emacs -Q not documented

emacs -Q not documented

[Werner LEMBERG]

[emacs 2005-03-29]

Option `-Q' isn't documented in emacs --help.


    Werner

Re: emacs -Q not documented

[Werner LEMBERG]

>     Option `-Q' isn't documented in emacs --help.
> 
> Would you like to fix that?

The patch is trivial, but we should probably add a long name for it.
What about `--bare'?


    Werner

Re: emacs -Q not documented

[David Kastrup]

Werner LEMBERG <wl@gnu.org> writes:

>>     Option `-Q' isn't documented in emacs --help.
>> 
>> Would you like to fix that?
>
> The patch is trivial, but we should probably add a long name for it.
> What about `--bare'?

How about --plain or --vanilla?

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: emacs -Q not documented

[Nick Roberts]

 > >>     Option `-Q' isn't documented in emacs --help.
 > >> 
 > >> Would you like to fix that?
 > >
 > > The patch is trivial, but we should probably add a long name for it.
 > > What about `--bare'?

This would be confusing because a 'bare Emacs' means something else.

 > How about --plain or --vanilla?

I don't think these are explanatory.

If we can't think of a suitable name why should we add one?


Nick

Re: emacs -Q not documented

[Miles Bader]

On Apr 2, 2005 8:02 PM, Nick Roberts <nickrob@snap.net.nz> wrote:
> I don't think these are explanatory.
> 
> If we can't think of a suitable name why should we add one?

Well obviously if we can't think of a good name, we shouldn't add a
bad one :-), but long option names, if well-chosen, are often _much_
easier to remember if you only use the option very occasionally...

So it's worth at least a bit of effort to come up with one.  I think
it needn't be _completely_ self-explanatory, as long as it's easy to
remember, and "makes sense" after reading the options documentation or
--help output.  Something like `--vanilla' is probably not going to
indicate to a completely naive user what's going on, but it's almost
certainly meaningful enough[*] to jog the memory of someone who has
read the documentation in the past (whereas -Q is more iffy).

How about `--default-settings' (which could be abbreviated
`--default'; perhaps it ought to also accept the plural of that,
`--defaults', without actually putting it in the option help)?

[*] For a native english speaker -- this is the big problem with
`--vanilla' I think.

-Miles
-- 
Do not taunt Happy Fun Ball.

Re: emacs -Q not documented

[David Kastrup]

Miles Bader <snogglethorpe@gmail.com> writes:

> How about `--default-settings' (which could be abbreviated
> `--default'; perhaps it ought to also accept the plural of that,
> `--defaults', without actually putting it in the option help)?

Since the settings are _not_ the defaults (for example, blink-cursor
is off), that would be extremely misleading.

--plain would be another proposal.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: emacs -Q not documented

[Kim F. Storm]

Miles Bader <snogglethorpe@gmail.com> writes:

> How about `--default-settings' (which could be abbreviated
> `--default'; perhaps it ought to also accept the plural of that,
> `--defaults', without actually putting it in the option help)?

I think that would be very confusing -- -Q actually turns off some of
the default features such as splash screen, toolbar, menubar, scroll
bars etc.

--no-nothing  :-)


-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: emacs -Q not documented

[Miles Bader]

On Apr 3, 2005 7:52 AM, Kim F. Storm <storm@cua.dk> wrote:
> I think that would be very confusing -- -Q actually turns off some of
> the default features such as splash screen, toolbar, menubar, scroll
> bars etc.
> 
> --no-nothing  :-)

`--minimal-defaults' ?

-Miles
-- 
Do not taunt Happy Fun Ball.

Re: emacs -Q not documented

[Henrik Enberg]

Miles Bader <snogglethorpe@gmail.com> writes:

> On Apr 3, 2005 7:52 AM, Kim F. Storm <storm@cua.dk> wrote:
>> I think that would be very confusing -- -Q actually turns off some of
>> the default features such as splash screen, toolbar, menubar, scroll
>> bars etc.
>> 
>> --no-nothing  :-)
>
> `--minimal-defaults' ?

How about just `--minimal'?  Should be fairly self-explanatory.

-- 
Vaya Con Satan

Re: emacs -Q not documented

[Miles Bader]

On Apr 3, 2005 8:21 AM, Henrik Enberg <henrik.enberg@telia.com> wrote:
> > `--minimal-defaults' ?
> 
> How about just `--minimal'?  Should be fairly self-explanatory.

I think it's somewhat misleading --

-Q apparently does _two_ things:  (1) it ignores all customizations,
and (2) uses `minimal settings' (turns off toolbars etc.)

`--minimal' implies that it only does (2).

`--minimal-defaults' explicitly refers to both (1) and (2).

-Miles
-- 
Do not taunt Happy Fun Ball.

Re: emacs -Q not documented

[Nick Roberts]

 > > How about `--default-settings' (which could be abbreviated
 > > `--default'; perhaps it ought to also accept the plural of that,
 > > `--defaults', without actually putting it in the option help)?
 > 
 > I think that would be very confusing -- -Q actually turns off some of
 > the default features such as splash screen, toolbar, menubar, scroll
 > bars etc.
 > 
 > --no-nothing  :-)

--no-toolbar-menubar-scrollbars-blinking-cursor-splash-screen-init-or-site-file

says it all, but

-Q

which one could guess is "quicker" than -q, is a bit more to the point.


Nick

Re: emacs -Q not documented

[Richard Stallman]

    The patch is trivial, but we should probably add a long name for it.

I guess so--if we keep it.  But the set of things it does is not
really coherent.

It turns off all init files; it also turns off various frame features
such as the menu bar and tool bar, it turns off tool tips,
and it selects a different type of cursor.  There is nothing particularly
natural or desirable about this combination.  I see no reason to have
a short option for it.  Perhaps we should just remove it.

Why was -Q installed, anyway?

Re: emacs -Q not documented

[Miles Bader]

On Apr 4, 2005 6:20 AM, Richard Stallman <rms@gnu.org> wrote:
> Why was -Q installed, anyway?

I seem to recall it was intended for some debugging scenario (e.g.
turning off lots of features helps simplify things).

I agree it's overly weird; the "no init files" functionality is very
useful though, and naturally maps to -Q (by analogy with -q).  Maybe
if those other features are useful they could be invoked by a separate
option.

-Miles
-- 
Do not taunt Happy Fun Ball.

Re: emacs -Q not documented

[Miles Bader]

On Apr 4, 2005 6:20 AM, Richard Stallman <rms@gnu.org> wrote:
> Why was -Q installed, anyway?

I seem to recall it was intended for some debugging scenario (e.g.
turning off lots of features helps simplify things).

I agree it's overly weird.  The "no init files" functionality is very
useful though, and naturally maps to -Q (by analogy with -q).  Maybe
if those other features are useful they could be triggered by a
separate option; specifying two short options is not much harder than
one.

-Miles
-- 
Do not taunt Happy Fun Ball.

Re: emacs -Q not documented

[Kim F. Storm]

Richard Stallman <rms@gnu.org> writes:

>     The patch is trivial, but we should probably add a long name for it.
>
> I guess so--if we keep it.  But the set of things it does is not
> really coherent.
>
> It turns off all init files; it also turns off various frame features
> such as the menu bar and tool bar, it turns off tool tips,
> and it selects a different type of cursor.  There is nothing particularly
> natural or desirable about this combination.  

It is intended to be an option turn off all "extra features".

This is _extremely_ useful for reporting and debugging redisplay
problems, as it removes a lot of extra stuff that is most often
irrelevant to a specific problem.

Just like we ask people to try emacs -q for reported problems in lisp
code, it is very useful to ask a person to try reproducing a redisplay
problem with emacs -Q and if the problem exists in that situation, you
can quickly eliminate a lot of possible causes of the problem.

But most important, running emacs -Q when debugging redisplay problems
makes it much easier to know what's going on.

>                                               I see no reason to have
> a short option for it.  Perhaps we should just remove it.

Just because we cannot find a long name for it -- that's silly, IMO!
Please keep the option.  It serves the purpose it was added for very
well!!

>
> Why was -Q installed, anyway?

You approved it...



-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: emacs -Q not documented

[Miles Bader]

On Apr 5, 2005 4:31 PM, Kim F. Storm <storm@cua.dk> wrote:
> Just because we cannot find a long name for it -- that's silly, IMO!
> Please keep the option.  It serves the purpose it was added for very
> well!!

There seems no reason not to split it into a couple of more coherent
options though.

-Miles
-- 
Do not taunt Happy Fun Ball.

Re: emacs -Q not documented

[Kim F. Storm]

Miles Bader <snogglethorpe@gmail.com> writes:

> On Apr 5, 2005 4:31 PM, Kim F. Storm <storm@cua.dk> wrote:
>> Just because we cannot find a long name for it -- that's silly, IMO!
>> Please keep the option.  It serves the purpose it was added for very
>> well!!
>
> There seems no reason not to split it into a couple of more coherent
> options though.

So instead of "emacs -Q" I have to ask a user to please try
"emacs -q -A -T --no-bla-bla -V --no-blinking-cursor"

As I said: 

Please keep the option.  It serves the purpose it was added for very well!!

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: emacs -Q not documented

[Miles Bader]

On Apr 5, 2005 7:04 PM, Kim F. Storm <storm@cua.dk> wrote:
> > There seems no reason not to split it into a couple of more coherent
> > options though.
> 
> So instead of "emacs -Q" I have to ask a user to please try
> "emacs -q -A -T --no-bla-bla -V --no-blinking-cursor"

That's a strawman.

I meant something like:

   emacs -Q -N

where -Q means "no init files or site-init files"
and -N means turn off all frame features

I think typing a few extra letters in such rare cases is hardly going
to cause much hardship, and the improvement in clarity makes it
worthwhile.

-Miles
-- 
Do not taunt Happy Fun Ball.

Re: emacs -Q not documented

[Miles Bader]

On Apr 5, 2005 7:27 PM, Miles Bader <snogglethorpe@gmail.com> wrote:
>    emacs -Q -N
> 
> where -Q means "no init files or site-init files"
> and -N means turn off all frame features

I should add that doing so is more flexible for debugging too -- for
instance, for reporting a bug with menus, you don't want frame
features turned off, but you _do_ want to disable site-init files.

This scenario (where you want to disable all customization, but not
change emacs' "default state") is probably the most common, so it
makes sense to give -Q that common and clear meaning, and add
something like -N for the rarer case where you suspect the frame
features are interfering.

-Miles
-- 
Do not taunt Happy Fun Ball.

Re: emacs -Q not documented

[Kim F. Storm]

Miles Bader <snogglethorpe@gmail.com> writes:

> On Apr 5, 2005 7:27 PM, Miles Bader <snogglethorpe@gmail.com> wrote:
>>    emacs -Q -N
>> 
>> where -Q means "no init files or site-init files"
>> and -N means turn off all frame features
>
> I should add that doing so is more flexible for debugging too -- for
> instance, for reporting a bug with menus, you don't want frame
> features turned off, but you _do_ want to disable site-init files.

-N also disables the splash screen, blinking cursor, and potentially
other stuff that is not closely related to "frame features".

What long name will you use for -N?

So the next thing somebody will tell us is that -N consists of
unrelated settings, and suggest to split it into -N and -M and ?

> This scenario (where you want to disable all customization, but not
> change emacs' "default state") is probably the most common, so it
> makes sense to give -Q that common and clear meaning, 

IMO, emacs -Q (in its current form) should be the reference for
reproducing errors -- if possible with even more features/settings
turned off (e.g. ignore .Xdefaults, Windows registry settings, etc).

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: emacs -Q not documented

[Robert J. Chassell]

   >  -Q

   It is intended to be an option turn off all "extra features".

Yes.  The option is helpful and should be kept.

It is easier than evoking Emacs with an equivalent long string of
options, which are:

    emacs -q             \
    --no-site-file       \
    --no-blinking-cursor \
    --eval '(progn
              (tooltip-mode -1)
              (tool-bar-mode -1)
              (menu-bar-mode -1)
              (scroll-bar-mode -1)
              (setq inhibit-startup-message t))'

(`nil' instead of `-1' causes the evaluated functions to produce a
message, so is no good.)

If we need a long name for `-Q', and do not wish to use `--plain',
please consider

    --no-init-site-splash-decorations-blinking

which is long and ugly but does tell you what the option avoids.

--
    Robert J. Chassell
    bob@rattlesnake.com                         GnuPG Key ID: 004B4AC8
    http://www.rattlesnake.com                  http://www.teak.cc

Re: emacs -Q not documented

[David Kastrup]

"Robert J. Chassell" <bob@rattlesnake.com> writes:

>    >  -Q
>
>    It is intended to be an option turn off all "extra features".
>
> Yes.  The option is helpful and should be kept.
>
> It is easier than evoking Emacs with an equivalent long string of
> options, which are:
>
>     emacs -q             \
>     --no-site-file       \
>     --no-blinking-cursor \
>     --eval '(progn
>               (tooltip-mode -1)
>               (tool-bar-mode -1)
>               (menu-bar-mode -1)
>               (scroll-bar-mode -1)
>               (setq inhibit-startup-message t))'

Well, that argument would hold for any option string.

> (`nil' instead of `-1' causes the evaluated functions to produce a
> message, so is no good.)
>
> If we need a long name for `-Q', and do not wish to use `--plain',
> please consider
>
>     --no-init-site-splash-decorations-blinking
>
> which is long and ugly but does tell you what the option avoids.

--basic-defaults

would be another possibility.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: emacs -Q not documented

[Kim F. Storm]

David Kastrup <dak@gnu.org> writes:

> --basic-defaults
>
> would be another possibility.

Still, "defaults" is non-informative in this context.

IMO --minimal would be good, but is too close to --minimize.

Some synonymes are:

--basic
--slim
--mean
--reduced
--nominal

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: emacs -Q not documented

[David Kastrup]

storm@cua.dk (Kim F. Storm) writes:

> David Kastrup <dak@gnu.org> writes:
>
>> --basic-defaults
>>
>> would be another possibility.
>
> Still, "defaults" is non-informative in this context.
>
> IMO --minimal would be good, but is too close to --minimize.
>
> Some synonymes are:
>
> --basic
> --slim
> --mean
> --reduced
> --nominal

--stripped-down

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: emacs -Q not documented

[Kim F. Storm]

David Kastrup <dak@gnu.org> writes:

> --stripped-down

--naked  :-)

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: emacs -Q not documented

[David Kastrup]

storm@cua.dk (Kim F. Storm) writes:

> David Kastrup <dak@gnu.org> writes:
>
>> --stripped-down
>
> --naked  :-)

Too close to --flashing.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: emacs -Q not documented

[Richard Stallman]

    It is intended to be an option turn off all "extra features".

On what criterion are these features "extra"?

    But most important, running emacs -Q when debugging redisplay problems
    makes it much easier to know what's going on.

Why does the blinking cursor relate to this?

    Just because we cannot find a long name for it -- that's silly, IMO!

The difficulty in finding a long name for it reflected a lack of
apparent coherence of the functionality of the option.  Nobody
could see what it was good for.  That is why I thought of removing
it or changing it.

Re: emacs -Q not documented

[Kim F. Storm]

Richard Stallman <rms@gnu.org> writes:

>     It is intended to be an option turn off all "extra features".
>
> On what criterion are these features "extra"?

They are various "decorations" to the basic Emacs frame or
functionality.  

Turning them off, skips over a fairly big (and complex) part of
internal redisplay code (e.g. building menu and tool bars which may
run quite a deal of lisp code).

>
>     But most important, running emacs -Q when debugging redisplay problems
>     makes it much easier to know what's going on.
>
> Why does the blinking cursor relate to this?

It means running timers and async Lisp code, and triggers redisplay
which may influense how a bug in redisplay manifests itself on the
screen.

Furthermore, it means you cannot just put a breakpoint on (or anywhere
inside) redisplay_internal and then execute some code to generate a
specific error...  The blinking cursor timer will trigger that
breakpoint every second, making it practically impossible to
debug such problems.

The -Q option has saved me many hours debugging rediplay problems!

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: emacs -Q not documented

[Nick Roberts]

 > The -Q option has saved me many hours debugging rediplay problems!

If Kim finds this option useful then it surely makes sense to keep it? It
presumably carries no overhead and, AFAIK, initial options aren't a limited
resource . If no-one else finds it useful, why are we arguing about adding a
long name for it?


Nick

Re: emacs -Q not documented

[Richard Stallman]

    Furthermore, it means you cannot just put a breakpoint on (or anywhere
    inside) redisplay_internal and then execute some code to generate a
    specific error...  The blinking cursor timer will trigger that
    breakpoint every second, making it practically impossible to
    debug such problems.

It seems like a bug that blinking the cursur should require a real
redisplay, if the buffer has not changed and point has not even moved.
Can't this be done by redrawing the glyph instead?

    The -Q option has saved me many hours debugging rediplay problems!

I think this should be split into two options.

-Q, or --quick, should turn off use of both kinds of init file
and the splash screen.

--basic-display should turn off various display features,
including the blinking cursor.

-Q --bas  would do them both.

Re: emacs -Q not documented

[Werner LEMBERG]

> The difficulty in finding a long name for it reflected a lack of
> apparent coherence of the functionality of the option.  Nobody could
> see what it was good for.  That is why I thought of removing it or
> changing it.

Well, -Q has definitely its merits, especially for reporting bugs.
After seeing this discussion I still think that my `--bare' proposal
is not that bad: The -Q options really strips off all features you
would normally like to have for daily work.


    Werner

Re: emacs -Q not documented

[Kim F. Storm]

Werner LEMBERG <wl@gnu.org> writes:

> Well, -Q has definitely its merits, especially for reporting bugs.
> After seeing this discussion I still think that my `--bare' proposal
> is not that bad: The -Q options really strips off all features you
> would normally like to have for daily work.

Yes, --bare-bones or --no-frills seem like good candidates, too.

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: emacs -Q not documented

[David Kastrup]

storm@cua.dk (Kim F. Storm) writes:

> Werner LEMBERG <wl@gnu.org> writes:
>
>> Well, -Q has definitely its merits, especially for reporting bugs.
>> After seeing this discussion I still think that my `--bare' proposal
>> is not that bad: The -Q options really strips off all features you
>> would normally like to have for daily work.
>
> Yes, --bare-bones or --no-frills seem like good candidates, too.

--bare-bones seems ok.  But things like menus are not "frills", and we
don't want to spread that impression.  I don't want to get bug reports
from people that did not bother using the help menu or tutorial
announced on the splash screen because they decided from the manual
page that they were not the type of person interested in frills.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: emacs -Q not documented

[Kim F. Storm]

David Kastrup <dak@gnu.org> writes:

> --bare-bones seems ok.

I have installed --bare-bones as an alias for -Q.

Case closed, I hope :-)

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: emacs -Q not documented

[Andreas Schwab]

storm@cua.dk (Kim F. Storm) writes:

> Werner LEMBERG <wl@gnu.org> writes:
>
>> Well, -Q has definitely its merits, especially for reporting bugs.
>> After seeing this discussion I still think that my `--bare' proposal
>> is not that bad: The -Q options really strips off all features you
>> would normally like to have for daily work.
>
> Yes, --bare-bones or --no-frills seem like good candidates, too.

Since the option is primarily useful for debugging maybe its name should
contain the word debug.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: emacs -Q not documented

[David Kastrup]

Andreas Schwab <schwab@suse.de> writes:

> storm@cua.dk (Kim F. Storm) writes:
>
>> Werner LEMBERG <wl@gnu.org> writes:
>>
>>> Well, -Q has definitely its merits, especially for reporting bugs.
>>> After seeing this discussion I still think that my `--bare' proposal
>>> is not that bad: The -Q options really strips off all features you
>>> would normally like to have for daily work.
>>
>> Yes, --bare-bones or --no-frills seem like good candidates, too.
>
> Since the option is primarily useful for debugging maybe its name
> should contain the word debug.

-debug-setup

This could also set debug-on-error.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: emacs -Q not documented

[Stefan Monnier]

>>> Yes, --bare-bones or --no-frills seem like good candidates, too.
>> Since the option is primarily useful for debugging maybe its name
>> should contain the word debug.
> -debug-setup

My experience when programming is that functions shouldn't be named based on
what they're used for, but based on what they do.
I think the same should apply to command line options.  Even if -Q is mostly
useful for debugging, that doesn't mean its name should mention anything
about debugging.
I think the current --bare-bones is a better name, so I think we should
close this discussion.  We have better things to do.


        Stefan