C-x SPC : rectangle-mark-mode or gud-break?

C-x SPC : rectangle-mark-mode or gud-break?

[Michael Welsh Duggan]

I know I am opening up a can of worms here, but I am going to argue that
`C-x SPC' be changed back to `gud-break'.  I've tried using the current
system for a couple of weeks now, and my fingers still automatically go
to `C-x SPC' instead of `C-x C-a C-b'.

I must admit to wondering why more people on this list aren't
experiencing this as well.  Do people not run gdb from within Emacs?  Do
they tend to use the toolbar to set breakpoints?  I'm don't get very
upset when new features are turned on by default in new versions of
Emacs, but I find myself annoyed when long-lived keybindings that I use
frequently are overwritten by functionality that (at least in my case)
is much less common.  Emacs is a programmers' editor, isn't it?

Now, please note that I do know how to rebind keys, and have re-bound
`C-x SPC' in my .emacs to that which I expect.  And my frustration with
this new rebinding isn't directed at anyone in particular.  It's just
frustration.  (And I have a co-worker who I know is going to go
ballistic once he types `C-x SPC' in the next version of Emacs and the
unexpected happens.)

If `C-x SPC' remains as `rectangle-mark-mode', please make sure that it
gets mentioned in the NEWS such that a search for some combination of
"gud", "gdb", and "breakpoint" will find the change, along with the `C-x
C-a C-b' binding.  Also please make sure the documentation for
`gud-break' in the "Commands of GUD" node of the manual gets updated.
(Currently it doesn't even mention the `C-x C-a C-b' binding.)

-- 
Michael Welsh Duggan
(md5i@md5i.com)

RE: C-x SPC : rectangle-mark-mode or gud-break?

[Drew Adams]

> I know I am opening up a can of worms here, but I am going to argue that
> `C-x SPC' be changed back to `gud-break'.

1. `gud-break' should never have been given a global binding on prefix
   `C-x'.  It is fine for it to have a binding on `gud-key-prefix', i.e.,
   `C-x C-a'.  Why does it need a global binding?  And if it really does,
   for some reason, why sacrifice such an important global prefix key as
   `C-x' for it?

2. `C-x SPC' should never have been given as a global key binding to
   `rectangle-mark-mode'.

`C-x SPC' should be left UNbound.  If and when a library or Emacs itself
binds `C-x SPC', there should be good reasons for that.  Other things
being equal, if bound, `C-x SPC' should be used as a prefix key, letting
users take advantage of `C-x SPC SPC SPC...' for a repeatable command.

There is no good reason (none) to bind `C-x SPC' to `rectangle-mark-mode'
in the global map.  That command should have been put on prefix key
`C-x r', like the other rectangle commands.  `C-x' is far too precious
to waste on this.

AFAICT, not one argument was ever given in favor of binding `C-x SPC'
to `rectangle-mark-mode'.  The only real support for it was this
non-argument from Eli: "`C-x SPC' is available", followed by "Why is
everybody ignoring the `C-x SPC' suggestion?"

And all objections to this binding, including the reasons given
against, were simply ignored.  The only replies to the objections were
messages in support of the objections - also ignored.

http://lists.gnu.org/archive/html/emacs-devel/2013-11/msg00046.html

Re: C-x SPC : rectangle-mark-mode or gud-break?

[Michael Welsh Duggan]

Drew Adams <drew.adams@oracle.com> writes:

>> I know I am opening up a can of worms here, but I am going to argue that
>> `C-x SPC' be changed back to `gud-break'.
>
> 1. `gud-break' should never have been given a global binding on prefix
> `C-x'.  It is fine for it to have a binding on `gud-key-prefix', i.e.,
> `C-x C-a'.  Why does it need a global binding?  And if it really does,
> for some reason, why sacrifice such an important global prefix key as
> `C-x' for it?

Nothing other than the fact that the keybinding has been around since at
least Emacs 18.58 (the first version I regularly used).  It's had a lot
of time for people to get used to it.

-- 
Michael Welsh Duggan
(md5i@md5i.com)

RE: C-x SPC : rectangle-mark-mode or gud-break?

[Drew Adams]

> >> I know I am opening up a can of worms here, but I am going to argue that
> >> `C-x SPC' be changed back to `gud-break'.
> >
> > 1. `gud-break' should never have been given a global binding on prefix
> >    `C-x'.  It is fine for it to have a binding on `gud-key-prefix', i.e.,
> >    `C-x C-a'.  Why does it need a global binding?  And if it really does,
> >    for some reason, why sacrifice such an important global prefix key as
> >    `C-x' for it?
> 
> Nothing other than the fact that the keybinding has been around since at
> least Emacs 18.58 (the first version I regularly used).

Not all bad decisions are recent. ;-)

More seriously, whether or not it might have made some sense to spend
a global keymap key on this in the 1980s, it does not make sense to do
that now, IMO.  What's wrong with using `gud-key-prefix' as the prefix
key for this?

> It's had a lot of time for people to get used to it.

And?

RE: C-x SPC : rectangle-mark-mode or gud-break?

[Stephen J. Turnbull]

Drew Adams writes:

 > > It's had a lot of time for people to get used to it.
 > 
 > And?

History matters, Drew.  AFAIK you don't prefer the current binding to
a rectangle function, but want the key sequence left for future use.

"Reserved for future use" is not really on the table here, given a
twenty-year history for one of the bindings.  Not everybody is going
to upgrade to the latest and greatest Emacs (eg, about 50% each of two
different corporate samples I've seen, both over 200 users, used
Emacsen released more than 5 years before the dates of the surveys),
but they do sometimes share others' Emacsen.

At least in my experience debugging sessions are more likely to end up
"hands on" than coding sessions; this is a binding that matters.

RE: C-x SPC : rectangle-mark-mode or gud-break?

[Drew Adams]

>  > > It's had a lot of time for people to get used to it.
>  > And?
> 
> History matters, Drew.

Pretty vague platitude.  So does chocolate matter.

Emacs Dev can decide whatever it wants for `C-x SPC'.

> AFAIK you don't prefer the current binding to
> a rectangle function, but want the key sequence left for future use.

Too narrow.

What I suggested was to unbind it by default, for now.  That means that
it would be open for libraries to use (and of course for users), and
it would be open for future default use by Emacs.  NOT only the latter.

> "Reserved for future use" is not really on the table here, given a
> twenty-year history for one of the bindings.

(Not just for future use.)

And yes, that can be on the table if Emacs Dev wants it on the table.
Gud has its own prefix key.  Why not use it for this command too?

> Not everybody is going to upgrade to the latest and greatest Emacs...,
> but they do sometimes share others' Emacsen.

Sure, I'm susceptible to that argument.

It is the reason, for instance, that I would prefer that we not swap
`C-j' and `RET': I use multiple Emacs versions everyday, and I prefer
to use those two keys the same way for each Emacs version.  And thank
goodness there is an easy way to keep the traditional `C-j' and `RET'
behavior.

But that inconvenience for me and my muscle memory did not prevent
Emacs Dev from swapping the two keys.  Same thing here.

Nothing prevents someone from binding `C-x SPC' to the same gud
command s?he has enjoyed since the 80s.  Freeing up the key makes
sense for all users, libraries, and future default use by Emacs.

It is a handy key.  What a shame to waste it on just setting a
gud breakpoint.  That's all.  There's no reason we cannot
reconsider this decades-old default global key binding.

But you are right that the other point here is also important:
put that rectangle command on the rectangle-commands prefix key.
Don't waste `C-x SPC' on it either.

RE: C-x SPC : rectangle-mark-mode or gud-break?

[Stephen J. Turnbull]

Drew Adams writes:

 > > History matters, Drew.
 > 
 > Pretty vague platitude.  So does chocolate matter.

Sure, and Richard wants a pony^Wwordprocessor.  It's only vague here
to you; I'm sure everybody else reading my post understood exactly
what I meant.

 > Emacs Dev can decide whatever it wants for `C-x SPC'.

Of course it can; what I (and M. Duggan) are suggesting is that it
*should* be reverted to unbound by core (or the rectangle library,
which also has a perfectly good prefix of C-x r and no history of
binding anything to C-x SPC), leaving it free for gud to bind (at
gud's risk if another good candidate for binding to C-x SPC should
appear in the future).

 > > AFAIK you don't prefer the current binding to
 > > a rectangle function, but want the key sequence left for future use.
 > 
 > Too narrow.
 > 
 > What I suggested was to unbind it by default, for now.

Well, in fact that's the way it is now (as of 24.3.1).  It's just that
gud installs its binding in the global keymap when loaded.

 > That's all.  There's no reason we cannot reconsider this
 > decades-old default global key binding.

We *are* reconsidering it.  I'm just expressing my opinion that when
there's a 20-year history behind a particular binding, there's good
reason to decide to leave it in place.  And that I see no similar
reason in favor of blocking it by using it for a rectangle command.

Re: C-x SPC : rectangle-mark-mode or gud-break?

[Stefan Monnier]

> I know I am opening up a can of worms here, but I am going to argue that
> `C-x SPC' be changed back to `gud-break'.

Oddly enough there hasn't been the expected deluge of opinions.


        Stefan

Re: C-x SPC : rectangle-mark-mode or gud-break?

[Michael Welsh Duggan]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>> I know I am opening up a can of worms here, but I am going to argue that
>> `C-x SPC' be changed back to `gud-break'.
>
> Oddly enough there hasn't been the expected deluge of opinions.

It surprises me, too.  My only guess is that people just don't use gdb
in emacs much, or they tend to use the mouse to set breakpoints.  (I
highly doubt people are using `C-x C-a C-b'.)  I guess I'll just have to
suggest rebinding it in gud-gdb-mode-hook to people who like it better
the way it was before.

I know that the documentation updates are still a work in progress, but
please make sure current references to `C-x SPC' in the manual are
changed to `C-x C-a C-b' respectively.

-- 
Michael Welsh Duggan
(md5i@md5i.com)

Re: C-x SPC : rectangle-mark-mode or gud-break?

[Stefan Monnier]

> highly doubt people are using `C-x C-a C-b'.)  I guess I'll just have to
> suggest rebinding it in gud-gdb-mode-hook to people who like it better
> the way it was before.

Some people suggested to rebind it in prog-mode buffers, based on the
assumption that the rectangle commands are not very useful in
programming language modes, whereas GUD bindings are only useful there.


        Stefan

Re: C-x SPC : rectangle-mark-mode or gud-break?

[Michael Welsh Duggan]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>> highly doubt people are using `C-x C-a C-b'.)  I guess I'll just have to
>> suggest rebinding it in gud-gdb-mode-hook to people who like it better
>> the way it was before.
>
> Some people suggested to rebind it in prog-mode buffers, based on the
> assumption that the rectangle commands are not very useful in
> programming language modes, whereas GUD bindings are only useful there.

That would be good, certainly.  Is there a reliable way to tell the
difference between a programming mode and another mode?

-- 
Michael Welsh Duggan
(md5i@md5i.com)

Re: C-x SPC : rectangle-mark-mode or gud-break?

[Stefan Monnier]

> That would be good, certainly.  Is there a reliable way to tell the
> difference between a programming mode and another mode?

A programming mode should derive from prog-mode (they don't all do,
because prog-mode is fairly recent).


        Stefan

Re: C-x SPC : rectangle-mark-mode or gud-break?

[Bozhidar Batsov]

[-- Attachment #1: Type: text/plain, Size: 542 bytes --]

Good to know. Sorry about the personal email. From time to time I use a web email client and I often
forget it behaves differently from my standard client.                  

-- 
Cheers,
Bozhidar


On Thursday, January 16, 2014 at 3:25 PM, Stefan Monnier wrote:

> > Personally, I find rectangle commands quite useful in programming modes
> > (and I'm pretty sure I'm not the only one).
> > 
> 
> 
> I was only repeating someone else's suggestion, as it seemed relevant.
> I already rejected that suggestion for the default.
> 
> 
> Stefan 


[-- Attachment #2: Type: text/html, Size: 1367 bytes --]

Re: C-x SPC : rectangle-mark-mode or gud-break?

[Bastien]

Hi Michael,

Michael Welsh Duggan <mwd@md5i.com> writes:

> I know that the documentation updates are still a work in progress, but
> please make sure current references to `C-x SPC' in the manual are
> changed to `C-x C-a C-b' respectively.

Done in trunk (revno 116054), thanks.

-- 
 Bastien