Jean-Marc Lasgouttes wrote:
> Two remarks:
>
> 1/ testing for Qt>=3.0 is
> #if QT_VERSION >= 300
> The semantics changed in 3.1

Yes, thanks. Actually, it changed with 3.0.5

> 2/ It would actually be simpler to enable the code only for Qt >= 3.2.

Agreed. I changed all version checks to #if QT_VERSION >= 0x030200

> Does this patch have an interest for Qt 3.1?

I don't think so. All qt-immodule patches are against Qt 3.3.x at least.

> Concerning the math superscript: is it because it is defined as a dead
> key? (on a french kbd we have both the dead ^ and the normal ^). I
> guess in the case of a dead key, it is ^<space> which is supposed to
> trigger a superscript.

I'm not really sure what the reason is. With the special code, the ^-handling 
inside mathed is as expected: whenever ^+Key makes some sense, the 
correspondig character is inserted (i.e. ^+e gives � even inside mathed). 
^+<space> triggers the superscript. Same as in xforms here.

Without the special code, it is impossible to trigger a superscript with the 
^-key, but "m-x accent-circumflex" works. Mathed has some special check for 
LFUN_CIRCUMFLEX to get the superscript. Seems like qt is returning some other 
key (maybe the second ^ you are talking about).

> I am not sure whether we want to add such
> special code.

I'm afraid I don't have an alternative solution.

J�rgen

Reply via email to