bug#2136: cleanup: Fresume_tty and Fsuspend_tty

bug#2136: cleanup: Fresume_tty and Fsuspend_tty

[Dan Nicolaescu]

These functions should not have any #ifdef MSDOS  or #ifdef subprocesses,
they cannot be used on such systems.  That code is clearly untested anyway.

Or maybe the can just have an #ifndef MSDOS around the whole body that
makes them do nothing.

bug#2136: cleanup: Fresume_tty and Fsuspend_tty

[Eli Zaretskii]

> Date: Sat, 31 Jan 2009 10:46:45 -0800 (PST)
> From: Dan Nicolaescu <dann@ics.uci.edu>
> Cc: 
> 
> These functions should not have any #ifdef MSDOS  or #ifdef subprocesses,
> they cannot be used on such systems.

The only uses up till now don't have any trouble.

> That code is clearly untested anyway.

I did test them at the time.

bug#2136: cleanup: Fresume_tty and Fsuspend_tty

[Dan Nicolaescu]

Eli Zaretskii <eliz@gnu.org> writes:

  > > Date: Sat, 31 Jan 2009 10:46:45 -0800 (PST)
  > > From: Dan Nicolaescu <dann@ics.uci.edu>
  > > Cc: 
  > > 
  > > These functions should not have any #ifdef MSDOS  or #ifdef subprocesses,
  > > they cannot be used on such systems.
  > 
  > The only uses up till now don't have any trouble.

Can you please describe exactly how you used them?

  > > That code is clearly untested anyway.
  > 
  > I did test them at the time.

bug#2136: cleanup: Fresume_tty and Fsuspend_tty

[Eli Zaretskii]

> Date: Sat, 31 Jan 2009 12:14:25 -0800 (PST)
> From: Dan Nicolaescu <dann@ics.uci.edu>
> Cc: 2136@emacsbugs.donarmstrong.com
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
>   > > Date: Sat, 31 Jan 2009 10:46:45 -0800 (PST)
>   > > From: Dan Nicolaescu <dann@ics.uci.edu>
>   > > Cc: 
>   > > 
>   > > These functions should not have any #ifdef MSDOS  or #ifdef subprocesses,
>   > > they cannot be used on such systems.
>   > 
>   > The only uses up till now don't have any trouble.
> 
> Can you please describe exactly how you used them?

For suspend-tty, I also used suspend-frame.  For resume-tty, I just
invoked it directly.

bug#2136: cleanup: Fresume_tty and Fsuspend_tty

[Eli Zaretskii]

> Date: Sat, 31 Jan 2009 22:40:22 +0200
> From: Eli Zaretskii <eliz@gnu.org>
> Cc: 2136@emacsbugs.donarmstrong.com
> 
> For suspend-tty, I also used suspend-frame.

"Also" here means "in addition to invoking it directly".  Sorry for
confusing wording.

bug#2136: cleanup: Fresume_tty and Fsuspend_tty

[Dan Nicolaescu]

Eli Zaretskii <eliz@gnu.org> writes:

  > > Date: Sat, 31 Jan 2009 12:14:25 -0800 (PST)
  > > From: Dan Nicolaescu <dann@ics.uci.edu>
  > > Cc: 2136@emacsbugs.donarmstrong.com
  > > 
  > > Eli Zaretskii <eliz@gnu.org> writes:
  > > 
  > >   > > Date: Sat, 31 Jan 2009 10:46:45 -0800 (PST)
  > >   > > From: Dan Nicolaescu <dann@ics.uci.edu>
  > >   > > Cc: 
  > >   > > 
  > >   > > These functions should not have any #ifdef MSDOS  or #ifdef subprocesses,
  > >   > > they cannot be used on such systems.
  > >   > 
  > >   > The only uses up till now don't have any trouble.
  > > 
  > > Can you please describe exactly how you used them?
  > 
  > For suspend-tty, I also used suspend-frame. 

Can you have multiple ttys in MSDOS?
Otherwise suspend-frame won't call supend-tty.

  >  For resume-tty, I just invoked it directly.

See above, this function only makes sens when using multiple ttys. 

bug#2136: cleanup: Fresume_tty and Fsuspend_tty

[Eli Zaretskii]

> Date: Sat, 31 Jan 2009 13:22:16 -0800 (PST)
> From: Dan Nicolaescu <dann@ics.uci.edu>
> Cc: 2136@emacsbugs.donarmstrong.com
> 
> Can you have multiple ttys in MSDOS?

No, of course not.  Not if by ``tty'' you mean an interactive
terminal.

> Otherwise suspend-frame won't call supend-tty.

It can be tricked into doing that, since the test in controlling-tty-p
is very naive.

>   >  For resume-tty, I just invoked it directly.
> 
> See above, this function only makes sens when using multiple ttys. 

If you mean the full functionality, then yes, I agree.  However, since
the current uses of these two functions are so few, I cannot yet say
that having them not fboundp will or won't do harm for the DOS port.
For now, they are more or less no-ops, which certainly doesn't do any
harm, at least AFAICS.

On the other hand, keeping them out of the DOS build would mean the
same #ifdef's will have to stay, just in different places, which won't
make the sources more clean, IMO.

bug#2136: cleanup: Fresume_tty and Fsuspend_tty

[Dan Nicolaescu]

Eli Zaretskii <eliz@gnu.org> writes:

  > > Date: Sat, 31 Jan 2009 13:22:16 -0800 (PST)
  > > From: Dan Nicolaescu <dann@ics.uci.edu>
  > > Cc: 2136@emacsbugs.donarmstrong.com
  > > 
  > > Can you have multiple ttys in MSDOS?
  > 
  > No, of course not.  Not if by ``tty'' you mean an interactive
  > terminal.
  > 
  > > Otherwise suspend-frame won't call supend-tty.
  > 
  > It can be tricked into doing that, since the test in controlling-tty-p
  > is very naive.

Then the test needs to be improved...

  > >   >  For resume-tty, I just invoked it directly.
  > > 
  > > See above, this function only makes sens when using multiple ttys. 
  > 
  > If you mean the full functionality, then yes, I agree.  However, since
  > the current uses of these two functions are so few, I cannot yet say
  > that having them not fboundp will or won't do harm for the DOS port.
  > For now, they are more or less no-ops, which certainly doesn't do any
  > harm, at least AFAICS.

It complicates the code, I just recently fixed in bug there, that's how
I run into this.

  > On the other hand, keeping them out of the DOS build would mean the
  > same #ifdef's will have to stay, just in different places, which won't
  > make the sources more clean, IMO.

Given that the functions are no-ops for DOS, just to put an #ifndef
inside the body and make them just return Qnil, that way it's pretty
clear they  are no-ops.

bug#2136: cleanup: Fresume_tty and Fsuspend_tty

[Glenn Morris]

Sorry, I see no value in keeping this report open.
Closing as wontfix.