Removing MULTI_KBOARD

Removing MULTI_KBOARD

[Eli Zaretskii]

> Date: Wed, 30 Jul 2008 22:47:53 -0400
> From: emacs-diffs-request@gnu.org
> 
> --- src/ChangeLog	30 Jul 2008 21:45:32 -0000	1.6815
> +++ src/ChangeLog	31 Jul 2008 02:47:26 -0000	1.6816
> @@ -1,3 +1,24 @@
> +2008-07-31  Dan Nicolaescu  <dann@ics.uci.edu>
> +
> +	* s/ms-w32.h (MULTI_KBOARD): Remove.
> +	* xterm.c:
> +	* xselect.c:
> +	* xfns.c:
> +	* window.c:
> +	* w32term.c:
> +	* w32fns.c:
> +	* terminal.c:
> +	* termhooks.h:
> +	* term.c:
> +	* sysdep.c:
> +	* keyboard.h:
> +	* keyboard.c:
> +	* frame.h:
> +	* frame.c:
> +	* frame.c: Remove references to MULTI_KBOARD, it is now the
> +	default.

Did we agree to remove these conditionals?  The MS-DOS port does not
support MULTI_KBOARD, so now it is more broken than it was before.

Re: Removing MULTI_KBOARD

[Dan Nicolaescu]

Eli Zaretskii <eliz@gnu.org> writes:

  > > Date: Wed, 30 Jul 2008 22:47:53 -0400
  > > From: emacs-diffs-request@gnu.org
  > > 
  > > --- src/ChangeLog	30 Jul 2008 21:45:32 -0000	1.6815
  > > +++ src/ChangeLog	31 Jul 2008 02:47:26 -0000	1.6816
  > > @@ -1,3 +1,24 @@
  > > +2008-07-31  Dan Nicolaescu  <dann@ics.uci.edu>
  > > +
  > > +	* s/ms-w32.h (MULTI_KBOARD): Remove.
  > > +	* xterm.c:
  > > +	* xselect.c:
  > > +	* xfns.c:
  > > +	* window.c:
  > > +	* w32term.c:
  > > +	* w32fns.c:
  > > +	* terminal.c:
  > > +	* termhooks.h:
  > > +	* term.c:
  > > +	* sysdep.c:
  > > +	* keyboard.h:
  > > +	* keyboard.c:
  > > +	* frame.h:
  > > +	* frame.c:
  > > +	* frame.c: Remove references to MULTI_KBOARD, it is now the
  > > +	default.
  > 
  > Did we agree to remove these conditionals?  The MS-DOS port does not
  > support MULTI_KBOARD, so now it is more broken than it was before.

The code now is cleaner and easier to read.  Unused code does not get in
your way as it used to happen before.
The difference in usage from a port point a view is minimal: a single
static object vs a single dynamically allocated object. The full history
is available in CVS.

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> From: Dan Nicolaescu <dann@ics.uci.edu>
> Cc: emacs-devel@gnu.org
> Date: Wed, 30 Jul 2008 20:43:01 -0700
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
>   > Did we agree to remove these conditionals?
> 
> The code now is cleaner and easier to read.

So is that a NO?

Re: Removing MULTI_KBOARD

[Dan Nicolaescu]

Eli Zaretskii <eliz@gnu.org> writes:

  > > From: Dan Nicolaescu <dann@ics.uci.edu>
  > > Cc: emacs-devel@gnu.org
  > > Date: Wed, 30 Jul 2008 20:43:01 -0700
  > > 
  > > Eli Zaretskii <eliz@gnu.org> writes:
  > > 
  > >   > Did we agree to remove these conditionals?
  > > 
  > > The code now is cleaner and easier to read.
  > 
  > So is that a NO?

Sorry I am not playing the partial citation to distort the meaning game.

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=failed 
> 	version=3.1.0
> From: Dan Nicolaescu <dann@ics.uci.edu>
> Cc: emacs-devel@gnu.org
> Date: Thu, 31 Jul 2008 12:21:58 -0700
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
>   > > From: Dan Nicolaescu <dann@ics.uci.edu>
>   > > Cc: emacs-devel@gnu.org
>   > > Date: Wed, 30 Jul 2008 20:43:01 -0700
>   > > 
>   > > Eli Zaretskii <eliz@gnu.org> writes:
>   > > 
>   > >   > Did we agree to remove these conditionals?
>   > > 
>   > > The code now is cleaner and easier to read.
>   > 
>   > So is that a NO?
> 
> Sorry I am not playing the partial citation to distort the meaning game.

Dan what happened here? what ``meaning game''?  I asked you a simple
question: did we agree to remove MULTI_KBOARD or didn't we?  I asked
because I don't remember whether such a discussion took place and if
it did, what was the conclusion.  Why won't you give me an answer to a
simple question?

Re: Removing MULTI_KBOARD

[Dan Nicolaescu]

Eli Zaretskii <eliz@gnu.org> writes:

  > did we agree to remove MULTI_KBOARD or didn't we?

Why?

Re: Removing MULTI_KBOARD

[Juanma Barranquero]

On Fri, Aug 1, 2008 at 14:09, Dan Nicolaescu <dann@ics.uci.edu> wrote:
> Eli Zaretskii <eliz@gnu.org> writes:
>
>  > did we agree to remove MULTI_KBOARD or didn't we?
>
> Why?

Well, obviously because, if it was not discussed, that's the kind of
wide-scale change that *should* have been discussed.

   Juanma

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> From: Dan Nicolaescu <dann@ics.uci.edu>
> Cc: emacs-devel@gnu.org
> Date: Fri, 01 Aug 2008 05:09:29 -0700
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
>   > did we agree to remove MULTI_KBOARD or didn't we?
> 
> Why?

Because if it was discussed and I missed that, or by my silence let it
look like I don't object, I have only myself to blame and will need to
live with the results.  If not, then there are other options.

But basically, I don't understand why a simple question cannot be
answered without further ado.  Is something wrong with how I asked it?

Re: Removing MULTI_KBOARD

[Dan Nicolaescu]

Eli Zaretskii <eliz@gnu.org> writes:

  > > From: Dan Nicolaescu <dann@ics.uci.edu>
  > > Cc: emacs-devel@gnu.org
  > > Date: Fri, 01 Aug 2008 05:09:29 -0700
  > > 
  > > Eli Zaretskii <eliz@gnu.org> writes:
  > > 
  > >   > did we agree to remove MULTI_KBOARD or didn't we?
  > > 
  > > Why?
  > 
  > Because if it was discussed and I missed that, or by my silence let it
  > look like I don't object, I have only myself to blame and will need to
  > live with the results.  If not, then there are other options.
  > 
  > But basically, I don't understand why a simple question cannot be
  > answered without further ado.  Is something wrong with how I asked it?

Yes, big time.

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> From: Dan Nicolaescu <dann@ics.uci.edu>
> Date: Fri, 01 Aug 2008 06:41:20 -0700
> Cc: emacs-devel@gnu.org
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
>   > > From: Dan Nicolaescu <dann@ics.uci.edu>
>   > > Cc: emacs-devel@gnu.org
>   > > Date: Fri, 01 Aug 2008 05:09:29 -0700
>   > > 
>   > > Eli Zaretskii <eliz@gnu.org> writes:
>   > > 
>   > >   > did we agree to remove MULTI_KBOARD or didn't we?
>   > > 
>   > > Why?
>   > 
>   > Because if it was discussed and I missed that, or by my silence let it
>   > look like I don't object, I have only myself to blame and will need to
>   > live with the results.  If not, then there are other options.
>   > 
>   > But basically, I don't understand why a simple question cannot be
>   > answered without further ado.  Is something wrong with how I asked it?
> 
> Yes, big time.

Apologies.

Anyway, does anyone else know if this was discussed, and if so, can
point me to the relevant thread?  TIA

Re: Removing MULTI_KBOARD

[Richard M Stallman]

    >  > did we agree to remove MULTI_KBOARD or didn't we?
    >
    > Why?

    Well, obviously because, if it was not discussed, that's the kind of
    wide-scale change that *should* have been discussed.

You are right about that.

Re: Removing MULTI_KBOARD

[Dan Nicolaescu]

Richard M Stallman <rms@gnu.org> writes:

  >     >  > did we agree to remove MULTI_KBOARD or didn't we?
  >     >
  >     > Why?
  > 
  >     Well, obviously because, if it was not discussed, that's the kind of
  >     wide-scale change that *should* have been discussed.
  > 
  > You are right about that.

Sorry, but this approach is not a good start.
The code was turned off by default, unused and unnecessary, and that has
been explained on the list almost a year ago.  If someone has a problem
with that, it would be much more productive to explain why, but not just
hand waving, with some real experience/examples.  Otherwise please let
bygones be bygones, it's a terrible waste of time.

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> From: Dan Nicolaescu <dann@ics.uci.edu>
> Cc: "Juanma Barranquero" <lekktu@gmail.com>, eliz@gnu.org, emacs-devel@gnu.org
> Date: Fri, 01 Aug 2008 23:55:57 -0700
> 
> The code was turned off by default, unused and unnecessary, and that has
> been explained on the list almost a year ago.  If someone has a problem
> with that, it would be much more productive to explain why, but not just
> hand waving, with some real experience/examples.  Otherwise please let
> bygones be bygones, it's a terrible waste of time.

I did explain why I have a problem with this: it breaks the MSDOS
port.

Now, since removing MULTI_KBOARD obviously was not discussed in
advance, I'm asking Yidong and Stefan to decide which of the following
is preferable:

  . Revert Dan's changes that removed MULTI_KBOARD.

  . Replace those #ifdef's that are needed by the MSDOS port (I don't
    know whether this means all of them or not) with #ifdef MSDOS.

Re: Removing MULTI_KBOARD

[Dan Nicolaescu]

Eli Zaretskii <eliz@gnu.org> writes:

  > > From: Dan Nicolaescu <dann@ics.uci.edu>
  > > Cc: "Juanma Barranquero" <lekktu@gmail.com>, eliz@gnu.org, emacs-devel@gnu.org
  > > Date: Fri, 01 Aug 2008 23:55:57 -0700
  > > 
  > > The code was turned off by default, unused and unnecessary, and that has
  > > been explained on the list almost a year ago.  If someone has a problem
  > > with that, it would be much more productive to explain why, but not just
  > > hand waving, with some real experience/examples.  Otherwise please let
  > > bygones be bygones, it's a terrible waste of time.
  > 
  > I did explain why I have a problem with this: it breaks the MSDOS
  > port.

You have handwaved that.  Can you please actually explain why you think that?
I have explained it is not a problem.

Given that you really insist on this, maybe it would be a good idea to
for have a look at the whole picture:

1. The MSDOS port is broken for other reasons, it has been like that for a
year.  Nobody has complained.
2. The MSDOS port was on a widely publicized list of platforms that we
indented to remove.  Nobody requested it to be kept, so RMS approved its
removal.
3. The MSDOS port might never be revived, you have never committed to do
so, the only statements you have made was "I might do it".

All working platforms use MULTI_KBOARD by default, removing the extra
code makes everything cleaner and easier to read, it was removed because
it has been getting in my way when trying to chase an infrequently
occurring event/input bug.

  > Now, since removing MULTI_KBOARD obviously was not discussed in
  > advance, 

This implies that there would be any need for such a discussion.  There
isn't, the code is better now without the #ifdefs.

  > I'm asking Yidong and Stefan to decide which of the following
  > is preferable:
  > 
  >   . Revert Dan's changes that removed MULTI_KBOARD.

Please do not do such a thing, this would be a very bad case of priority
inversion: forcing to keep code just because it might (although it is
doubtful) make it easier to revive a port to a very old, proprietary
platform with NO ACTIVE USERS, otherwise broken and unmaintained for one
year, hence a low priority task for a GNU project.  The code is an
inconvenience and requires more work when doing maintenance and
debugging maintenance on GNU/Linux, i.e a high priority task for GNU.

If the MSDOS port ever gets fixed (which, again, is not sure to happen),
it can be fixed from a CVS checkout prior to the removal of the
MULTI_KBOARD code.  If it turns out that the MULTI_KBOARD code is
needed, it can be restored at that time.  I stand by my claim that such
a thing would not be necessary.

  >   . Replace those #ifdef's that are needed by the MSDOS port (I don't
  >     know whether this means all of them or not) with #ifdef MSDOS.

That is even worse, the unnecessary code is still in the way of all
the GNU/Linux maintainers.

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> From: Dan Nicolaescu <dann@ics.uci.edu>
> Cc: rms@gnu.org, Juanma Barranquero <lekktu@gmail.com>, emacs-devel@gnu.org
> Date: Sat, 02 Aug 2008 02:27:49 -0700
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
>   > > From: Dan Nicolaescu <dann@ics.uci.edu>
>   > > Cc: "Juanma Barranquero" <lekktu@gmail.com>, eliz@gnu.org, emacs-devel@gnu.org
>   > > Date: Fri, 01 Aug 2008 23:55:57 -0700
>   > > 
>   > > The code was turned off by default, unused and unnecessary, and that has
>   > > been explained on the list almost a year ago.  If someone has a problem
>   > > with that, it would be much more productive to explain why, but not just
>   > > hand waving, with some real experience/examples.  Otherwise please let
>   > > bygones be bygones, it's a terrible waste of time.
>   > 
>   > I did explain why I have a problem with this: it breaks the MSDOS
>   > port.
> 
> You have handwaved that.  Can you please actually explain why you think that?
> I have explained it is not a problem.

How it is not a problem, if the MSDOS port does not support
MULTI_KBOARD?  I must have missed the explanation, please repeat it.

> 1. The MSDOS port is broken for other reasons, it has been like that for a
> year.  Nobody has complained.
> 2. The MSDOS port was on a widely publicized list of platforms that we
> indented to remove.  Nobody requested it to be kept, so RMS approved its
> removal.
> 3. The MSDOS port might never be revived, you have never committed to do
> so, the only statements you have made was "I might do it".

All of these reasons might be invalid tomorrow, if I decide to revive
it tomorrow.

>   >   . Revert Dan's changes that removed MULTI_KBOARD.
> 
> Please do not do such a thing, this would be a very bad case of priority
> inversion: forcing to keep code just because it might (although it is
> doubtful) make it easier to revive a port to a very old, proprietary
> platform with NO ACTIVE USERS, otherwise broken and unmaintained for one
> year, hence a low priority task for a GNU project.  The code is an
> inconvenience and requires more work when doing maintenance and
> debugging maintenance on GNU/Linux, i.e a high priority task for GNU.
> 
> If the MSDOS port ever gets fixed (which, again, is not sure to happen),
> it can be fixed from a CVS checkout prior to the removal of the
> MULTI_KBOARD code.  If it turns out that the MULTI_KBOARD code is
> needed, it can be restored at that time.  I stand by my claim that such
> a thing would not be necessary.
> 
>   >   . Replace those #ifdef's that are needed by the MSDOS port (I don't
>   >     know whether this means all of them or not) with #ifdef MSDOS.
> 
> That is even worse, the unnecessary code is still in the way of all
> the GNU/Linux maintainers.

I'm open to other suggestions, but please don't suggest to drop the
MSDOS port until I decide to do that.

Re: Removing MULTI_KBOARD

[Jason Rumney]

Eli Zaretskii wrote:

>> You have handwaved that.  Can you please actually explain why you think that?
>> I have explained it is not a problem.
> 
> How it is not a problem, if the MSDOS port does not support
> MULTI_KBOARD?  I must have missed the explanation, please repeat it.

If anyone is going to get the MSDOS port working again, then the effort
in getting it working with MULTI_KBOARD is going to be similar to or
less than the work required to get the rest of post-multi-tty Emacs
working without MULTI_KBOARD. The reason W32 and Mac now use
MULTI_KBOARD is that the multi-tty changes basically forced it. The
changes were minor compared with what else needs doing to revive the
MSDOS port after the unicode and multi-tty merges.

Re: Removing MULTI_KBOARD

[Chong Yidong]

Jason Rumney <jasonr@gnu.org> writes:

> Eli Zaretskii wrote:
>
>>> You have handwaved that.  Can you please actually explain why you think that?
>>> I have explained it is not a problem.
>> 
>> How it is not a problem, if the MSDOS port does not support
>> MULTI_KBOARD?  I must have missed the explanation, please repeat it.
>
> If anyone is going to get the MSDOS port working again, then the effort
> in getting it working with MULTI_KBOARD is going to be similar to or
> less than the work required to get the rest of post-multi-tty Emacs
> working without MULTI_KBOARD. The reason W32 and Mac now use
> MULTI_KBOARD is that the multi-tty changes basically forced it. The
> changes were minor compared with what else needs doing to revive the
> MSDOS port after the unicode and multi-tty merges.

That's my impression as well.

It would have been nice for Dan to mention he was removing MULTI_KBOARD,
but I can understand the reasoning for what he did (maybe he was rushing
for time due to the feature freeze deadline too).  So let's keep his
changes.

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> From: Chong Yidong <cyd@stupidchicken.com>
> Date: Sat, 02 Aug 2008 10:07:12 -0400
> Cc: lekktu@gmail.com, Eli Zaretskii <eliz@gnu.org>,
> 	Dan Nicolaescu <dann@ics.uci.edu>, rms@gnu.org, emacs-devel@gnu.org
> 
> Jason Rumney <jasonr@gnu.org> writes:
> 
> > Eli Zaretskii wrote:
> >
> >> How it is not a problem, if the MSDOS port does not support
> >> MULTI_KBOARD?  I must have missed the explanation, please repeat it.
> >
> > If anyone is going to get the MSDOS port working again, then the effort
> > in getting it working with MULTI_KBOARD is going to be similar to or
> > less than the work required to get the rest of post-multi-tty Emacs
> > working without MULTI_KBOARD. The reason W32 and Mac now use
> > MULTI_KBOARD is that the multi-tty changes basically forced it. The
> > changes were minor compared with what else needs doing to revive the
> > MSDOS port after the unicode and multi-tty merges.
> 
> That's my impression as well.

I guess we will see if these impressions are true.

Re: Removing MULTI_KBOARD

[Nick Roberts]

 > It would have been nice for Dan to mention he was removing MULTI_KBOARD,
 > but I can understand the reasoning for what he did (maybe he was rushing
 > for time due to the feature freeze deadline too).  So let's keep his
 > changes.

Well, I can't understand the reasoning.  If I was Eli, it would leave a very
bitter taste in my mouth.  Maybe because we don't see one another, there is
sometimes scant regard for the feelings of others on this list, and I'm sure
this loses developers.

If the logical thing to do is remove MULTI_KBOARD, and therefore the MSDOS
port, I'm sure that after discussion Eli would see that.   The outcome may
well be the same but it would be reached in a much more agreeable manner.

-- 
Nick                                           http://www.inet.net.nz/~nickrob

Re: Removing MULTI_KBOARD

[Lennart Borgman (gmail)]

Nick Roberts wrote:
>  > It would have been nice for Dan to mention he was removing MULTI_KBOARD,
>  > but I can understand the reasoning for what he did (maybe he was rushing
>  > for time due to the feature freeze deadline too).  So let's keep his
>  > changes.
> 
> Well, I can't understand the reasoning.  If I was Eli, it would leave a very
> bitter taste in my mouth.  Maybe because we don't see one another, there is
> sometimes scant regard for the feelings of others on this list, and I'm sure
> this loses developers.

Yes, I have been told that this has happened on the w32 side.

> If the logical thing to do is remove MULTI_KBOARD, and therefore the MSDOS
> port, I'm sure that after discussion Eli would see that.   The outcome may
> well be the same but it would be reached in a much more agreeable manner.

It is a bit said that two good efforts, one from Dan and one from Eli, 
clash. However it is also natural that such situations arise within a 
complex system.

I think that both Eli and Dan deserve our support here.

Trying to give them that might help avoid situations like this in the 
future.

Re: Removing MULTI_KBOARD

[Richard M. Stallman]

You have presented good arguments for deciding to delete the
MULTI_KBOARD conditionals, but you must let the maintainers
decide whether to do so.

Re: Removing MULTI_KBOARD

[Richard M. Stallman]

    The code was turned off by default, unused and unnecessary, and that has
    been explained on the list almost a year ago.

Was it discussed on the list a year ago, and was a decision properly made then?

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> From: Nick Roberts <nickrob@snap.net.nz>
> Date: Sun, 3 Aug 2008 10:49:42 +1200
> Cc: emacs-devel@gnu.org
> 
> If I was Eli, it would leave a very bitter taste in my mouth.

Make no mistake about it: the taste is _very_ bitter.

Re: Removing MULTI_KBOARD

[Miles Bader]

Chong Yidong <cyd@stupidchicken.com> writes:
> It would have been nice for Dan to mention he was removing MULTI_KBOARD,
> but I can understand the reasoning for what he did (maybe he was rushing
> for time due to the feature freeze deadline too).  So let's keep his
> changes.

This seems reasonable to me.

Still, it's not _that_ hard to drop a note to emacs-devel when you're
going to make a big or possibly controversial change, and I think it can
helpful for preserving the (relatively :-) congenial atmosphere.

-Miles

-- 
My books focus on timeless truths.  -- Donald Knuth

Re: Removing MULTI_KBOARD

[Stefan Monnier]

> Well, I can't understand the reasoning.  If I was Eli, it would leave a very
> bitter taste in my mouth.  Maybe because we don't see one another, there is
> sometimes scant regard for the feelings of others on this list, and I'm sure
> this loses developers.

Agreed.

> If the logical thing to do is remove MULTI_KBOARD, and therefore the MSDOS
> port, I'm sure that after discussion Eli would see that.

FWIW, removing MULTI_KBOARD does not entail removing MSDOS support: it
just requires a bit more effort to update the MSDOS code.  The problem
is that Dan should have made this effort (even if he can't test it
because the MSDOS port currently doesn't build/work), unless someone
else (i.e. Eli) had agreed to do it, or unless we had decided to drop
the MSDOS port (which I'd favor, but not enough to impose it on Eli).

So Dan, rather than revert your removal of MULTI_KBOARD, could you do
a best-effort update of the MSDOS code to adapt it to the new "always
multi kbd" code?


        Stefan

Re: Removing MULTI_KBOARD

[Dan Nicolaescu]

Stefan Monnier <monnier@IRO.UMontreal.CA> writes:

  > > If the logical thing to do is remove MULTI_KBOARD, and therefore the MSDOS
  > > port, I'm sure that after discussion Eli would see that.
  > 
  > FWIW, removing MULTI_KBOARD does not entail removing MSDOS support: it
  > just requires a bit more effort to update the MSDOS code.  The problem
  > is that Dan should have made this effort (even if he can't test it
  > because the MSDOS port currently doesn't build/work), unless someone
  > else (i.e. Eli) had agreed to do it, or unless we had decided to drop
  > the MSDOS port (which I'd favor, but not enough to impose it on Eli).
  > 
  > So Dan, rather than revert your removal of MULTI_KBOARD, could you do
  > a best-effort update of the MSDOS code to adapt it to the new "always
  > multi kbd" code?

What makes you think I didn't?  To the best of my knowledge no extra
changes are required for the "always multi kbd" code.  No evidence of
the contrary has been given this far.

Re: Removing MULTI_KBOARD

[Stefan Monnier]

>> So Dan, rather than revert your removal of MULTI_KBOARD, could you do
>> a best-effort update of the MSDOS code to adapt it to the new "always
>> multi kbd" code?

> What makes you think I didn't?

Sorry, I just pessimistically assumed so.

> To the best of my knowledge no extra changes are required for the
> "always multi kbd" code.  No evidence of the contrary has been given
> this far.

Thanks.  Eli, could you explain what is the remaining problem with his
change, then?


        Stefan

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> From: Stefan Monnier <monnier@IRO.UMontreal.CA>
> Date: Mon, 04 Aug 2008 15:10:14 -0400
> Cc: Nick Roberts <nickrob@snap.net.nz>, Chong Yidong <cyd@stupidchicken.com>,
> 	emacs-devel@gnu.org
> 
> Eli, could you explain what is the remaining problem with his
> change, then?

I will be able to explain that when I actually make the DOS port
working again.  Everyone else is telling me that there are no
problems, but no one of them is actually familiar with how the DOS
port works, so I still have my doubts.

Re: Removing MULTI_KBOARD

[Stefan Monnier]

>> Eli, could you explain what is the remaining problem with his
>> change, then?

> I will be able to explain that when I actually make the DOS port
> working again.  Everyone else is telling me that there are no
> problems, but no one of them is actually familiar with how the DOS
> port works, so I still have my doubts.

OK.  So I hereby grant you a voucher for 32KB of complaints about the
MULTI_KBOARD merge, to be redeemed when the DOS port is revived,


        Stefan "Sorry, the vacation makes me a bit lighthearted"

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> From: Stefan Monnier <monnier@iro.umontreal.ca>
> Cc: dann@ics.uci.edu,  nickrob@snap.net.nz,  cyd@stupidchicken.com,  emacs-devel@gnu.org
> Date: Thu, 07 Aug 2008 15:37:22 -0400
> 
> >> Eli, could you explain what is the remaining problem with his
> >> change, then?
> 
> > I will be able to explain that when I actually make the DOS port
> > working again.  Everyone else is telling me that there are no
> > problems, but no one of them is actually familiar with how the DOS
> > port works, so I still have my doubts.
> 
> OK.  So I hereby grant you a voucher for 32KB of complaints about the
> MULTI_KBOARD merge, to be redeemed when the DOS port is revived,

Beware: I will hold you to your word.

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> Date: Tue, 12 Aug 2008 21:08:29 +0300
> From: Eli Zaretskii <eliz@gnu.org>
> Cc: nickrob@snap.net.nz, dann@ics.uci.edu, cyd@stupidchicken.com,
> 	emacs-devel@gnu.org
> 
> > From: Stefan Monnier <monnier@iro.umontreal.ca>
> > Cc: dann@ics.uci.edu,  nickrob@snap.net.nz,  cyd@stupidchicken.com,  emacs-devel@gnu.org
> > Date: Thu, 07 Aug 2008 15:37:22 -0400
> > 
> > >> Eli, could you explain what is the remaining problem with his
> > >> change, then?
> > 
> > > I will be able to explain that when I actually make the DOS port
> > > working again.  Everyone else is telling me that there are no
> > > problems, but no one of them is actually familiar with how the DOS
> > > port works, so I still have my doubts.
> > 
> > OK.  So I hereby grant you a voucher for 32KB of complaints about the
> > MULTI_KBOARD merge, to be redeemed when the DOS port is revived,
> 
> Beware: I will hold you to your word.

Well, it turned out to be much worse than my worst fears, but I did it
anyway: the MS-DOS build of Emacs 23 is operational again.  There are
still a few problems I need to take care of, but it compiles, links,
dumps itself, and comes up as a useful editor.

Please don't ask for more details, or I will have hard time resisting
the temptation to stop being civilized and say everything I think
about the code changes that broke the DOS port so hard for such
relatively small changes in functionality.  And I don't mean just the
MULTI_KBOARD thing.

P.S.  That's just 548 bytes out of the 32KB I'm entitled to.  I
reserve the rest for a rainy day.

Re: Removing MULTI_KBOARD

[Dan Nicolaescu]

Eli Zaretskii <eliz@gnu.org> writes:

  > > Date: Tue, 12 Aug 2008 21:08:29 +0300
  > > From: Eli Zaretskii <eliz@gnu.org>
  > > Cc: nickrob@snap.net.nz, dann@ics.uci.edu, cyd@stupidchicken.com,
  > > 	emacs-devel@gnu.org
  > > 
  > > > From: Stefan Monnier <monnier@iro.umontreal.ca>
  > > > Cc: dann@ics.uci.edu,  nickrob@snap.net.nz,  cyd@stupidchicken.com,  emacs-devel@gnu.org
  > > > Date: Thu, 07 Aug 2008 15:37:22 -0400
  > > > 
  > > > >> Eli, could you explain what is the remaining problem with his
  > > > >> change, then?
  > > > 
  > > > > I will be able to explain that when I actually make the DOS port
  > > > > working again.  Everyone else is telling me that there are no
  > > > > problems, but no one of them is actually familiar with how the DOS
  > > > > port works, so I still have my doubts.
  > > > 
  > > > OK.  So I hereby grant you a voucher for 32KB of complaints about the
  > > > MULTI_KBOARD merge, to be redeemed when the DOS port is revived,
  > > 
  > > Beware: I will hold you to your word.
  > 
  > Well, it turned out to be much worse than my worst fears, but I did it
  > anyway: the MS-DOS build of Emacs 23 is operational again.  There are
  > still a few problems I need to take care of, but it compiles, links,
  > dumps itself, and comes up as a useful editor.
  > 
  > Please don't ask for more details, or I will have hard time resisting
  > the temptation to stop being civilized and say everything I think
  > about the code changes that broke the DOS port so hard for such
  > relatively small changes in functionality.  And I don't mean just the
  > MULTI_KBOARD thing.

Was the removed MULTI_KBOARD code actually needed?

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> From: Dan Nicolaescu <dann@ics.uci.edu>
> Cc: monnier@iro.umontreal.ca, emacs-devel@gnu.org
> Date: Sat, 23 Aug 2008 11:24:38 -0700
> 
> Was the removed MULTI_KBOARD code actually needed?

I don't know.  I debugged the current code, and never looked back,
because I knew it was pointless, and because in my experience
comparing with older versions is not an efficient way of debugging a
program.  It's all in the logs and in CVS, if you really want to know.

(Actually, the right question is whether the _additional_ code exposed
by removing the MULTI_KBOARD did any harm to the MSDOS build.)

Re: Removing MULTI_KBOARD

[Dan Nicolaescu]

Eli Zaretskii <eliz@gnu.org> writes:

  > > From: Dan Nicolaescu <dann@ics.uci.edu>
  > > Cc: monnier@iro.umontreal.ca, emacs-devel@gnu.org
  > > Date: Sat, 23 Aug 2008 11:24:38 -0700
  > > 
  > > Was the removed MULTI_KBOARD code actually needed?
  > 
  > I don't know.  I debugged the current code, and never looked back,
  > because I knew it was pointless, and because in my experience
  > comparing with older versions is not an efficient way of debugging a
  > program.  It's all in the logs and in CVS, if you really want to know.

So you are in effect saying that you preemptive attack was baseless and
without any technical justification (the HR approach was so wrong that's
not even worth disscussing).  Refusing to provide a technical
justification after so many requests is quite telling.  Thank you.

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> From: Dan Nicolaescu <dann@ics.uci.edu>
> Cc: monnier@iro.umontreal.ca, emacs-devel@gnu.org
> Date: Sat, 23 Aug 2008 12:13:24 -0700
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
>   > > From: Dan Nicolaescu <dann@ics.uci.edu>
>   > > Cc: monnier@iro.umontreal.ca, emacs-devel@gnu.org
>   > > Date: Sat, 23 Aug 2008 11:24:38 -0700
>   > > 
>   > > Was the removed MULTI_KBOARD code actually needed?
>   > 
>   > I don't know.  I debugged the current code, and never looked back,
>   > because I knew it was pointless, and because in my experience
>   > comparing with older versions is not an efficient way of debugging a
>   > program.  It's all in the logs and in CVS, if you really want to know.
> 
> So you are in effect saying that you preemptive attack was baseless and
> without any technical justification (the HR approach was so wrong that's
> not even worth disscussing).  Refusing to provide a technical
> justification after so many requests is quite telling.  Thank you.

Look, Dan, it's quite clear that you want to pick up this fight right
where you left off.  But I'm done fighting.  The code and the logs are
there for everyone to see and study, and they speak for themselves.
They are the technical justification you are asking for, as far as I'm
concerned.  Whatever conclusions you draw from that is your own
business and responsibility.

Re: Removing MULTI_KBOARD

[Chong Yidong]

Eli Zaretskii <eliz@gnu.org> writes:

> Well, it turned out to be much worse than my worst fears, but I did it
> anyway: the MS-DOS build of Emacs 23 is operational again.  There are
> still a few problems I need to take care of, but it compiles, links,
> dumps itself, and comes up as a useful editor.

Amazing.  Thanks.

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> From: Chong Yidong <cyd@stupidchicken.com>
> Date: Sat, 23 Aug 2008 16:18:41 -0400
> Cc: monnier@iro.umontreal.ca, emacs-devel@gnu.org
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> > Well, it turned out to be much worse than my worst fears, but I did it
> > anyway: the MS-DOS build of Emacs 23 is operational again.  There are
> > still a few problems I need to take care of, but it compiles, links,
> > dumps itself, and comes up as a useful editor.
> 
> Amazing.  Thanks.

You are welcome.

Re: Removing MULTI_KBOARD

[Dan Nicolaescu]

Eli Zaretskii <eliz@gnu.org> writes:

  > > From: Dan Nicolaescu <dann@ics.uci.edu>
  > > Cc: monnier@iro.umontreal.ca, emacs-devel@gnu.org
  > > Date: Sat, 23 Aug 2008 12:13:24 -0700
  > > 
  > > Eli Zaretskii <eliz@gnu.org> writes:
  > > 
  > >   > > From: Dan Nicolaescu <dann@ics.uci.edu>
  > >   > > Cc: monnier@iro.umontreal.ca, emacs-devel@gnu.org
  > >   > > Date: Sat, 23 Aug 2008 11:24:38 -0700
  > >   > > 
  > >   > > Was the removed MULTI_KBOARD code actually needed?
  > >   > 
  > >   > I don't know.  I debugged the current code, and never looked back,
  > >   > because I knew it was pointless, and because in my experience
  > >   > comparing with older versions is not an efficient way of debugging a
  > >   > program.  It's all in the logs and in CVS, if you really want to know.
  > > 
  > > So you are in effect saying that you preemptive attack was baseless and
  > > without any technical justification (the HR approach was so wrong that's
  > > not even worth disscussing).  Refusing to provide a technical
  > > justification after so many requests is quite telling.  Thank you.
  > 
  > Look, Dan, it's quite clear that you want to pick up this fight right
  > where you left off.  

Unfounded accusations as usual.  I was just stating facts.

  > But I'm done fighting.  The code and the logs are there for everyone
  > to see and study, and they speak for themselves.  They are the
  > technical justification you are asking for, as far as I'm concerned.
  > Whatever conclusions you draw from that is your own business and
  > responsibility.

You are just repeating yourself.  Still no facts.  But yes, I am done
talking about this, and with you in general because it's pointless.

Re: Removing MULTI_KBOARD

[Richard M. Stallman]

    So you are in effect saying that you preemptive attack was baseless and
    without any technical justification (the HR approach was so wrong that's
    not even worth disscussing).  Refusing to provide a technical
    justification after so many requests is quite telling.  Thank you.

It was entirely reasonable for Eli object to a change that looked like
it would make things more difficult for him -- he did not have to try
to actually do the work before he could object to this.

The fact that he ultimately went along with the change, rather than
undoing it, argues that removing the no-MULTI_KBOARD code was a good
change.  But it doesn't change the conclusion that you should have
asked.

Re: Removing MULTI_KBOARD

[Dan Nicolaescu]

"Richard M. Stallman" <rms@gnu.org> writes:

  >     So you are in effect saying that you preemptive attack was baseless and
  >     without any technical justification (the HR approach was so wrong that's
  >     not even worth disscussing).  Refusing to provide a technical
  >     justification after so many requests is quite telling.  Thank you.
  > 
  > It was entirely reasonable for Eli object to a change that looked like
  > it would make things more difficult for him -- he did not have to try
  > to actually do the work before he could object to this.

It would have been reasonable to object, but what he did cannot be
called object.  Had he just been objecting, this issue would have been
easily solved.  It was much worse than that, and it was totally
unfounded too.  At least 3 people told him that at the time.

  > The fact that he ultimately went along with the change, rather than
  > undoing it, 

That what he wants us to believe, but it is not the reality.  After
reviewing his changes I can confirm what I have been saying from day 1:
there's no relation between fixing the MSDOS port and removing the
non-MULTI_KBOARD support.

  > argues that removing the no-MULTI_KBOARD code was a good
  > change.  But it doesn't change the conclusion that you should have
  > asked.

I wouldn't put too much base on a conclusion based on a false premise.

Now, can we please drop this topic?  It has been a terrible waste of
time.  Putting so much energy into discussing a platform which, given
the evidence we have has no users, it is close to being dead, and it is
of no strategic importance to GNU does not seem like a good idea.

Re: Removing MULTI_KBOARD

[Juanma Barranquero]

On Sun, Aug 24, 2008 at 11:36, Dan Nicolaescu <dann@ics.uci.edu> wrote:

> I wouldn't put too much base on a conclusion based on a false premise.

That you should have asked is not a conclusion of any premise. Anyone
who does wide-scale changes should at least warn beforehand.

   Juanma

Re: Removing MULTI_KBOARD

[Eli Zaretskii]

> From: "Richard M. Stallman" <rms@gnu.org>
> CC: eliz@gnu.org, monnier@iro.umontreal.ca, emacs-devel@gnu.org
> Date: Sun, 24 Aug 2008 01:34:54 -0400
> 
> The fact that he ultimately went along with the change, rather than
> undoing it, argues that removing the no-MULTI_KBOARD code was a good
> change.

Actually, the main reason is that I have a great deal of respect for
other people's work, and would not consider undoing it except as the
very last resort.

When I said I didn't know whether MULTI_KBOARD had a significant part
in having me invest many-many hours in bringing the DOS port back to
life, I really meant it.  There were two large merges to struggle
with -- unicode-2 and multi-tty -- and many smaller changes, one of
them MULTI_KBOARD.  Trying to understand how many percents each one
contributed to the plight would be a complete waste of my time.

Re: Removing MULTI_KBOARD

[Stefan Monnier]

>> > Was the removed MULTI_KBOARD code actually needed?
>> I don't know.  I debugged the current code, and never looked back,
>> because I knew it was pointless, and because in my experience
>> comparing with older versions is not an efficient way of debugging a
>> program.  It's all in the logs and in CVS, if you really want to know.
> So you are in effect saying that you preemptive attack was baseless and
> without any technical justification (the HR approach was so wrong that's
> not even worth disscussing).  Refusing to provide a technical
> justification after so many requests is quite telling.  Thank you.

Let's please stop this silly thread right here right now,


        Stefan

Re: Removing MULTI_KBOARD

[Stefan Monnier]

> Well, it turned out to be much worse than my worst fears, but I did it
> anyway: the MS-DOS build of Emacs 23 is operational again.  There are
> still a few problems I need to take care of, but it compiles, links,
> dumps itself, and comes up as a useful editor.

Thank you very much, Eli.

> P.S.  That's just 548 bytes out of the 32KB I'm entitled to.
> I reserve the rest for a rainy day.

I count 647 here (or even 1554 with the quoted material), but I'll let
it slide,


        Stefan