ADB German Keyboard layout ^ and < reversed (Windows) #35

DOCaCola · 2013-05-24T13:07:22Z

This is more or less a follow up on the older bug #22 that was fixed in the adb_usb converter.
I have now noticed something else. While these keys work as expected on my mac, They are reversed when the keyboard is plugged in on Windows (M3501 Extended II keyboard).
So when i press the ^ key (left of the 1 key), i get <.
If i press <, i get ^.
I thought it to be some kind of issue with Windows, so i googled and found this site (it is german though), but it might be helpful in analyzing this issue.
He describes some low level patch for the keyboard driver of Mac OS X.
http://snark.de/mac/usbkbpatch/
As it is already in the adb_usb code, he describes the adb keyboard codes as
0x32 <
0x0a ^

but he has an interesting note on the usb codes for these keys:
0x64 ^ (caret - on wintel-usb switched with < (smaller) )
0x35 < (smaller - on wintel-usb switched with ^ (caret) )

Not sure what this means though. But there seems some kind of difference how Mac OS X interpretes these keys.

Switching the ADB code in the keymap.c obviously did the trick, but i guess a better solution exists, since other keyboards do not have this issue.

tmk · 2013-05-24T15:36:56Z

Hmm,
How does windows keyboard work on Mac? How does apple keyboard work on Windows(Not bootcamp)? Those keys work correctlly?

Actually I want to know what codes German keyboard sends for those keys. Though that'll be difficult or need some hassle...

DOCaCola · 2013-05-24T16:35:30Z

yes, a windows keyboard (logitech) connected to my macbook -> keys work correct
apple aluminium keyboard connected to my windows pc -> keys also work correct. no special drivers loaded

if you explain what is required, i will try to help

DOCaCola · 2013-05-24T16:57:46Z

i have loaded an usb monitor software on my windows machine
data transmitted from the apple keyboard AND logitech keyboard are both:
<
00 00 64 00 00 00 00 00

and
^
00 00 35 00 00 00 00 00

as expected, when having the adb_usb teensy connected, it behaves the other way round...
now it would be interesting to get this data on mac...

edit: i think i found the issue.....actually i was...
When you plug in a new keyboard to mac, the keyboard setup assistant appears and asks you to press the key right from the left shift key...however i never saw that dialog on my mac..
I just found another site where someone describes this very issue. So if i press the caret key (which is actually the < in adb_usb) the behaviour is switched (like it is wrong now under windows).

Those two keys should be interchanged in adb_usb. Sorry for reporting it wrong in the first place.

The keyboard layout can be reset by deleting the /Library/Preferences/com.apple.keyboardtype.plist file. Then restart. Mac OS X now brings up the keyboard assistant again. With the correct adb_usb keyboard layout, after using this assistant, the keys should also work correct then.

tmk · 2013-05-25T06:35:03Z

I seemed to be confused about those keys, if I can get them correctlly they should be like following table, right?
ADB ISO(DE) keyboard has different code0A) for key left of 1 from US layout(32). This is very confusing.
I think we should have another KEYMAP macro for ISO in keymap.c.

                                        Code
                    -----------------------------------------------
Key Position        ADB-US      ADB-DE(ISO)        USB
-------------------------------------------------------------------
Left of 1           0x32        0x0A               0x35 Grave
Right of LShift     NONE        0x32               0x64 Non-US \

DOCaCola · 2013-05-27T09:21:44Z

i am really 100% sure. As it is now in the keycode.c file it is not correct for the german keyboard. It will work under OS X, but it will turn out wrong on Windows etc.

tmk · 2014-04-30T23:45:31Z

Looks like two solutions can exist.

adb_usb needs ISO keymap macro. US and ISO layout are incompatible due to two keys above.
Or if we are able to know layout(US, ISO or JIS?) with ADB command firmware can handle the two keys especially and we can have one universal keymap. I don't know this command really exists.

French AZERTY has same problem.
http://geekhack.org/index.php?topic=14290.msg1312242#msg1312242

tmk · 2014-05-01T02:46:19Z

OSX can recognize its layout(ANSI, ISO or JIS) automatically for Apple products only so that faithful users doens't need 'Keyboard Setup Assistant' to start using it. Meanwhile third party products including TMK are handled as ANSI layout by default unless user specified its layout in some way.

OSX checks only Apple USB vendor/product ID.(and ADB product ID, probably.) See IOHIDKeyboard::handlerID.
http://opensource.apple.com/source/IOHIDFamily/IOHIDFamily-503.1.13/IOHIDFamily/IOHIDKeyboard.cpp

OSX uses ADB scan codes as internal key codes and USB codes(usages) should be converted into ADB codes. This file defines USB-ADB code convert table.
http://opensource.apple.com/source/IOHIDFamily/IOHIDFamily-503.1.13/IOHIDFamily/Cosmo_USB2ADB.c

OSX checks only Apple USB vendor/product ID and ADB product ID, probably.(See IOHIDKeyboard::handlerID) And the table is patched when Apple ISO keyboard is using. This circumvents the key swap problem.(See IOHIDKeyboard::deviceType)
http://opensource.apple.com/source/IOHIDFamily/IOHIDFamily-503.1.13/IOHIDFamily/IOHIDKeyboard.cpp

But our keyboard is handled as ANSI layout so we suffer from the key swap problem.

[DELETE]
YOU NEED TO SPECIFY LAYOUT UNLESS YOUR KEYBOARD IS APPLE PRODUCT OR ANSI.
Maybe you can use Keyboard Setup Assistant. This itself is not problem at all.
[/DELETE]

Problem is USB-ADB convert table to support both ISO and ANSI ADB keyboard layouts which are incompatile with eatch other. ANSI uses ADB code 0x32 to register '^'(left of '1') while ISO uses 0x32 for '<'(right of Left Shift) and 0x0A for '^'. How iMate resolves this? Maybe solution 2.

Methinks. Just two cents.

tmk · 2014-05-20T17:52:16Z

Use ISO keymap with make KEYMAP=iso to circumvent this problem.

ADB ISO keyboard has weird layout which swaps key on left of 1(_a) and key on right of Shift(_b). We need to use keymap for this ISO layout.

ADB ANSI/ISO Keymapping Note
============================
ANSI
,-----------    ----------.
| *a|  1|  2     =|Backspa|
|-----------    ----------|
|Tab  |  Q|     |  ]|   *c|
|-----------    ----------|
|CapsLo|  A|    '|Return  |
|-----------    ----------|
|Shift   |      Shift     |
`-----------    ----------'

ISO
,-----------    ----------.
| *a|  1|  2     =|Backspa|
|-----------    ----------|
|Tab  |  Q|     |  ]|Retur|
|-----------    -----`    |
|CapsLo|  A|    '| *c|    |
|-----------    ----------|
|Shif| *b|      Shift     |
`-----------    ----------'

ADB Keyboard scan code:
        ADB scan code   USB usage
        -------------   ---------
Key     ANSI    ISO     ANSI    ISO
---------------------------------------------
*a      0x32    0x0A    0x35    0x35
*b      ----    0x32    ----    0x64
*c      0x2A    0x2A    0x31    0x31(or 0x32)


TMK ADB-USB mapping:
ADB     USB(ANSI)   USB(ISO)
---------------------------------
0x32    0x35        0x64
0x0A    ----        0x35
0x2A    0x31        0x31(or 0x32)

* upstream/master: (93 commits) Add ISO keymap(issue tmk#35) and use new keymap API Fix set LED for NKRO keyboard interface Unused endpoint of console OUT is commentout'd Add Makefiles for TMK converters. Fix remotewakeup of PJRC stack(Fix tmk#121) Check if LUFA-git really exists(Fix tmk#118) NKRO is disable when SET_PROTOCOL(boot) Make NKRO deafult when it is available Change MCU setting to ATmega32U2 Fix description of pin usage Add LUFA git repository(release 140302) Move macro definitions from lufa.c to lufa.h Fix 'Shift Parentheses' example Change LUFA build options Return when ready check loop is timeouted(Fix tmk#115) Change PollingIntervalMS to 10ms(Fix tmk#114) New macro: ACTION_BACKLIGHT_LEVEL(level) Show correct function name for default layer set Added Lightsaber FN-layer Lightsaber clean ups ...

tmk · 2016-05-10T01:49:07Z

I found this post, seems 'Griffin iMate' and 'New Motion uKey' also have similar or worse problem with ISO keyboards.
https://geekhack.org/index.php?topic=4185.msg61270#msg61270

tmk · 2016-05-10T02:27:01Z

ISO AEK layout:
https://deskthority.net/w/images/1/18/AEKII-front.jpg
http://maximac.se/wordpress/wp-content/uploads/2014/12/apple-extended-keyboard-ii.jpg

US AEK layout:
https://deskthority.net/w/images/1/1a/Apple_M0115_top.jpg

tmk · 2016-05-19T06:54:02Z

how to determine keyboard layout with handler id.
http://lxr.free-electrons.com/source/drivers/macintosh/adbhid.c?v=4.4#L802

tmk · 2016-05-19T15:54:25Z

Added auto detection of ISO layout and swap of the problematic scan codes.
f43a2df

I hope this works for ISO keyboards.

tmk · 2016-05-31T22:34:45Z

As for this ISO keyboard auto detection test and report from ISO users are required.

Informations like your OS version, keyboard layout setting on OS and etc would be uselful.

scytho · 2017-01-28T16:08:54Z

I don’t know if this is still an issue.

I have an AEKII ISO German and use Pins 4 and 5 of a Teensy 2.0 for Data and Power. I didn’t change anything in the tmk code, so it should expect Data and Power on Pins 0 and 1 (in adb_usb/config.h). It works almost perfect with Linux Mint 17.3. Only »problem«: The equal sign on the Numpad generates an accent (»´«). I’d prefer a dot (».«, as alternative to the comma). < and ^ work as expected.

Same with Windows 8.1.

scytho · 2017-01-28T16:19:25Z

I changed the code to Pins 4 and 5, now < and ^ are switched. I changed it back to 0 and 1, but the Teensy only would go into bootloops and make the mouse pointer jump. After several tries and with Pins 4 and 5 in config.h it’s working again, < and ^ not switched. Weird.

scytho · 2017-01-28T16:31:43Z

After another programming of the Teensy < and ^ were switched again, plugging off and on the AEKII on the Teensy switched them back.

kuerb · 2017-10-23T09:16:23Z

I too have this issue with my German AEKII: NUBS and GRV are switched after programming the micro-controller. Once the keyboard is unplugged they are switched again, to whatever is defined in the keymap.

It seems like the auto detection only works if the keyboard is being programmed.

tmk · 2017-10-24T11:35:03Z

@kuerb yes, you are right. It detects ISO keyboard only once.

Fix and improvement for this is welcomed.

EDIT: ADB conveter checks ISO keyboard only at startup, if you have ISO keyboard you will have to plug it into the converter before pluggin computer.

egla89 · 2017-10-26T17:40:10Z

I have the same problem as a previous commentator where the EQUAL key on the numpad, registers as ‘ on an AEK2 iso. Mapped it as iso

tmk · 2017-10-26T21:53:48Z

@egla89 did you try firmware got from keyboard editor?
what is your keyboard layout on your os?

egla89 · 2017-10-27T05:19:28Z

I flashed a teensy in my iso-de aek2 with ‘make -f Makefile.teensy KEYMAP=iso‘. Os is macOS 10.13 and I ran the layout recognition (so the < and ^ are in the right spot). Every other key works correctly

tmk · 2017-10-27T06:06:54Z

OK, then, ~~key next to backspace(=) spits out equal correctly?~~
what character does key next to backspace spit out?

tmk · 2017-10-27T06:51:13Z

And is this pic what you referred as 'iso-de aek2'?
https://deskthority.net/w/images/1/18/AEKII-front.jpg

if not can you post pic of your aek2?

kuerb · 2017-10-27T10:59:37Z

Here is a picture of my German AEKII, which should be the same as the one egla89 is using.

In my own keymap I switched the numpads = key to PEQL. Maybe that's a simple fix for the general iso keymap?!

tmk · 2017-10-27T11:27:19Z

thanks for the pic and it makes sense.
EQL is recognized as ' on german layout and numpad = key is defined as EQLin keymap_iso.c. PEQL may be less confused, in fact unimap_plain.c uses it.

anyway, you can change keymap as you want, there is no universal keymap for everyone.

egla89 · 2017-10-27T11:42:52Z

Ok yes that makes sense. I‘ll change my layout accordingly. Thanks a lot!

tmk · 2017-10-27T12:10:34Z

Just for future reference, PEQL is not recognized on Windows and which is reason why some keymap use EQL for numpad '=' instead of PEQL.
#238

tmk · 2019-06-21T04:43:06Z

Open new issue if you have problem.

tmk closed this as completed May 25, 2013

tmk reopened this Apr 30, 2014

tmk added the TODO label Apr 30, 2014

tmk closed this as completed May 1, 2014

tmk added the NOTE label May 13, 2014

tmk pushed a commit that referenced this issue May 20, 2014

Add ISO keymap(issue #35) and use new keymap API

ed8cf12

tmk added a commit that referenced this issue Apr 13, 2016

adb_usb: Fix for ISO swapped keys(issue #35)

7613efc

tmk removed the TODO label Apr 13, 2016

jiexi pushed a commit to jiexi/tmk_keyboard that referenced this issue Apr 28, 2016

adb_usb: Fix for ISO swapped keys(issue tmk#35)

af462c5

tmk reopened this May 19, 2016

tmk added the REQUEST_FOR_HELP label May 31, 2016

louisrousseau mentioned this issue Dec 19, 2016

adb_usb/matrix.c : changed size of handler_id from uint16_t to uint8_t #416

Merged

tmk removed the REQUEST_FOR_HELP label Oct 27, 2017

tmk closed this as completed Jun 21, 2019

kls0e mentioned this issue Dec 14, 2019

german Apple Extended Keyboard II on uKey adb-usb adapter (DE-DE) randyrants/sharpkeys#225

Closed

tmk mentioned this issue Feb 4, 2020

[adb2usb] weird Keyboard recognition #640

Closed

siggie0815 mentioned this issue Dec 9, 2020

Update ADB impelemtation in TMK Core qmk/qmk_firmware#11168

Merged

14 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

ADB German Keyboard layout ^ and < reversed (Windows) #35

ADB German Keyboard layout ^ and < reversed (Windows) #35

DOCaCola commented May 24, 2013

tmk commented May 24, 2013

DOCaCola commented May 24, 2013

DOCaCola commented May 24, 2013

tmk commented May 25, 2013

DOCaCola commented May 27, 2013

tmk commented Apr 30, 2014

tmk commented May 1, 2014

tmk commented May 20, 2014

tmk commented May 10, 2016 •

edited

Loading

tmk commented May 10, 2016

tmk commented May 19, 2016

tmk commented May 19, 2016

tmk commented May 31, 2016 •

edited

Loading

scytho commented Jan 28, 2017 •

edited

Loading

scytho commented Jan 28, 2017

scytho commented Jan 28, 2017

kuerb commented Oct 23, 2017

tmk commented Oct 24, 2017 •

edited

Loading

egla89 commented Oct 26, 2017

tmk commented Oct 26, 2017

egla89 commented Oct 27, 2017

tmk commented Oct 27, 2017 •

edited

Loading

tmk commented Oct 27, 2017

kuerb commented Oct 27, 2017

tmk commented Oct 27, 2017

egla89 commented Oct 27, 2017

tmk commented Oct 27, 2017

tmk commented Jun 21, 2019

ADB German Keyboard layout ^ and < reversed (Windows) #35

ADB German Keyboard layout ^ and < reversed (Windows) #35

Comments

DOCaCola commented May 24, 2013

tmk commented May 24, 2013

DOCaCola commented May 24, 2013

DOCaCola commented May 24, 2013

tmk commented May 25, 2013

DOCaCola commented May 27, 2013

tmk commented Apr 30, 2014

tmk commented May 1, 2014

tmk commented May 20, 2014

tmk commented May 10, 2016 • edited Loading

tmk commented May 10, 2016

tmk commented May 19, 2016

tmk commented May 19, 2016

tmk commented May 31, 2016 • edited Loading

scytho commented Jan 28, 2017 • edited Loading

scytho commented Jan 28, 2017

scytho commented Jan 28, 2017

kuerb commented Oct 23, 2017

tmk commented Oct 24, 2017 • edited Loading

EDIT: ADB conveter checks ISO keyboard only at startup, if you have ISO keyboard you will have to plug it into the converter before pluggin computer.

egla89 commented Oct 26, 2017

tmk commented Oct 26, 2017

egla89 commented Oct 27, 2017

tmk commented Oct 27, 2017 • edited Loading

tmk commented Oct 27, 2017

kuerb commented Oct 27, 2017

tmk commented Oct 27, 2017

egla89 commented Oct 27, 2017

tmk commented Oct 27, 2017

tmk commented Jun 21, 2019

tmk commented May 10, 2016 •

edited

Loading

tmk commented May 31, 2016 •

edited

Loading

scytho commented Jan 28, 2017 •

edited

Loading

tmk commented Oct 24, 2017 •

edited

Loading

tmk commented Oct 27, 2017 •

edited

Loading