@ncafferkey:
You're welcome.

Most probably my memory acting up again.... I can scan my note-block on which i wrote it down if you want though

@magorium

Thanks for checking, although I'm not sure what you mean about "41.1"

@ncafferkey:
Yups, i was also not sure. Now i am

input.device v41.3 (although i wrote down 41.1, to copy from the initial post ;-p)

The sequence as mentioned produced a small o and capital O respectively (without accent).

Thanks for the sequence, as i wondered about that.

There was a system-wide problem with dead keys on USB keyboards, which was fixed just before Icaros 2.0 was released. I'm not sure if the fix made it into 2.0 (I'm still using a pre-release version here).

Could someone please check the version of input.device? The fixed version is 41.4.

BTW, a French keyboard isn't required to check for this bug: just switch your layout to French in Input prefs, and type '[' on English keyboards, then the 'O' key. You should get '?'.

Thank you very much for the explanation ball000 :thumbsup:

At least i was now able to make sense of the special declaration in the keymap source for that key... however, it seems the French keymap declaration is too difficult for me to grasp in relation to what could be wrong.

Is there a possibility that people use keyboards that is not compatible to the pc-105 keycodes and by that have a different keycode for that 'special' dead key ?

Regarding windows... could that be caused because the host-os (keyboard driver) interferes with the special key, intercepts it and make it's own interpretation so that it confuses AROS ? (just spilling my thoughts).

The last improvement made on French keymap was only done on ABI_v1 so far. It has only one purpose: allow to send the "char 28" (CTRL-\) which is recognised by the console window (Shell) as a terminating sequence, and thus closes the window. The problem is that with the French Keyboard, the '\' character can only be typed using the "ALTGR-8" sequence. So on French keyboard, the "char 28" is in fact done with the "CTRL-ALTGR-8" sequence.
Now there is another new problem on mingw32-hosted, as Windows sends a "CTRL-ALT" sequence when the ALTGR key is pressed, instead of just a "ALT"... So on this system, typing just "ALTGR-8" does actually close the console window, which is wrong indeed. I don't know how to correct the sequence sent by Windows.

Now I have just tried the use of the ^/? dead key (magorium: this dead key is hit alone on the French keyboard, or with SHIFT for the '?', but without ALTGR. It changes the behaviour of the next key if it's a vowel or space, mainly) and it behaves correctly here. I never went in the problem described by oan3 and noppatoppa, and I don't know how to reproduce it, it just works, on all the applications I could try on a nightly build or on Icaros 2.0 running on QEMU... But on those sessions I never had to tweak Poseidon at all, I didn't even start it.

Hi oan3,

hmz that's weird. i always assumed the alt-gr key was used to used to input accented characters.

The last PS remark you made, let me think there is still a dead-key not recognized properly (e.g. recognized as normal key, and loses it 'accent' state because of it).

I have no idea if that's still a keymap related issue. Unfortunately i do not have a french or alt-gr keyboard myself atm.

PS: did you at least managed to discover the 'life' mode so that you can see your keys typed life in action with poseidon. ?

fwiw: I also have trouble understanding the advanced poseidon options (not to mention some options not being saved for re-use, or i must be doing that wrong as well).

hi all,

AKAIK last improvements did not cure the problem with some accentuations like ? and ? (the same for a, e, i...), like if the key itself isn't recognised.

To explain the way this key is used : we first type it, then the system waits for a second key to be typed to add the accent on the key. For example : we type ^ then after we type a, and we get ? on the screen.

I tried using poseidon like it was adviced to me, but i couldn't find a way to get through this problem.
Anyone can help me understanding how we can manage to get the "typed code" of a key and set it in poseidon parameters ?

Thanx,
Jess.
PS : this is not related to OWB as every apps behave the same.

Hi noppatoppa,

afaik recently some improvements were made in keymap for french keyboards. Whether or not that made it into Icaros 2.0 release i am not sure. Maybe a fellow french-keyboard user could verify.

Besides that (and again as far as my knowledge goes) there are plans to further update OWB.

Hi!

Sorry to bring up an old thread, but Googling brought me here. I'm experiencing same problems as oan3 with Icaros 2.0 and OWB 1.23. Elsewhere in the OS accented characters (such as a with dots and o with dots, kinda like M_o_torhead) work just fine, but in HTML text fields I get nothing when pressing the keys. Anywhere I should go looking for the solution and / or can I help? Not sure how, but still :)

Tried UTF-8 and ISO8859-15, but neither work.