DEL C-h l
>
> No funny characters.  Just a DEL.

So (global-set-key (kbd "DEL") 'delete-char) is not working?  Or,
like Stefan suggests, (global-set-key (kbd "DEL") (kbd "C-d"))?

(If both work, use the second one.)
-- 
This line is not blank.

Re: simple editor required

[Stefan Monnier]

>>>>> "Paul" == Paul Edwards <kerravon@nosppaam.w3.to> writes:
> ESC O B ESC O B ESC O B ESC O B ESC O B ESC O B ESC
> O B ESC O B ESC O B ESC O B ESC O B ESC O B ESC O B
> x DEL C-h l

Oh, so you're on a text-terminal.  As I said, my code wouldn't work
in such a case.

   (global-set-key "?\C-?" 'delete-char)

might do what you want, but it might also change the behavior of backspace
because backspace might send the same key sequence to Emacs.  Or maybe
backspace sends C-h, which then clashes with Emacs's help key.


        Stefan

Re: simple editor required

[Paul Edwards]

"Stefan Monnier" <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> wrote in message news:5lfzmgbrio.fsf@rum.cs.yale.edu...
> >>>>> "Paul" == Paul Edwards <kerravon@nosppaam.w3.to> writes:
> > ESC O B ESC O B ESC O B ESC O B ESC O B ESC O B ESC
> > O B ESC O B ESC O B ESC O B ESC O B ESC O B ESC O B
> > x DEL C-h l
>
> Oh, so you're on a text-terminal.  As I said, my code wouldn't work
> in such a case.
>
>    (global-set-key "?\C-?" 'delete-char)
>
> might do what you want, but it might also change the behavior of backspace
> because backspace might send the same key sequence to Emacs.

Backspace and delete are two different keys on my terminal.

> Or maybe
> backspace sends C-h, which then clashes with Emacs's help key.

Yeah, but that is easily remapped, which is what I do.  This basic
delete vs backspace I have working in xemacs already, so I know
it can be done via the editor.  But before you say "use xemacs then",
xemacs has different problems, completely unrelated to key
mappings.

Anyway, here are the results of today's tests:


(global-set-key "?\C-?" 'delete-char)

Error in init file: error: "Key sequence ? DEL uses invalid prefix
characters"

In xemacs has no effect



(global-set-key "\C-?" 'delete-char)

works in xemacs, has no effect in emacs



(global-set-key [DEL] 'delete-char) has no effect



The following work in xemacs, don't work in emacs...

(global-set-key (kbd "DEL") 'delete-char)
(global-set-key (kbd "DEL") (kbd "C-d"))

BFN.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> The following work in xemacs, don't work in emacs...
>
> (global-set-key (kbd "DEL") 'delete-char)
> (global-set-key (kbd "DEL") (kbd "C-d"))

What does it mean to not work?  Does Emacs crash?
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 800 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:84el1zcze6.fsf@lucy.is.informatik.uni-duisburg.de...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > The following work in xemacs, don't work in emacs...
> >
> > (global-set-key (kbd "DEL") 'delete-char)
> > (global-set-key (kbd "DEL") (kbd "C-d"))
>
> What does it mean to not work?  Does Emacs crash?

No, it just has no effect.  The DEL key remains bound to
delete-backward-char and keeps trying to delete the previous
character.

No "error in init file", no crash, just completely ignored.

And last time I tried binding DEL to a different function,
e.g. find, it ignored that too, it isn't just the delete-char
function it doesn't like.  But I can retry that to be sure if
you think it would be worth a try?

BFN.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> "Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:84el1zcze6.fsf@lucy.is.informatik.uni-duisburg.de...
>> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>>
>> > The following work in xemacs, don't work in emacs...
>> >
>> > (global-set-key (kbd "DEL") 'delete-char)
>> > (global-set-key (kbd "DEL") (kbd "C-d"))
>>
>> What does it mean to not work?  Does Emacs crash?
>
> No, it just has no effect.  The DEL key remains bound to
> delete-backward-char and keeps trying to delete the previous
> character.

Gah?  How can this be?  I mean, C-h l tells that the key in question
is the DEL key.  Hm.

Is the statement really executed?  Try putting (message "Hi there")
(sit-for 3) before and after the statement to see if Emacs ever gets
there.  Also try M-: to eval the statement from a running Emacs, to
see if that has any effect.

Hm.  Oh!  It could be that the key is not bound in the global map,
but instead in some local map, instead.  Try C-x b foo RET, which
creates a buffer named foo in fundamental mode.  Maybe the DEL key
works in that mode?

I think that emacs-lisp-mode or lisp-interaction-mode have their own
bindings for DEL, so you'd have to do like

(define-key emacs-lisp-mode-map (kbd "DEL") 'delete-char)

for those.

-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1993 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:841xxzto54.fsf@lucy.is.informatik.uni-duisburg.de...
> >> > (global-set-key (kbd "DEL") 'delete-char)
> >> > (global-set-key (kbd "DEL") (kbd "C-d"))
> >>
> >> What does it mean to not work?  Does Emacs crash?
> >
> > No, it just has no effect.  The DEL key remains bound to
> > delete-backward-char and keeps trying to delete the previous
> > character.
>
> Gah?  How can this be?  I mean, C-h l tells that the key in question
> is the DEL key.  Hm.
>
> Is the statement really executed?  Try putting (message "Hi there")
> (sit-for 3) before and after the statement to see if Emacs ever gets
> there.

Yes it does.

> Also try M-: to eval the statement from a running Emacs, to
> see if that has any effect.

just says "delete-char" doesn't do anything

> Hm.  Oh!  It could be that the key is not bound in the global map,
> but instead in some local map, instead.  Try C-x b foo RET, which
> creates a buffer named foo in fundamental mode.  Maybe the DEL key
> works in that mode?

created buffer, but no effect.

from another post...
>>>>(add-hook 'c-common-mode-hook
>>>>  (lambda () (local-set-key "\C-?" 'delete-char)))

had no effect when editting .c file.

> I think that emacs-lisp-mode or lisp-interaction-mode have their own
> bindings for DEL, so you'd have to do like
>
> (define-key emacs-lisp-mode-map (kbd "DEL") 'delete-char)

AND THE WINNER IS!!!  :-)

Thanks a lot for figuring that out Kai.

And since I now use text mode for everything, all I need is...
(define-key text-mode-map (kbd "DEL") 'delete-char)

and it works!

For technical curiosity though, why was everyone convinced
the global-set-key would work?  And indeed it did work on
Xemacs.  I would have thought this situation would have
come up all the time for the presumably large number of
people who want that behaviour of DEL doing a delete-char.

I'll now go back and check the relatively minor problems I
have within text mode.

BFN.  Paul.

Re: simple editor required

[Sam Halliday]

Paul Edwards wrote:
> For technical curiosity though, why was everyone convinced
> the global-set-key would work?  And indeed it did work on
> Xemacs.  I would have thought this situation would have
> come up all the time for the presumably large number of
> people who want that behaviour of DEL doing a delete-char.

did anyone suggest

  (normal-erase-is-backspace-mode 0)

in ~/.emacs

if that doesnt do what you want, then replace the 0 with a 1. i know you
have it working, this is just out of curiosity.

this was something i had issues with when i started using emacs, but i
wanted the opposite to what you wanted. when i read the docs, it
basically said all DEL/BACKSPACE issues could be solved with that call.
but that may be more to do with emacs picking up key mappings
incorrectly.

reference:
http://www.gnu.org/manual/emacs/html_node/emacs_512.html

please excuse me if this post is redundant, but i guess everyone is a
little confused as to why things are a little bit borked on Paul's
machine (from the shear length of the thread for such a trivial issue)

Re: simple editor required

[Paul Edwards]

"Sam Halliday" <devnull@example.com> wrote in message news:20030616114712.73c4d27e.devnull@example.com...
> Paul Edwards wrote:
> > For technical curiosity though, why was everyone convinced
> > the global-set-key would work?  And indeed it did work on
> > Xemacs.  I would have thought this situation would have
> > come up all the time for the presumably large number of
> > people who want that behaviour of DEL doing a delete-char.
>
> did anyone suggest
>
>   (normal-erase-is-backspace-mode 0)

No.

> in ~/.emacs

Error in init file: Symbol's function definition is void:
normal-erase-is-backspace-mode

Client is using 20.7.1, maybe it is a new feature.

BFN.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> Client is using 20.7.1, maybe it is a new feature.

Yes, normal-erase-is-backspace-mode is new in Emacs 21.

-- 
This line is not blank.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> For technical curiosity though, why was everyone convinced
> the global-set-key would work?  And indeed it did work on
> Xemacs.  I would have thought this situation would have
> come up all the time for the presumably large number of
> people who want that behaviour of DEL doing a delete-char.

Normally, DEL does like backspace (!) in Emacs.  The delete key is
not the same as ascii 127 (DEL is another name for ascii 127).
Under X11, the delete key is known as itself, and on a tty the delete
key often sends an escape sequence such as ESC [ 3 ~ (on my system).

That's why people don't run into this problem.
-- 
This line is not blank.

Re: simple editor required

[Stefan Monnier]

>> (define-key emacs-lisp-mode-map (kbd "DEL") 'delete-char)
> AND THE WINNER IS!!!  :-)

Actually, I think the best answer for you is

   (define-key function-key-map "\C-?" "\C-d")

> Xemacs.  I would have thought this situation would have
> come up all the time for the presumably large number of
> people who want that behaviour of DEL doing a delete-char.

DEL is for backward deletion, so the number of people who want to
use it for forward deletion is rather small.  If your `delete' key
sends a DEL, then the best answer is generally to fix the thing
that makes it work that way (typically, the console or the text-terminal)
so that it sends some other key sequence.


        Stefan

Re: simple editor required

[Kai Großjohann]

"Stefan Monnier" <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> writes:

>>> (define-key emacs-lisp-mode-map (kbd "DEL") 'delete-char)
>> AND THE WINNER IS!!!  :-)
>
> Actually, I think the best answer for you is
>
>    (define-key function-key-map "\C-?" "\C-d")

<aol/>

(Though I'd say (kbd "DEL") instead of "\C-?", as Stefan knows 8-)
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 829 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:84n0git5vn.fsf@lucy.is.informatik.uni-duisburg.de...

Can I ask the next (minor) thing on my list of things I've
been living with for x years?

In text mode, if my file looks like:

aaaa
bbbb
cccc

Hitting enter at beginning of aaaa line creates a newline plus an indent of
8 spaces.  (my tab is only 4 spaces), but I didn't actually expect any
indentation at all here.

Hitting enter at beginning of cccc line creates a newline only (as wanted).

My workaround is to type "x" then hit enter, then go up and delete
the line with "x" on it.

What's the rationale and fix for that?

And one more.  If I cursor down in xemacs, xemacs doesn't
automatically start expanding the size of my file when I reach
the end, but emacs does.  How do I stop that?

Thanks.  Paul.

Re: simple editor required

[Stefan Monnier]

> In text mode, if my file looks like:

> aaaa
> bbbb
> cccc

> Hitting enter at beginning of aaaa line creates a newline plus an indent of
> 8 spaces.

I don't see this with any combination of Emacs-20.7, Emacs-21.3, text-mode,
indented-text-mode, paragraph-indent-text-mode, newline, newline-and-indent.
It's probably a bad interaction with your config.

> (my tab is only 4 spaces),

This is ambiguous since it doesn't say whether it's the displayed size of
a TAB character, or the number of spaces-equivalent inserted when you hit
M-i or maybe TAB (in which case it also depends on what TAB is bound to,
which depends on the major mode via the variable indent-line-function).

> but I didn't actually expect any indentation at all here.

Rightly so.

> And one more.  If I cursor down in xemacs, xemacs doesn't
> automatically start expanding the size of my file when I reach
> the end, but emacs does.  How do I stop that?

C-h k <cursor down> will give you the documentation of the function
that moves the cursor down.  It explains how to change this part
of the behavior.


        Stefan

Re: simple editor required

[Paul Edwards]

"Stefan Monnier" <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> wrote in message news:5lr85sbs0w.fsf@rum.cs.yale.edu...
> > In text mode, if my file looks like:
>
> > aaaa
> > bbbb
> > cccc
>
> > Hitting enter at beginning of aaaa line creates a newline plus an indent of
> > 8 spaces.

Note that there are no blank lines above aaa.  My goal is to insert one.
Also, I have seen this problem before, probably with 18.x.

> I don't see this with any combination of Emacs-20.7, Emacs-21.3, text-mode,
> indented-text-mode, paragraph-indent-text-mode, newline, newline-and-indent.
> It's probably a bad interaction with your config.
>
> > (my tab is only 4 spaces),
>
> This is ambiguous since it doesn't say whether it's the displayed size of
> a TAB character, or the number of spaces-equivalent inserted when you hit
> M-i or maybe TAB (in which case it also depends on what TAB is bound to,
> which depends on the major mode via the variable indent-line-function).

I don't generate tabs myself, so it is display size.

> > but I didn't actually expect any indentation at all here.
>
> Rightly so.

Here is my .emacs.  I will now try out the various commands!

BFN.  Paul.


; Switch off any crap the installation has
(setq inhibit-default-init 't)

; Don't let emacs guess what file types I am editting
(setq auto-mode-alist '() interpreter-mode-alist '())

; Always assume text mode
(setq-default default-major-mode 'text-mode)

; Make backspace backspace
(global-set-key [?\C-h] 'delete-backward-char)

; Set default tab width
(setq-default tab-width 4)

; Don't create junk backup files
(setq make-backup-files nil)
(setq auto-save-default nil)

; make insert toggle overwrite mode
(global-set-key [insertchar] 'overwritemode)

; Don't use tabs to indent
(setq-default indent-tabs-mode nil)

; Indent whenever you hit enter
(global-set-key [?\C-m] 'newline-and-indent)

; micro-emacs lookalikes
(global-set-key "\M-g" 'goto-line)
(global-set-key "\M-r" 'replace-string)

;(global-unset-key [delete])
;(global-set-key (kbd "<delete>") #'delete-char)
;(define-key emacs-lisp-mode-map (kbd "DEL") 'delete-char)
(define-key text-mode-map (kbd "DEL") 'delete-char)
;(normal-erase-is-backspace-mode 0)
;(define-key function-key-map "\C-?" "\C-d")
;(global-set-key [delete] 'delete-char)
;(define-key function-key-map [delete] "\C-d")

; potential alternatives
;(global-unset-key [deletechar])
;(global-set-key [deletechar] 'delete-char)
;(global-unset-key (kbd "<delete>"))
;(global-set-key (kbd "<delete>") 'delete-char)

Re: simple editor required

[Paul Edwards]

"Stefan Monnier" <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> wrote in message news:5lr85sbs0w.fsf@rum.cs.yale.edu...
> > In text mode, if my file looks like:
>
> > aaaa
> > bbbb
> > cccc
>
> > Hitting enter at beginning of aaaa line creates a newline plus an indent of
> > 8 spaces.
>
> I don't see this with any combination of Emacs-20.7, Emacs-21.3, text-mode,
> indented-text-mode, paragraph-indent-text-mode, newline, newline-and-indent.
> It's probably a bad interaction with your config.

I've used a simpler version of .emacs, and what I got was a
blank line, and then on the next line I got a single tab in front
of my "aaaa".  I confirmed that if either of the last two things
in my .emacs are absent, I do not get that behaviour, I simply
get a blank line.

; Switch off any crap the installation has
(setq inhibit-default-init 't)

; Don't let emacs guess what file types I am editting
(setq auto-mode-alist '() interpreter-mode-alist '())

; Always assume text mode
(setq-default default-major-mode 'text-mode)

; Indent whenever you hit enter
(global-set-key [?\C-m] 'newline-and-indent)

BFN.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> In text mode, if my file looks like:
>
> aaaa
> bbbb
> cccc
>
> Hitting enter at beginning of aaaa line creates a newline plus an indent of
> 8 spaces.  (my tab is only 4 spaces), but I didn't actually expect any
> indentation at all here.

Whee.  Which version of Emacs is this?  Emacs 20.7?  Hm.

What does C-h k RET say?
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 996 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:844r2opsuy.fsf@lucy.is.informatik.uni-duisburg.de...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > In text mode, if my file looks like:
> >
> > aaaa
> > bbbb
> > cccc
> >
> > Hitting enter at beginning of aaaa line creates a newline plus an indent of
> > 8 spaces.  (my tab is only 4 spaces), but I didn't actually expect any
> > indentation at all here.
>
> Whee.  Which version of Emacs is this?  Emacs 20.7?  Hm.

20.7.1

> What does C-h k RET say?

What does C-h k RET say?

RET runs the command newline-and-indent
   which is an interactive compiled Lisp function in `simple'.
(newline-and-indent)

Insert a newline, then indent according to major mode.
Indentation is done using the value of `indent-line-function'.
In programming language modes, this is the same as TAB.
In some text modes, where TAB inserts a tab, this command indents to the
column specified by the function `current-left-margin'.

BFN.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> "Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:844r2opsuy.fsf@lucy.is.informatik.uni-duisburg.de...
>
>> What does C-h k RET say?
>
> What does C-h k RET say?
>
> RET runs the command newline-and-indent
>    which is an interactive compiled Lisp function in `simple'.
> (newline-and-indent)
>
> Insert a newline, then indent according to major mode.
> Indentation is done using the value of `indent-line-function'.
> In programming language modes, this is the same as TAB.
> In some text modes, where TAB inserts a tab, this command indents to the
> column specified by the function `current-left-margin'.

As you can see, it refers to the variable indent-line-function.  Use
C-h v to find its value, then tell us.

What does BFN stand for?
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 965 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:84r85ry9ft.fsf@lucy.is.informatik.uni-duisburg.de...
> > What does C-h k RET say?
> >
> > RET runs the command newline-and-indent
> >    which is an interactive compiled Lisp function in `simple'.
> > (newline-and-indent)
> >
> > Insert a newline, then indent according to major mode.
> > Indentation is done using the value of `indent-line-function'.
> > In programming language modes, this is the same as TAB.
> > In some text modes, where TAB inserts a tab, this command indents to the
> > column specified by the function `current-left-margin'.
>
> As you can see, it refers to the variable indent-line-function.  Use
> C-h v to find its value, then tell us.

indent-line-function's value is
indent-relative-maybe
Local in buffer .emacs; global value is
indent-to-left-margin

Documentation:
Function to indent current line.

Defined in `indent'.

> What does BFN stand for?

Bye For Now.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> indent-line-function's value is
> indent-relative-maybe
> Local in buffer .emacs; global value is
> indent-to-left-margin

Okay, so indent-relative-maybe has the effect at the beginning of the
file.  Put point at the beginning of the buffer and invoke M-x
indent-relative-maybe RET, you'll see where the RET behavior comes
from.

It's now clear that you want a function to use as the value of
indent-line-function that works like indent-relative-maybe in some
cases but not in others.

What should happen in which cases?
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2001 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:84vfv2ulf7.fsf@lucy.is.informatik.uni-duisburg.de...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > indent-line-function's value is
> > indent-relative-maybe
> > Local in buffer .emacs; global value is
> > indent-to-left-margin
>
> Okay, so indent-relative-maybe has the effect at the beginning of the
> file.  Put point at the beginning of the buffer and invoke M-x
> indent-relative-maybe RET, you'll see where the RET behavior comes
> from.
>
> It's now clear that you want a function to use as the value of
> indent-line-function that works like indent-relative-maybe in some
> cases but not in others.
>
> What should happen in which cases?

Ok, the problem as I described is more generic than that.
After having used it for more I found that the following:

    aaaa
    bbbb

cccc

If I am at the beginning of the "cccc" line and I hit enter,
emacs is inspired to indent cccc to be under bbbb.  I don't
want that, I want cccc to stay where it is.  If I wanted cccc
to be indented I would do something, e.g. press tab or hit
spaces.

I expect that in the above, if I am at the end of aaaa and hit
enter, then a blank line is inserted and I am positioned under
the first "a".  This already happens.

As I continue to type and hit enter, I expect the cursor to
stay at that same indentiation level.  However, if I break
the sequence of non-blank lines, by leaving a blank line,
then I don't mind emacs giving up and deciding to leave me
at the left hand side until I've finished entering blank lines,
and then the obligation being on me to press spaces to start
the indentation sequence again.

If emacs was super smart and let me hit enter without
changing the indent level, so long as I didn't do anything
strange, such as a cursor movement, then that would be
great.  But if it doesn't have that feature, that's fine too.

Micro-emacs works as I described above, there are no
surprises that I am aware of.

BFN.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> If I am at the beginning of the "cccc" line and I hit enter,
> emacs is inspired to indent cccc to be under bbbb.  I don't
> want that, I want cccc to stay where it is.  If I wanted cccc
> to be indented I would do something, e.g. press tab or hit
> spaces.

Okay, so if the previous line exists and is not blank, then indent
the new line to the previous line after RET.

Yeah, I'm afraid you'll have to write your own indentation function.
Maybe this one?

(defun pe-indent-relative ()
  "Like `indent-relative' but don't indent after blank line."
  (when (save-excursion
          (and (zerop (forward-line -1))
               (not (looking-at "\s-*$"))))
    (indent-relative-maybe)))

(defun pe-text-setup ()
  (setq indent-line-function 'pe-indent-relative))

(add-hook 'text-mode-hook 'pe-text-setup)

Maybe it works, but I haven't tested it.
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2589 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:84wufgke19.fsf@lucy.is.informatik.uni-duisburg.de...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > If I am at the beginning of the "cccc" line and I hit enter,
> > emacs is inspired to indent cccc to be under bbbb.  I don't
> > want that, I want cccc to stay where it is.  If I wanted cccc
> > to be indented I would do something, e.g. press tab or hit
> > spaces.
>
> Okay, so if the previous line exists and is not blank, then indent
> the new line to the previous line after RET.

> Yeah, I'm afraid you'll have to write your own indentation function.

Wow.  How does everyone else cope with text mode?  This is
basic logical indentation, that comes with micro-emacs.  There's
no existing equivalent?

> Maybe this one?
>
> (defun pe-indent-relative ()
>   "Like `indent-relative' but don't indent after blank line."

Actually, I don't mind it indenting after a blank line, micro-emacs
does that.  But micro-emacs won't indent the CURRENT
line when I hit enter at the BEGINNING of the line.  Emacs not
only inserts a blank line (as I wanted), but is also inspired to
indent the current line for no reason whatsoever.  And when it
is at the beginning of the file, and thus doesn't even have any
previous line to match up with, it still forges ahead and decides
I must want to indent the CURRENT line (with text on it) by
8 characters or 1 tab, as well as putting in the blank line.

>   (when (save-excursion
>           (and (zerop (forward-line -1))
>                (not (looking-at "\s-*$"))))
>     (indent-relative-maybe)))
>
> (defun pe-text-setup ()
>   (setq indent-line-function 'pe-indent-relative))
>
> (add-hook 'text-mode-hook 'pe-text-setup)
>
> Maybe it works, but I haven't tested it.

Thanks!  I'll give it a go.

So what about when you are editting text, and it is lined up in
semi-paragraphs, e.g.

My doctor's details are:
  Name = Fred
  Phone = 1234


My dentist's details are:
  Name = Mary
  Phone = 5678

And the cursor happens to be at the beginning of "My dentist's details"
and you decide that you want an extra blank line to separate doctor
from dentist, so you hit enter.  You're not disturbed that emacs is
inspired to not only insert a blank line, but also to indent "My dentist"
in a couple of characters, to line up with "Phone = "?

I wouldn't mind if I was ABOVE "my dentist" and it decided to indent
a blank line a couple of characters in case I wanted to add more
details, but moving "My dentist" from its current indentation level is
definitely anti-social.

BFN.  Paul.

Re: simple editor required

[Stefan Monnier]

> Wow.  How does everyone else cope with text mode?  This is
> basic logical indentation, that comes with micro-emacs.  There's
> no existing equivalent?

I guess I'm wondering myself how can people use micro-emacs
without the basic logical indentation that comes with Emacs.


        Stefan

Re: simple editor required

[Paul Edwards]

"Stefan Monnier" <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> wrote in message news:5ly8zw7hah.fsf@rum.cs.yale.edu...
> > Wow.  How does everyone else cope with text mode?  This is
> > basic logical indentation, that comes with micro-emacs.  There's
> > no existing equivalent?
>
> I guess I'm wondering myself how can people use micro-emacs
> without the basic logical indentation that comes with Emacs.

There is nothing illogical about micro-emacs's behaviour.

BFN.  Paul.

Re: simple editor required

[Stefan Monnier]

> There is nothing illogical about micro-emacs's behaviour.

And there's nothing illogical about Emacs's behavior either,


        Stefan

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> "Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:848yrsgd4s.fsf@lucy.is.informatik.uni-duisburg.de...
>> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>>
>> > It is.  The "maybe" means it doesn't always indent.  And it
>> > shouldn't, when it is quite obvious the user already has text
>> > on that line, and doesn't need it indented.  If they had wanted
>> > it indented, they would have pressed tab, not enter.
>>
>> Huh?  I find it really convenient that I can just hit TAB (my binding
>> for indent-relative) at the beginning of a line to increase its
>> indentation.
>
> But that's what I just said.  You'd press TAB, not enter.
> TAB is logical, enter is not.  No-one expects enter, at
> the beginning of a line, to go to a newline and indent
> the existing text.  enter doesn't do that sort of thing.  It
> does start a newline, beginning under the previous one
> though.  That is logical and reasonable.

But the fact that RET does newline and then does like TAB is your own
customization!  So you can't blame Emacs that it is doing what you
told it to do.

That said, I understand that the behavior of RET that you see in
text-mode is not the best behavior.  Clearly it should be changed.

>> > Indentation is meant to apply to a NEW line, not an existing
>> > line.  It's meant to put blanks there ready for you to optionally
>> > start typing real characters, and then delete them later if they
>> > are not already there.
>>
>> I use it for existing lines, too.
>
> You don't hit enter at the beginning of a line of text, and expect
> it to insert a newline and indent the current text.

Actually, I've never minded until now.

But I see how it would be strange in text mode.  But in text mode,
indentation is not done very often.  I have told Emacs that RET
invokes newline-and-indent in programming modes (C, Java, Perl, ...)
whereas it does just newline in other modes, like text.

[time passes]

Oh, now I remember that I sent a bug report about newline-and-indent
in text mode, some years back.  Whee.  Yeah.  I guess Emacs just
doesn't like you to use newline-and-indent in text-like modes :-)

FWIW, I get by with just newline quite nicely.  For the kind of text
I write, indentation is not often required.

>> > It is not meant to trash existing lines!  Nor is it meant to get
>> > so confused on the first line, that it just indents 8 spaces, even
>> > when my tab setting is 4, and there's not even a prior line to
>> > inspire it to do something that silly.
>>
>> That might be a misconfiguration on your part.  (Not that I blame you;
>> the subject is complex.)  indent-relative looks at tab-stop-list and
>> configuring tab-width is not going to have the effect you might
>> expect.  (tab-width refers to how tab characters in the text are
>> displayed, but the TAB key does not always insert tab characters.)
>
> So it sometimes (ie at beginning of file) looks at tab-stop-list,
> and everywhere else it looks at the indentation of the previous
> line.  Seems very strange to me.

Hm.  I was going to answer that indent-relative does that, but in
fact indent-relative-maybe also has this strange behavior in the
first line of the file.  I think that might be a bug.

Note that indent-relative goes by tab-stop-list after the end of the
previous line, whereas indent-relative-maybe does not do that.
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 4910 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:84fzlzsdj6.fsf@lucy.is.informatik.uni-duisburg.de...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > "Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:848yrsgd4s.fsf@lucy.is.informatik.uni-duisburg.de...
> >> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
> >>
> >> > It is.  The "maybe" means it doesn't always indent.  And it
> >> > shouldn't, when it is quite obvious the user already has text
> >> > on that line, and doesn't need it indented.  If they had wanted
> >> > it indented, they would have pressed tab, not enter.
> >>
> >> Huh?  I find it really convenient that I can just hit TAB (my binding
> >> for indent-relative) at the beginning of a line to increase its
> >> indentation.
> >
> > But that's what I just said.  You'd press TAB, not enter.
> > TAB is logical, enter is not.  No-one expects enter, at
> > the beginning of a line, to go to a newline and indent
> > the existing text.  enter doesn't do that sort of thing.  It
> > does start a newline, beginning under the previous one
> > though.  That is logical and reasonable.
>
> But the fact that RET does newline and then does like TAB is your own
> customization!  So you can't blame Emacs that it is doing what you
> told it to do.

Well in that case I am blaming emacs for not coming with a
"logical indentation mode", e.g. (indent-text-properly 1)
which is preferably enabled by default.  I didn't specifically
want RET bound to newline-and-strange-indent, what I
wanted was logical indentation.

> That said, I understand that the behavior of RET that you see in
> text-mode is not the best behavior.  Clearly it should be changed.

Thanks for agreeing!  What's more surprising is that 10%
of the user base didn't report this 10 years ago, and thus
was fixed years before I they installed 20.7.1 on this box.

> >> > Indentation is meant to apply to a NEW line, not an existing
> >> > line.  It's meant to put blanks there ready for you to optionally
> >> > start typing real characters, and then delete them later if they
> >> > are not already there.
> >>
> >> I use it for existing lines, too.
> >
> > You don't hit enter at the beginning of a line of text, and expect
> > it to insert a newline and indent the current text.
>
> Actually, I've never minded until now.
>
> But I see how it would be strange in text mode.  But in text mode,
> indentation is not done very often.  I have told Emacs that RET
> invokes newline-and-indent in programming modes (C, Java, Perl, ...)
> whereas it does just newline in other modes, like text.

I am mostly using text mode to write C code.  I expect text
mode to allow me to hop in and start changing C code,
regardless of the existing style of the C code.

> [time passes]
>
> Oh, now I remember that I sent a bug report about newline-and-indent
> in text mode, some years back.  Whee.  Yeah.  I guess Emacs just
> doesn't like you to use newline-and-indent in text-like modes :-)

Is there any mode I can switch too, where I can get the behaviour
I want, without it making a whole set of new problems for me
(e.g. automatic assumptions about C coding style).  The only
assumption I want is that it is indented text, so if the previous
line started in column 16, then the cursor should be positioned
there when I start a new line.

> FWIW, I get by with just newline quite nicely.  For the kind of text
> I write, indentation is not often required.

Remember that doctor's and dentist's appointment?  I really
do have that stuff, in a big file called "notes.txt" and that is
how I indent too.

But more importantly I need it for my C code.

> >> > It is not meant to trash existing lines!  Nor is it meant to get
> >> > so confused on the first line, that it just indents 8 spaces, even
> >> > when my tab setting is 4, and there's not even a prior line to
> >> > inspire it to do something that silly.
> >>
> >> That might be a misconfiguration on your part.  (Not that I blame you;
> >> the subject is complex.)  indent-relative looks at tab-stop-list and
> >> configuring tab-width is not going to have the effect you might
> >> expect.  (tab-width refers to how tab characters in the text are
> >> displayed, but the TAB key does not always insert tab characters.)
> >
> > So it sometimes (ie at beginning of file) looks at tab-stop-list,
> > and everywhere else it looks at the indentation of the previous
> > line.  Seems very strange to me.
>
> Hm.  I was going to answer that indent-relative does that, but in
> fact indent-relative-maybe also has this strange behavior in the
> first line of the file.  I think that might be a bug.

Ok.

> Note that indent-relative goes by tab-stop-list after the end of the
> previous line, whereas indent-relative-maybe does not do that.

Ok.  Hopefully with your latest lisp stuff I will finally get the
basic editting that I am after.

BFN.  Paul.

Re: simple editor required

[Stefan Monnier]

> Thanks for agreeing!  What's more surprising is that 10%
> of the user base didn't report this 10 years ago, and thus
> was fixed years before I they installed 20.7.1 on this box.

It's probably because most users actually like to use the tool
in the way it was intended to work rather than work hard to
use it in a different way and then be surprised that it doesn't
do what they want.

> I am mostly using text mode to write C code.

No wonder it doesn't work well.  Try to use tar-mode to edit those
C files and you'll have even more fun ;-)


        Stefan

Re: simple editor required

[Paul Edwards]

"Stefan Monnier" <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> wrote in message news:5l1xxj6pqf.fsf@rum.cs.yale.edu...
> > Thanks for agreeing!  What's more surprising is that 10%
> > of the user base didn't report this 10 years ago, and thus
> > was fixed years before I they installed 20.7.1 on this box.
>
> It's probably because most users actually like to use the tool
> in the way it was intended to work rather than work hard to
> use it in a different way and then be surprised that it doesn't
> do what they want.

So I'm the only person, in the last 10 years, who has ever
needed to maintain C code in various unpredictable formats,
and just needs simple, logical indentation, to avoid having
to hit space a lot?

> > I am mostly using text mode to write C code.
>
> No wonder it doesn't work well.  Try to use tar-mode to edit those
> C files and you'll have even more fun ;-)

:-)

BFN.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> Is there any mode I can switch too, where I can get the behaviour
> I want, without it making a whole set of new problems for me
> (e.g. automatic assumptions about C coding style).  The only
> assumption I want is that it is indented text, so if the previous
> line started in column 16, then the cursor should be positioned
> there when I start a new line.

Well, recent versions of CC mode have the variable
c-syntactic-indentation that you can set to nil from a mode hook:

(defun paul-c-setup ()
  (setq c-syntactic-indentation nil))
(add-hook 'c-mode-hook 'paul-c-setup)

I won't be all that different from text mode, but you can get syntax
highlighting.  Maybe you will like that.

-- 
~/.signature

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> I don't need syntax highlighting, what I need is indent to work
> logically.  Is cc's indenting more logical?

My thinking was if you have the choice between two modes that don't
behave logically, you might prefer the one which at least does syntax
highlighting ;-)

You could configure C mode in a similar way, like text mode.  So you
can get the now-found logical behavior and also syntax highlighting.
-- 
~/.signature

Re: simple editor required

[Paul Edwards]

"Paul Edwards" <kerravon@nosppaam.w3.to> wrote in message news:tscKa.709$9g2.6501@news-server.bigpond.net.au...
> I guess while I'm on a roll I should see if I can get xemacs to
> behave like emacs, so that next time I'm on a site with xemacs
> only, I can get it to behave as logically as emacs (now) does!  :-)

I've now got both emacs and xemacs behaving properly.

Thanks everyone for your help.  The final solution is below,
for anyone who wants a simple editor...

BFN.  Paul.


; Switch off any crap the installation has
(setq inhibit-default-init 't)

; Don't let emacs guess what file types I am editting
(setq auto-mode-alist '() interpreter-mode-alist '())

; Always assume text mode
(setq-default default-major-mode 'text-mode)

; Make backspace backspace
(global-set-key [?\C-h] 'delete-backward-char)

; Set default tab width
(setq-default tab-width 4)

; Don't create junk backup files
(setq make-backup-files nil)
(setq auto-save-default nil)

; make insert toggle overwrite mode
(global-set-key [insertchar] 'overwritemode)

; Don't use tabs to indent
(setq-default indent-tabs-mode nil)

; Indent whenever you hit enter
(defun newline-and-maybe-indent ()
  (interactive)
  (if (bolp) (newline) (newline-and-indent)))
(global-set-key [?\C-m] 'newline-and-maybe-indent)
(setq indent-line-function 'indent-relative-maybe)

; Don't add newlines when cursoring down
(setq-default next-line-add-newlines nil)

; micro-emacs lookalikes
(global-set-key "\M-g" 'goto-line)
(global-set-key "\M-r" 'replace-string)
(global-set-key "\M- " 'set-mark-command)

; For xemacs at least, don't destroy marker on goto
(setq-default zmacs-regions nil)

; Make delete delete
(define-key key-translation-map "\C-?" "\C-d")

; potential alternatives to make delete delete
;(define-key text-mode-map (kbd "DEL") 'delete-char)
;(global-unset-key [delete])
;(global-set-key (kbd "<delete>") #'delete-char)
;(define-key emacs-lisp-mode-map (kbd "DEL") 'delete-char)
;(normal-erase-is-backspace-mode 0)
;(define-key function-key-map "\C-?" "\C-d")
;(global-set-key [delete] 'delete-char)
;(define-key function-key-map [delete] "\C-d")
;(global-unset-key [deletechar])
;(global-set-key [deletechar] 'delete-char)
;(global-unset-key (kbd "<delete>"))
;(global-set-key (kbd "<delete>") 'delete-char)

; If some things are meant to be emacs/xemacs-specific, put them here
(if (featurep 'xemacs)
    (progn
      ;; do Xemacs stuff
      )
  ;; do Emacs stuff
  )

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> Actually, I don't mind it indenting after a blank line, micro-emacs
> does that.  But micro-emacs won't indent the CURRENT
> line when I hit enter at the BEGINNING of the line.  Emacs not
> only inserts a blank line (as I wanted), but is also inspired to
> indent the current line for no reason whatsoever.

In Emacs, newline-and-indent does like newline first, and then like
indent.  It seems to be logical to me, even though I agree that it is
not always what I want.  But the behaivor is quite predictable, so
I'm not bothered too much.

FWIW, I do NOT use newline-and-indent in text mode.  Instead, I use
newline (bound to RET) and indent-relative (bound to TAB).

    When I start a new paragraph that's indented, like this one, I
    usually just keep on typing, like here, and the subsequent lines
    will be similarly indented.

This is because of auto-fill.

> So what about when you are editting text, and it is lined up in
> semi-paragraphs, e.g.
>
> My doctor's details are:
>   Name = Fred
>   Phone = 1234
>
>
> My dentist's details are:
>   Name = Mary
>   Phone = 5678
>
> And the cursor happens to be at the beginning of "My dentist's details"
> and you decide that you want an extra blank line to separate doctor
> from dentist, so you hit enter.  You're not disturbed that emacs is
> inspired to not only insert a blank line, but also to indent "My dentist"
> in a couple of characters, to line up with "Phone = "?

No, I'm not disturbed.  Above I explained why.
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1188 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:84wufgke19.fsf@lucy.is.informatik.uni-duisburg.de...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > If I am at the beginning of the "cccc" line and I hit enter,
> > emacs is inspired to indent cccc to be under bbbb.  I don't
> > want that, I want cccc to stay where it is.  If I wanted cccc
> > to be indented I would do something, e.g. press tab or hit
> > spaces.
>
> Okay, so if the previous line exists and is not blank, then indent
> the new line to the previous line after RET.
>
> Yeah, I'm afraid you'll have to write your own indentation function.
> Maybe this one?
>
> (defun pe-indent-relative ()
>   "Like `indent-relative' but don't indent after blank line."
>   (when (save-excursion
>           (and (zerop (forward-line -1))
>                (not (looking-at "\s-*$"))))
>     (indent-relative-maybe)))
>
> (defun pe-text-setup ()
>   (setq indent-line-function 'pe-indent-relative))
>
> (add-hook 'text-mode-hook 'pe-text-setup)
>
> Maybe it works, but I haven't tested it.

This didn't work (had no effect), but am I meant to change the
binding of my RET or something like that?

Thanks.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> This didn't work (had no effect), but am I meant to change the
> binding of my RET or something like that?

No, but you need to be in text mode.  Are you in text mode?

Use M-x debug-on-entry RET to find out if the pe-indent-relative
indentation function is ever called.
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 790 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:84isqwlxf8.fsf@lucy.is.informatik.uni-duisburg.de...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > This didn't work (had no effect), but am I meant to change the
> > binding of my RET or something like that?
>
> No, but you need to be in text mode.  Are you in text mode?

Yes.

> Use M-x debug-on-entry RET to find out if the pe-indent-relative
> indentation function is ever called.

Ok, will try.  But actually, I've thought of a better way of
abstracting it that will make it more useful.  If RET could
be newline-and-indent except when I am positioned at the
beginning of the line (in which case it should be newline
only) then that is what would be ideal.

Can you think of a way to do that?

Thanks.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> Ok, will try.  But actually, I've thought of a better way of
> abstracting it that will make it more useful.  If RET could
> be newline-and-indent except when I am positioned at the
> beginning of the line (in which case it should be newline
> only) then that is what would be ideal.
>
> Can you think of a way to do that?

(defun pe-newline-and-maybe-indent ()
  (interactive)
  (if (bolp) (newline) (newline-and-indent)))

-- 
This line is not blank.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> And one more.  If I cursor down in xemacs, xemacs doesn't
> automatically start expanding the size of my file when I reach
> the end, but emacs does.  How do I stop that?

C-h k C-n ought to give you a pointer to the variable
next-line-add-newlines.
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 474 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:84znkgoe9v.fsf@lucy.is.informatik.uni-duisburg.de...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > And one more.  If I cursor down in xemacs, xemacs doesn't
> > automatically start expanding the size of my file when I reach
> > the end, but emacs does.  How do I stop that?
>
> C-h k C-n ought to give you a pointer to the variable
> next-line-add-newlines.

Yes, that worked, thanks.

BFN.  Paul.

Re: simple editor required

[Kevin Dziulko]

I'm glad you got your del key to work as desired.  These are not answers 
to your questions, but are related, and are usefull features to know 
about.

First, Emacs with the tramp or ange-ftp packages, one can edit remote 
files as if they were local.  So you could use your favorite version of
emacs all the time.

Second, Emacs and XEmacs are separate entities.  You might consider
fixing all your problems with just one of them first.  Then you can do
something like:

(if (featurep 'xemacs)
    (progn
      ;; do Xemacs stuff
      )
  ;; do Emacs stuff
  )

in your .emacs file, because the same settings may not be possible 
globally.  You may even need to if/then/else the different versions of 
emacs within your .emacs file.

Just wanted to raise awareness of possibilities ...

Kevin

On Tue, 17 Jun 2003, Paul Edwards wrote:

> "Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:84n0git5vn.fsf@lucy.is.informatik.uni-duisburg.de...
> 
> Can I ask the next (minor) thing on my list of things I've
> been living with for x years?
> 
> In text mode, if my file looks like:
> 
> aaaa
> bbbb
> cccc
> 
> Hitting enter at beginning of aaaa line creates a newline plus an indent of
> 8 spaces.  (my tab is only 4 spaces), but I didn't actually expect any
> indentation at all here.
> 
> Hitting enter at beginning of cccc line creates a newline only (as wanted).
> 
> My workaround is to type "x" then hit enter, then go up and delete
> the line with "x" on it.
> 
> What's the rationale and fix for that?
> 
> And one more.  If I cursor down in xemacs, xemacs doesn't
> automatically start expanding the size of my file when I reach
> the end, but emacs does.  How do I stop that?
> 
> Thanks.  Paul.
> 
> 
> _______________________________________________
> Help-gnu-emacs mailing list
> Help-gnu-emacs@gnu.org
> http://mail.gnu.org/mailman/listinfo/help-gnu-emacs
> 

Re: simple editor required

[Stefan Monnier]

> aaaa
> bbbb
> cccc

> Hitting enter at beginning of aaaa line creates a newline plus an indent of

Aaah... stupid me, I tested it by hitting C-j (aka newline-and-indent)
after `aaaa' rather than before.  Of course I can reproduce the problem,
since it's a problem I fixed in Emacs-CVS.


        Stefan

Re: simple editor required

[Paul Edwards]

"Stefan Monnier" <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> wrote in message news:5lu1aqcev2.fsf@rum.cs.yale.edu...
> >> (define-key emacs-lisp-mode-map (kbd "DEL") 'delete-char)
> > AND THE WINNER IS!!!  :-)
>
> Actually, I think the best answer for you is
>
>    (define-key function-key-map "\C-?" "\C-d")

(define-key function-key-map "\C-?" "\C-d")

after commenting out the thing that actually works, the
above doesn't do anything, ie DEL is still backspace.

BFN.  Paul.

Re: simple editor required

[Stefan Monnier]

> (define-key function-key-map "\C-?" "\C-d")
> after commenting out the thing that actually works, the
> above doesn't do anything, ie DEL is still backspace.

I think you're right.  My brain is still full of crap.
The mapping is ignored because there's a binding for ?\C-?,
so you'd want to use key-translation-map instead:

    (define-key key-translation-map "\C-?" "\C-d")

As I said, it's probably better to fix your text-terminal to send
another key sequence than DEL.


        Stefan

Re: simple editor required

[Paul Edwards]

"Stefan Monnier" <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> wrote in message news:5lvfv4bsep.fsf@rum.cs.yale.edu...
> > (define-key function-key-map "\C-?" "\C-d")
> > after commenting out the thing that actually works, the
> > above doesn't do anything, ie DEL is still backspace.
>
> I think you're right.  My brain is still full of crap.
> The mapping is ignored because there's a binding for ?\C-?,
> so you'd want to use key-translation-map instead:
>
>     (define-key key-translation-map "\C-?" "\C-d")

This worked, so I have adopted it and commented out the
other thing that worked, as your one is more generic, in
case I ever use non-text mode.

> As I said, it's probably better to fix your text-terminal to send
> another key sequence than DEL.

I'm not sure that DEL sending DEL is actually wrong!  :-)

Anyway, thanks for the new fix.

BFN.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> I'm not sure that DEL sending DEL is actually wrong!  :-)

For Emacs, DEL and <delete> are two very different things.  The
<delete> key is near <home> and <end>, and usually deletes forward.
The DEL key, however, is conventionally known as <backspace> and
deletes backward.

Do not be confused by the similarity in names.

Do you still uphold that statement?  ;-)
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1066 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:84vfv3y9ha.fsf@lucy.is.informatik.uni-duisburg.de...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > I'm not sure that DEL sending DEL is actually wrong!  :-)
>
> For Emacs, DEL and <delete> are two very different things.  The
> <delete> key is near <home> and <end>, and usually deletes forward.
> The DEL key, however, is conventionally known as <backspace> and
> deletes backward.
>
> Do not be confused by the similarity in names.
>
> Do you still uphold that statement?  ;-)

I didn't get the bit after "now listen carefully Max".  :-)

Ok, it is the <delete> key I am pressing then (near <home>).
But when Emacs is reporting DEL, it doesn't mean it just
saw ascii 127, it means it saw a backspace (C-h).  Incidentally,
people are always referring to "press C-h then xxx", but of
course I have remapped C-h to be backspace, I mean, C-h is
backspace, and we all know what backspace does.  So I
duly replace C-h with Esc x help.  It seems a very surprising
key binding to me.

BFN.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> Ok, it is the <delete> key I am pressing then (near <home>).
> But when Emacs is reporting DEL, it doesn't mean it just
> saw ascii 127, it means it saw a backspace (C-h).

Huh?  What does C-h l say after you hit whatever key it is that Emacs
reports as DEL?

(Is Emacs reporting DEL when you hit <delete>?)

It is possible that normal-erase-is-backspace has an influence, here,
and that it is set to the wrong value in your Emacs.  (The variable
normal-erase-is-backspace should not be set via setq, if you want to
use Lisp, use the function normal-erase-is-backspace-mode.)

> Incidentally, people are always referring to "press C-h then xxx",
> but of course I have remapped C-h to be backspace, I mean, C-h is
> backspace, and we all know what backspace does.  So I duly replace
> C-h with Esc x help.  It seems a very surprising key binding to me.

Whee.  Does <f1> work as a help key?

Maybe it's the normal-erase-is-backspace thing.  If not, then I'm
flabbergasted.
-- 
This line is not blank.

Re: simple editor required

[Paul Edwards]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1628 bytes --]

"Kai Großjohann" <kai.grossjohann@gmx.net> wrote in message news:8465n1983n.fsf@lucy.is.informatik.uni-duisburg.de...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > Ok, it is the <delete> key I am pressing then (near <home>).
> > But when Emacs is reporting DEL, it doesn't mean it just
> > saw ascii 127, it means it saw a backspace (C-h).
>
> Huh?  What does C-h l say after you hit whatever key it is that Emacs
> reports as DEL?

DEL ESC x h e l p RET l

Note that the first key I hit was the <delete> (next to <home>).

> (Is Emacs reporting DEL when you hit <delete>?)

Yep!

> It is possible that normal-erase-is-backspace has an influence, here,

I'm on emacs 20.7.1 which doesn't have that.

> and that it is set to the wrong value in your Emacs.  (The variable
> normal-erase-is-backspace should not be set via setq, if you want to
> use Lisp, use the function normal-erase-is-backspace-mode.)
>
> > Incidentally, people are always referring to "press C-h then xxx",
> > but of course I have remapped C-h to be backspace, I mean, C-h is
> > backspace, and we all know what backspace does.  So I duly replace
> > C-h with Esc x help.  It seems a very surprising key binding to me.
>
> Whee.  Does <f1> work as a help key?

Yes, it does!  Although seems a bit different from M-x help, the
latter comes up with the menu straight away, the former suggests
I press "?" to get it.

> Maybe it's the normal-erase-is-backspace thing.  If not, then I'm
> flabbergasted.

Everything to do with backspace/delete is fine, just that help
I have to (now) use F1 instead of C-h.  This is not a problem
at all.

BFN.  Paul.

Re: simple editor required

[Stefan Monnier]

> I'm not sure that DEL sending DEL is actually wrong!  :-)

You're confusing the key and the character it sends (and the behavior you
want to associate with the key, and the behavior you want to associate
with the character).

The key you call DEL is the key you want to delete-forward, but the
character DEL is the event sequence sent by the terminal to Emacs that is
normally associated with delete-backward: clearly the "DEL" key should not
send a "DEL" character in your case (and this depends on your expectations,
not on the name of the key or the name of the character).


        Stefan

Re: simple editor required

[Stefan Monnier]

> (global-set-key "?\C-?" 'delete-char)

Boy, how stupid can I be?

> (global-set-key "\C-?" 'delete-char)

Yes, that's what I meant.  Try

(add-hook 'c-common-mode-hook
  (lambda () (local-set-key "\C-?" 'delete-char)))

Since c-mode and derivatives insist on rebind the key.


        Stefan

Re: simple editor required

[Stefan Monnier]

> The second problem I encountered was that in the version where
> text mode actually took effect, was that it inserted spaces instead
> of leaving the line empty, when I inserted a blank line.  I haven't
> seen emacs do that before, even when it used to work on an older
> version.  But this is perhaps a case of faulty memory.

I believe that's faulty memory.  Emacs has to insert spaces in order to
place the cursor at the proper place.  But newline-and-indent does delete
trailing space before inserting the newline, so that should work as long
as you bind RET to newline-and-indent (which you obviously would like to see
be the default).

> The third problem I encountered was when trying to manually
> clean up the blanks, ctrl-k would not simply delete the blanks on
> the line, it would delete the entire line.

> Someone has suggested putting:
> (setq kill-whole-line nil)

> to get around this problem.  I didn't have to do that in other
> versions of emacs, it was the basic standard consistent behaviour.

As far as I know, C-k has always behaved like that (at least since
version 18.59): it deletes either "[ \t]*\n" or ".*[^ \t\n].*".
Basically, Emacs encourages you to ignore trailing whitespace and
consider it as non-present.  You can ask Emacs to automatically erase all
trailing whitespace when you save the file, although I don't think that was
a standard feature feature before Emacs-21, so it probably won't help you.

>> Try indented-text-mode instead.

indented-text-mode is the same as text-mode since Emacs-20, so it will
only make a difference if you ever use Emacs-19.

> 2. Block-marking not to be destroyed by ctrl-f and esc-g.  Xemacs
> 20.4 has a you-beaut highlighting of the block, which is great, but
> I'd trade it any day for the ability to do a find/goto to position to
> end of block.  Like emacs 20.7.1.  And like the last time I used a
> version of xemacs, IIRC.  I don't know why some people think it
> is appropriate to forget the marker, seems pretty basic to me, but
> besides, other emacs (like 20.7.1) don't forget the marker.

For XEmacs complaints, go to comp.emacs.xemacs, this an Emacs group.

> 3. DEL to delete.  Just like xemacs 20.4 with appropriate .emacs entry.

You can try (define-key function-key-map [delete] [?\C-d]),
but be warned that on some systems, the backspace key supposedly
sends a `delete' event.
Also it won't do anything on text-terminals where there's simply no
agreement as to what key corresponds to DEL (aka C-?) or to C-h, so that
there's no generic answer.

> I consider these to be extremely basic things that should be provided

Everyone has his own definition of normality.


        Stefan

Re: simple editor required

[Eli Zaretskii]

> From: "Stefan Monnier" <monnier+gnu.emacs.help/news/@flint.cs.yale.edu>
> Newsgroups: gnu.emacs.help
> Date: 07 Jun 2003 11:06:49 -0400
> 
> >> Try indented-text-mode instead.
> 
> indented-text-mode is the same as text-mode since Emacs-20, so it will
> only make a difference if you ever use Emacs-19.

I have no idea what version the OP uses, and he did say that in one of
the versions text-mode didn't do what we both know it should have
done.  So I thought there could be no harm to try indented-text-mode.
I thought that line of reasoning was obvious.

Re: simple editor required

[Johan Bockgård]

"Ben Key" <bkey1@tampabay.rr.com> writes:

> Stefan Monnier wrote:
>> Should be "\\.c\\'", although it only makes a difference if there's a
> newline in your file name.
>
> I did not know it was possible to have a new line character in a file
> name.  At least I do not think it is possible in DOS or MS Windows.  Is
> it possible to have a new line character in a file name in Gnu Linux /
> Unix?

Solaris:

drwxr-xr-x   8 bojohan  ftekn        512 Jun  2 01:14 emacs/
drwx------   2 bojohan  ftekn       1024 May 27 00:42 files/
-rw-r--r--   1 bojohan  ftekn         12 Jun 11 04:51 foo
bar
drwxr-xr-x   3 bojohan  ftekn       1024 May 27 18:32 haskell/

Apparently :-)

-- 
The day Microsoft makes something that doesn't suck is probably the
day they start making vacuum cleaners. -- Ernst Jan Plugge

Re: simple editor required

[Stefan Monnier]

> I did not know it was possible to have a new line character in a file
> name.  At least I do not think it is possible in DOS or MS Windows.  Is
> it possible to have a new line character in a file name in Gnu Linux /
> Unix?

Sure: anything but \0 and / is allowed.
Newline chars in file names are great to exploit some security holes.
Typically when a cron script does `find /tmp | blabla'.


        Stefan

Re: simple editor required

[Paul Edwards]

"Kevin Dziulko" <dziulko@klaatu.canisius.edu> wrote in message news:mailman.8100.1055869193.21513.help-gnu-emacs@gnu.org...
> First, Emacs with the tramp or ange-ftp packages, one can edit remote
> files as if they were local.  So you could use your favorite version of
> emacs all the time.

For some reason I can't ftp to my remote systems.  I've already
asked why not.  sysadmin mentioned firewall, but it was
presumably all too hard.  I can ftp from the remote system to
the local though.

> Second, Emacs and XEmacs are separate entities.  You might consider
> fixing all your problems with just one of them first.

Yes, I'm starting with emacs.  99+% of my work is done on local
machines, and emacs 20.7.1 is available on all of them.  I only
found xemacs on the remote, but I don't need to do complex stuff
there.

> Then you can do something like:

> (if (featurep 'xemacs)
>    (progn
>      ;; do Xemacs stuff
>      )
>  ;; do Emacs stuff
>  )

Thanks!  I will stick that in too.

BFN.  Paul.

Re: simple editor required

[David Kastrup]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> Yes, I'm starting with emacs.  99+% of my work is done on local
> machines, and emacs 20.7.1 is available on all of them.

20.7 is really old.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: simple editor required

[Paul Edwards]

"David Kastrup" <dak@gnu.org> wrote in message news:x5d6hcjosm.fsf@lola.goethe.zz...
> "Paul Edwards" <kerravon@nosppaam.w3.to> writes:
>
> > Yes, I'm starting with emacs.  99+% of my work is done on local
> > machines, and emacs 20.7.1 is available on all of them.
>
> 20.7 is really old.

And I'm the only emacs user at my site too.  I believe the
features I am after should be part of much earlier versions
than that anyway, I just need to coax emacs into giving
them to me.

BFN.  Paul.

Re: simple editor required

[Kai Großjohann]

"Paul Edwards" <kerravon@nosppaam.w3.to> writes:

> "Kevin Dziulko" <dziulko@klaatu.canisius.edu> wrote in message news:mailman.8100.1055869193.21513.help-gnu-emacs@gnu.org...
>> First, Emacs with the tramp or ange-ftp packages, one can edit remote
>> files as if they were local.  So you could use your favorite version of
>> emacs all the time.
>
> For some reason I can't ftp to my remote systems.  I've already
> asked why not.  sysadmin mentioned firewall, but it was
> presumably all too hard.  I can ftp from the remote system to
> the local though.

If you can log in to the remote system, Tramp will probably help with
remote editing.  Warning: I wrote Tramp.

-- 
This line is not blank.