Bug#379292: xkb-data hr layout broke the normal AltGr+numbers behaviour

[Denis Barbier <barbier@linuxfr.org>]

On Sun, Aug 27, 2006 at 04:49:13PM +0200, Josip Rodin wrote:
[...]
> BTW, a note on organization of files. It seems to me that the upstream went
> generalizing, and created a new file "cs" (country code for Serbia and
> Montenegro), and then went on to base ba, hr and si on it because the
> layouts are mostly equal.

Absolutely, this has been discussed last year, mostly in this thread
  http://listserv.bat.ru/xkb/Message/708.html

> While such a condensation was probably a decent idea since it makes the
> files more logical, it was probably the root of this whole problem.
> Since cs(cyralpha) includes its own version of level3 which is more like
> English, this change propagated into hr, si, ba so that they now started
> to have AltGr'ed keys like on an English keyboard, which is incorrect.

Right, it looks like we did not check carefully enough and some unwanted
changes have been introduced.

> Another problem cropped up earlier this year, when the state union of
> Serbia and Montenegro ceased to exist. This made their provisional country
> code (cs) defunct, too. So the file name will need to be changed anyway.
> 
> The commonality in keyboards between Slovenian, Croatian, Bosnian and
> Serbian (Latin) alphabets can be grouped under the linguistically recognized
> grouping called the Western branch of the South Slavic languages. Perhaps
> the common file should be called "wsslavic"?

Ok, I will send your suggestion to upstream when the cs layout is renamed.

> On related note, the optimization/generalization should also encompass
> cz(basic) and sk(basic) which differ in only six keys, and could be
> similarly merged.

Thanks for the tip, I will forward it upstream.

Denis