First Midas Tests...

[Christian Stocker]

=2E.. look quite promising ;)

I was able to insert arbitary elements and intercept/override
key/mouseevents, so everything I had doubts about, seems to be possible.
Therefore I will give it a try for the BXE NG and try to integrate it.
This will save us certainly a lot of hassle with the keyevents handling
and hopefully make it faster. Furthermore, the big =D6, resp. dead keys do
also work with midas. I couldn't get them working with JS keypress events
(http://bugzilla.mozilla.org/show_bug.cgi?id=3D192935 for the record).

Concerning Mozilla < 1.3 /Netscape 7, I think, it should be possible to
still support them. We'll use Midas mainly for the typing stuff, not the
formatting stuff, so it should be just a matter of additional keyevents in
Mozilla <1.3 as it is now (of course, it's much more debugging/testing
work, but as long as Mozilla 1.3 isn't that widespread (it's not even out
yet..) it's well worth the effort.)

chregu



--=20
christian stocker | bitflux GmbH | sch=F6neggstrasse 5 | ch-8004 zurich
phone [phone] | mobile [phone] | fax [phone]
http://www.bitflux.ch | ***@bitflux.ch | gnupg-keyid 0x5CE1DECB