freebsd stable - Aug 2014 - TESTING required: keyboard maps for NEWCONS (committed to -CURRENT and available for -STABLE)

I have just committed keymap definitions that are converted to work with
NEWCONS. Since the new console expects a Unicode locale, all
the different encodings are no longer needed - but all files had to
be converted from their respective encodings to Unicode.

Since the character encoding is no longer required in the file
names, all names where converted to be named by ISO country code
(plus ISO language code, where required) followed by modifiers
(e.g. .acc for accent keys).


It is likely, that I introduced regressions, even though I spent
quite a few hours on the conversion process and the manual
verification of the results.


These files are currently only committed to -CURRENT and I plan
to MFC to 10-STABLE as soon as possible (and permitted), which is
in 3 days.

If you are able to test the keymap files under -CURRENT, I'd love
to receive your feedback and remarks. Please do not commit any
fixes directly to -CURRENT, since I'm still working on these files
(I'm afraid of merge conflicts and higher effort required for the
MFC in a few days ...).


If you are a (potential) NEWCONS user on -STABLE, you may want to
extract the contents of

	http://people.freebsd.org/~se/vt-keymaps.tar.bz2

into /usr/share/vt/keymaps. (The Makefile is not required, if you
directly extract to that directory, but it does no harm.)


We really want to have verified keymaps for NEWCONS in 10-STABLE
and the more "special" your language (i.e. the farther it is away
from ISO8859-1 ;-) ) the bigger the risk that some specialty has
been lost during the conversion.

Please test and report your expected deadkeys vs. nodeadkeys
experience, the combinations of Shift, Ctrl, Alt, Alt-Gr and
all keys that you expect to support such modifiers, and the
Caps Lock behavior (on keys that respect or ignore Caps Lock).

I'll be reserving a few hours each day to get this into good
shape for 10.1.

Best regards, STefan

On Sun, 17 Aug 2014 22:36:24 +0200 Stefan Esser wrote:
> I have just committed keymap definitions that are converted to work with
> NEWCONS. Since the new console expects a Unicode locale, all
> the different encodings are no longer needed - but all files had to
> be converted from their respective encodings to Unicode.
Slovenian (si) and Kroatian (hr) are ISO-8859-2 or ISO-8859-16 but were
converted as if they were ISO-8859-1.  Alt+l should be ? for instance.

Thanks for your work, Stefan!

Please consider modifying scan code 125 in jp.kbd and jp.capsctrl.kbd.
Changing '\' to 165 (unicode XML entity ¥ is YEN SIGN) matches
major keyboards in Japan (would be OADG 109A or its variants) with
unicode locales.

Please see below for key layouts.

  http://ja.wikipedia.org/wiki/JIS%E3%82%AD%E3%83%BC%E3%83%9C%E3%83%BC%E3%83%89

This page is in Japanese, but figures here would be useful.
 
*Historical background:
 For historical reason, starting from JIS X 0201, code point 0x5C is
 used for YEN SIGN (currency JPY) in Japan. This code point may be
 used for backslash in other countries. (So, YEN SIGN should not be
 supported in syscons.)

 But now in unicode, YEN SIGN has its own code point, and safely
 coexists with backslash.

 So, enabling both in vt would be nice. And particularly for users
 having both FreeBSD and Windoze, matching keytop and displayed
 character should be important to avoid confusion.


CC'ed to freebsd-users-jp list for other Japanese users.
Any objections? Possibly, someone would want modifying scan code 115
instead of 125.


On Sun, 17 Aug 2014 22:36:24 +0200
Stefan Esser <se at freebsd.org> wrote:
> I have just committed keymap definitions that are converted to work with
> NEWCONS. Since the new console expects a Unicode locale, all
> the different encodings are no longer needed - but all files had to
> be converted from their respective encodings to Unicode.
> 
> Since the character encoding is no longer required in the file
> names, all names where converted to be named by ISO country code
> (plus ISO language code, where required) followed by modifiers
> (e.g. .acc for accent keys).
> 
> 
> It is likely, that I introduced regressions, even though I spent
> quite a few hours on the conversion process and the manual
> verification of the results.
> 
> 
> These files are currently only committed to -CURRENT and I plan
> to MFC to 10-STABLE as soon as possible (and permitted), which is
> in 3 days.
> 
> If you are able to test the keymap files under -CURRENT, I'd love
> to receive your feedback and remarks. Please do not commit any
> fixes directly to -CURRENT, since I'm still working on these files
> (I'm afraid of merge conflicts and higher effort required for the
> MFC in a few days ...).
> 
> 
> If you are a (potential) NEWCONS user on -STABLE, you may want to
> extract the contents of
> 
> 	http://people.freebsd.org/~se/vt-keymaps.tar.bz2
> 
> into /usr/share/vt/keymaps. (The Makefile is not required, if you
> directly extract to that directory, but it does no harm.)
> 
> 
> We really want to have verified keymaps for NEWCONS in 10-STABLE
> and the more "special" your language (i.e. the farther it is away
> from ISO8859-1 ;-) ) the bigger the risk that some specialty has
> been lost during the conversion.
> 
> Please test and report your expected deadkeys vs. nodeadkeys
> experience, the combinations of Shift, Ctrl, Alt, Alt-Gr and
> all keys that you expect to support such modifiers, and the
> Caps Lock behavior (on keys that respect or ignore Caps Lock).
> 
> I'll be reserving a few hours each day to get this into good
> shape for 10.1.
> 
> Best regards, STefan
> _______________________________________________
> freebsd-stable at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-stable
> To unsubscribe, send any mail to "freebsd-stable-unsubscribe at
freebsd.org"
> 

-- 
Tomoaki AOKI    junchoon at dec.sakura.ne.jp