XT/AT/PS2/Terminal to USB Converter with NKRO

[Krogenar]

Halvar posted a 122-key config that might be a good starting point. It's also in the v1.10 zip as configs\Halvar.sc.

Yes, Halvar's config file got me in the right direction for my 122-key project. Glad it was included in the new documentation!

[poxeclipse]

Thank you Soarer. I'm typing on it now. I used Halvar's configuration file. 
As typing experience, it is identical to the F AT 84-key I have.
This 122-key 6110668 is in very good condition, and I got it for $25. I like it.

[poxeclipse]

I do have another terminal keyboard, a Model M. I kept it as a donor, but it is in such a good shape, almost new, that I might try to mod it with your converter. But, I don't like model m so much. Will see.
The effort of Sourer & co. is great, for so many high quality keyboards can now be used again. So many have been lost, but still lots of them will survive.

[fohat.digs]

After you are spoiled by the Model F, even a Model M is unsatisfying.

I don't understand the appeal of the M-122, though, the only real reason I did the F-122 mod was to get close to the ANSI (Model M) standard layout.

If I could get a capacitive switch Model M, I would look no further.

But, having said that, after a year I have to admit that I have really gotten used to the function keys on the left and 2 Escapes!

[poxeclipse]

If only was a way to make the switches lighter on Model M ! I like the layout and the construction, but they are too heavy for me. I'm still keeping my Model Ms, hoping that will be a way to mod those switches in the near future. 70-80 g is heavy.
When I touched an F AT, I was hooked, for this was exactly what I wanted a Model M to be. And I got it for only $5.
I have now two  F  PC AT and one F terminal.

[Krogenar]

After you are spoiled by the Model F, even a Model M is unsatisfying.

Gah! Stop saying this sort of stuff! My addiction is already too far gone as it is! LOL!

I don't understand the appeal of the M-122, though, the only real reason I did the F-122 mod was to get close to the ANSI (Model M) standard layout.

I love having a lot of keys. Makes me feel important. Also, the thing looks like a tank. Still questing for a Cherry G80-2100 HDU, because I love the layout.

That said, the constant talk of how awesome the Model F is, compared to the Model M is eventually going to make me get one. I just don't really like the vertical Enter. But maybe someone will eventually offer up a modded one.

[Soarer]

Model Fs are far easier to modify than Model Ms... so there's really no excuse not to acquire one... or several!

[poxeclipse]

I just don't really like the vertical Enter. But maybe someone will eventually offer up a modded one.

I had the same preconceived idea about the vertical Enter. Now I'm typing on this 122-key model f and the vertical Enter doesn't bother me at all.
I feel no difference from the ANSI Enter. You should try first.
Model Fs have the identical exterior as the corresponding Model Ms, but they shine at the key switch level.
And it feels good, indeed, with so many keys. But an SSK Model F would be nice too. I don't know if somebody tried to cut a Model F board to a smaller size.

[metalliqaz]

They definitely did.  I'm too lazy to look it up...

[fohat.digs]

I seem to remember an unsaver F, but I wouldn't want to try it.

The ISO-style Enter does not bother me at all - what I cannot abide is the small Backspace. I must make a lot of mistakes, and apparently I always go for the left side of Backspace to fix them.

The plethora of keys is nice, but I still have about a half-dozen that I have not assigned (F15-F19). I really like the function keys on the left.

Also, the incredible weight makes for an amazing keystroke experience, it is only about the keys - the board moving around on the desk is simply not part of the scenario (I do employ a rubber mat, too, of course). And I don't think the effect would be the same with a 1 kg keyboard, even if you glued it down to the desk.

PS - Krogenar - I took your advice and changed out the switches in my Joy mouse knock-off, and it made a big difference! I was pretty nervous, but it went well. I am sure that you can mod an F-122 if you can get your hands on one.

[poxeclipse]

Question for Soarer.
Does your converter work with an IBM 6052101 display writer for 3187 ? The real question is what protocol is used.
Thank you

[Soarer]

No The problem is that they don't send release codes for all but a couple of keys. I started on a project to replace the CPU with a Teensy (but it's stalled with what could be noise problems causing KRO to be poor). I guess the most surefire approach would be to use a reprogrammed 8048 CPU, but that's not trivial and it would still need a converter to USB as well.

[Hubbert]

Is it possible to connect one keyboard to two computers using two Converters?  If so, then what's a good way to disable one converter at a time from sending keystrokes: it only needs to maintain the USB connection.  My quick thought is to use diodes on the power connection and maybe something on the grounds (just hope they're the same since the computers are connected to the same AC outlet?), but the clock & data connections are potential problems.

I have tried USB KVM switches, but the handshaking takes too long, and usually after about 5 swaps my Windows box would stop recognizing the keyboard.  I found a manufacturer of ultra high performance KVM switches for professional audio/visual media uses, but they either didn't list a price or I got the impression that they cost $5000 or so.  I don't need to switch monitors, and I don't really need to switch pointing devices (in a pinch, an extra Marble Mouse trackball doesn't take up too much space).

[kishy]

For information's sake:
My previously reported success with the converter 'surviving' from cold boot all the way through to the Windows 7 desktop on my ThinkPad W520 is no longer the case - one or two firmware updates ago (for the W520 - UEFI/BIOS), it started killing the converter during boot.

This seems to be intermittent/inconsistent though as it sometimes works, but seemingly does not work during POST to enter setup. I haven't done any comprehensive testing, and it could well differ from docking station 2.0 ports to built in 2.0 ports to 3.0 ports...

[Soarer]

Is it possible to connect one keyboard to two computers using two Converters?  If so, then what's a good way to disable one converter at a time from sending keystrokes: it only needs to maintain the USB connection.  My quick thought is to use diodes on the power connection and maybe something on the grounds (just hope they're the same since the computers are connected to the same AC outlet?), but the clock & data connections are potential problems.

I have tried USB KVM switches, but the handshaking takes too long, and usually after about 5 swaps my Windows box would stop recognizing the keyboard.  I found a manufacturer of ultra high performance KVM switches for professional audio/visual media uses, but they either didn't list a price or I got the impression that they cost $5000 or so.  I don't need to switch monitors, and I don't really need to switch pointing devices (in a pinch, an extra Marble Mouse trackball doesn't take up too much space).

That's an interesting idea, and I think it could work. Connecting grounds ought to be fine, and as you say, diodes should be enough to prevent problems on the power line and allow it to be powered from either converter. Switching just clock might be sufficient, but since DPDT switches are just as cheap, you may well switch both data and clock.

The main thing is that each converter will need to talk to the keyboard when it (the converter) powers up, so it knows what kind of keyboard it is. You'll just have to remember to switch it accordingly when powering things up!

That said, it would be an experiment, and I can't say for sure that it will all work

I use an Aten CS1784 4 port KVM that works very well, but of course in its normal mode it can't deal with NKRO or any of the extra interfaces on my converter. It can be set to switch the keyboard as a general USB device, but then that means a delay as the OS processes the new device(s) it sees when switching. It was about $200, so if you don't need to switch monitors, it would be overkill!

[Soarer]

For information's sake:
My previously reported success with the converter 'surviving' from cold boot all the way through to the Windows 7 desktop on my ThinkPad W520 is no longer the case - one or two firmware updates ago (for the W520 - UEFI/BIOS), it started killing the converter during boot.

This seems to be intermittent/inconsistent though as it sometimes works, but seemingly does not work during POST to enter setup. I haven't done any comprehensive testing, and it could well differ from docking station 2.0 ports to built in 2.0 ports to 3.0 ports...

Hmm, not much to go on, and more difficult if it sometimes works, sometimes doesn't

Is that using my latest firmware, or at least the one that has the USB compliance fixes (v1.03)?
(I didn't fix anything that seemed likely to cause such problems though).

[poxeclipse]

Soarer, kishi already pointed out problems with the converter at boot phase, but I still want to ask. It is possible to make the keyboard available at boot time ? This would be desirable for Linux, for instance, when you have to select the image you want to load. I played with it and it did not work. The keyboard is recognised only after the boot.

[Soarer]

It should already work in BIOS and during boot, on pretty much everything. As long as any perfectly standard USB keyboard would, of course. Kishy is just adept at finding problem machines!

I'm not sure exactly what the Linux bootloader does for USB keyboards... I guess it might use a simlar mechanism to BIOS. But if it's a problem with my converter, I'm surprised that nobody has mentioned it before. (On the other hand, if more people have a problem, it will probably make it easier to fix!)

What Linux bootloader and version are you using?
Does the converter work to get into, and in, your BIOS? If not, what's your machine, and what BIOS does it use?

[Soarer]

Just noting a report of a kernel panic in OpenBSD on a Samsung n130 netbook.

[fohat.digs]

I think I have mentioned this before, but, again, this morning, I wired up my "universal" Teensy outboard converter, and it does not recognize a PS2 keyboard.

Having done the Teensy thing half a dozen times, I am pretty good at it. But, with my final Teensy, I want an outboard box with female PS2 connector, so I can use it for "everything" with various adapter plugs.

It is built, continuity checked properly on all 4 leads (the mini-DIN-6 has 6 total - 2 "extras" that I ignored), and programmed using teensy.exe.

The computer acknowledges that the USB device is connected, and the programming routine seemed to complete successfully.

However, when I plug in a PS2 keyboard (I have tried it with a Dell AT101 and an Alps brand board, both of which work properly, otherwise) I get nothing. I have tried rebooting in a couple of different ways, and hot-plugging the USB from the Teensy before and after, all the usual combinations. Also, I learned this before, there is no other Teensy connected, I am doing this after a reboot with a PS2 keyboard in the PS2 connector to the motherboard.

Is there some other trick to this, or what?

I know that it is pointless to use the Teensy when an ordinary adapter would work, but am I missing something here?

At the moment, the 2 XTs (for which I actually want to use this thing) are dismantled, so are not available for alternate testing.

Thanks!

[Soarer]

I can't remember if you have mentioned this problem before! I remember the other outboard, with the real IBM terminal connector, which I think worked in the end...?

The order of connection should always be keyboard to Teensy, then Teensy to computer, so that the Teensy can talk to the keyboard as soon as they are powered up.

Do the LEDs on the keyboards flash just after the Teensy is connected to the computer? That would at least indicate that they were getting power OK.

Does hid_listen say anything when the Teensy is connected to the computer? Or when you press keys?

[fohat.digs]

Thank you for your reply.

I never had a problem with the official Model F terminal connector, it worked great as soon as I got it all hooked up properly.

This is about my 3rd or 4th attempt at this one, I try it, give up, set it aside for a month or 2, and come back.

Going through your steps, I realize that it is completely dead - no LEDs, no hid_listen.

Twice, in recent months, I have had problems with switches and cables, formerly never an issue. Although the continuity checks out, I have a sewing needle soldered to one of my probes, and I am starting to wonder if my female PS2 plug is the problem. Perhaps I can stick the needle farther back into the socket and find a connection, but one or more of the pins on my keyboard cable can't hit it?

Tomorrow, or sometime, I will replace the receptacle pigtail and see if that is it. Otherwise, I am stumped.

The computer sees the Teensy and reports it as operating properly, continuity checks out all around, and the keyboards are OK, so the socket is about the only thing I can't verify.

[Soarer]

Yes, it does sound like a problem with the connector. You'll hate me for saying it, but could it be wired wrongly, with the connections mirrored? It would explain the symptoms.

Do you get some output to hid_listen when plugging the converter in without having any keyboard connected to it?

[fohat.digs]

Can't mess with it until tomorrow, but if I wired it wrong I will feel like a monumental dolt!

[Soarer]

I can't think of anything else that would explain zero output to hid_listen (since you already know to only have one Teensy connected)... but only if hid_listen output is working when no keyboard is attached. The logic being that the keyboard gets reverse power, tries to draw too much current, and the USB port shuts down in response. I could be wrong of course!

[fohat.digs]

I re-worked all my wiring, and the keyboard LEDs do flash when the USB is plugged in, so I know that power is getting through, but that is all.

The computer chimes and reports that the device is connected or disconnected, and is working properly.

hid_listen also reports the device being connected or disconnected, and "listens" for the device for a minute or so with a string of dots marching across the screen.

But it never "hears" any keystrokes from the keyboard.

I have re-programmed and rebooted the Teensy a couple of times, and also rebooted the computer.

And, hid_listen gives the same response without the keyboard plugged into the Teensy as with it.

Got any more ideas?

[Soarer]

I re-worked all my wiring, and the keyboard LEDs do flash when the USB is plugged in, so I know that power is getting through, but that is all.

The computer chimes and reports that the device is connected or disconnected, and is working properly.

hid_listen also reports the device being connected or disconnected, and "listens" for the device for a minute or so with a string of dots marching across the screen.

But it never "hears" any keystrokes from the keyboard.

I have re-programmed and rebooted the Teensy a couple of times, and also rebooted the computer.

And, hid_listen gives the same response without the keyboard plugged into the Teensy as with it.

Got any more ideas?

OK, power is good.

Does hid_listen ever actually say 'Listening:', which comes after the dots have stopped?

Without a keyboard connected, we'd be hoping to see something like this (although without any config loaded it would be slightly different)...

Code: [Select]

Waiting for device:..........................
Listening:
wEE

remaining: 0041
41@0008 remaps !set
layerdefs: 00
max_layer: 00
total_macros: 00
alloc ok.


Keyboard ID: 0000
Code Set: 1
I can't think what would prevent that, if the Teensy was programmed with the converter code correctly (and only that one converter was connected).

Getting it programmed shouldn't need lots of reconnecting, or reboots of the computer. You can run hid_listen, run the Teensy loader, and connect the Teensy (without keyboard attached) - in no particular order. Then just use the reset button and the Teensy loader controls to get it programmed and talking to hid_listen. If it doesn't work, pressing the reset button is usually enough to be able to try again.

[fohat.digs]

Thank you for your patience.

I gave up on that problem, maybe I will revisit it in another month (as I have, multiple times, already). But, to answer that question, after waiting for device, and listening, nothing further happens. If I unplug it, and hot plug it, it waits, and listens, again, to no avail.

Here is my new problem with my F-122 and Teensy. For no reason, (well, the only thing I did was to):

hot plug a iMate ADB-to-USB converter and working AEK2 to the USB port on the front of the computer, for testing, with the F-122/Teensy  running as usual. The Apple tested good and that was that. I unplugged it.

Now, I have lost my numpad, just as if I had pressed the "Numlock" button, which has not even existed for a year. I rebooted and went into BIOS, in expectation of setting "Numlock on at Boot" but apparently there is no such option on recent Gigabyte motherboards. The other keys I remapped around the perimeter of the Numpad still operate properly (obviously, since there is no Numlock any more!) but numpad 7 is "Home" no matter what I try to do.

For the time being, I suppose that I will try to re-map an unused key to Numlock so that I can press it if I need to. What is going on?

Finally, is there any advantage to remapping the Numpad, for instance, there seems to be a subtle difference between "3" and "numpad 3" ?

Thank you.

[alaricljs]

Are you running Windows 7?  The on screen keyboard (Control Panel -> Ease of Access) has an options button that toggles num lock.

The 3 and the numpad 3 are 2 distinctly different keycodes.  You can only do extended Alt-codes with the numpad.  You can only get symbols via the number line.  Num Lock doesn't interact with the number line in any way.  Any advantage would be for you to determine based on your usage.

[Soarer]

I have no further ideas on standalone converter problem; there's just not enough clues

Apparently num-lock on Win7 ends up 'off' at boot if various registry settings get non-matching (or wrong) values set in them, see the answer here... http://answers.microsoft.com/en-us/windows/forum/windows_7-hardware/want-to-set-num-lock-to-be-on-and-active-at/7b48ae6f-fa9c-4bc2-8359-81e0c5e5d6a2.

I don't know why that would get shot by plugging in an iMate! I must admit my first thought was that you'd simply pressed 'clear' on the AEK when testing it, which (I think) gets mapped to num-lock by the iMate... but I don't think that would still affect anything after a reboot.

On-screen keyboard sounds like a good solution if it's only rare cases when you need to toggle num-lock.

Although I don't tend to use num-lock much, I do keep it mapped somewhere. You could put it in a macro if you don't have any single keys left you could use.

If you really only ever use the numpad for numbers, and never with any modifiers pressed, then I suppose it shouldn't really make any difference which codes are produced by it.

[fohat.digs]

Good answers, both.

After sleeping on it, I realized that I am an idiot.

I use keyboardtest.exe for testing, (I like it better than any of the others and it finds all the keys on the 122) and clearly I had tested Numlock on the AEK2.

BIOS was simply remembering that, so I rebooted with a standard PS2 keyboard and pressed the Numlock a couple of times and now I am back in business.

Still, Gigabyte should allow a "Numlock on at Bootup" selection in BIOS, I know a lot of people don't use it, but a lot of us do.

Thank you.

[Soarer]

Interesting... it must be Windows 7 remembering the num-lock state. I don't know if Vista did that as well, but XP certainly doesn't. So we learnt that from this

[rknize]

I finally got off my butt and got the layouts on this F122 and my AT the way I want.  Thanks to the documentation, it was fairly straight-forward.  LANG_4 was the only tricky bit on the F122.  Perhaps an alias or at least a hint in the doc would be helpful.    The AT is from wcass, so that one is mainly some tweaks to his.

FWIW, they are attached.

[metalliqaz]

Where do you get that keyboardtest.exe?  I googled it but there seem to be many different "shareware" tools that go by the same name.

[Soarer]

rknize - Thanks for posting configs Is it OK if I include them in the next release zip?

Re. LANG_4, I won't add a specific hint, but I will try to make clearer the method for finding out what the codes are... it has been mentioned before, but since hid_listen output is detailed on the Trouble page, it's not obvious that it has 'regular' uses as well!

[fohat.digs]

keyboardtest.exe

is a horrible thing to Google, you would think that it would turn up exactly as you typed it, but no .....

I did not know that anything except photos could be posted here directly.

This is the one I like to use:

PS -

I forgot to mention, the program has a few odd quirks (who doesn't?)

It ignores "Tab", it locks up after "F10", and it respects "CapLock"

If it goes funny during or after "F10", "Alt" or "Spacebar" you should use your mouse to highlight whatever is in the dialog box, then when you press a key it will go back to normal.

These are not bugs, they are "idiosyncrasies" - HAHA

[rknize]

rknize - Thanks for posting configs Is it OK if I include them in the next release zip?

Re. LANG_4, I won't add a specific hint, but I will try to make clearer the method for finding out what the codes are... it has been mentioned before, but since hid_listen output is detailed on the Trouble page, it's not obvious that it has 'regular' uses as well!

That would be great.  Even dropping a hint in the comment column of the codes page would help.  It's a common enough scenario, I think.  Perhaps some "cook books" for common projects would be helpful as well.  It would have saved me multiple look ups in multiple keycode tables on the Internet if I just knew which switch translated to which soarerese HID code for the mapping table. 

Feel free to include them.  I left some comments in the remapblocks.

[Halvar]

I posted my M-122 configuration in this earlier post, and it's also in the newest distribution from Soarer in a file named halvar.sc

http://geekhack.org/index.php?topic=17458.msg658161#msg658161

I put a lot of comments in it, also about LANG_4, numpad and the other keys on terminal keyboards, so I'm sorry that you had to find it all out for yourself again.   

Halvar

[Krogenar]

I finally got off my butt and got the layouts on this F122 and my AT the way I want.  Thanks to the documentation, it was fairly straight-forward.  LANG_4 was the only tricky bit on the F122.  Perhaps an alias or at least a hint in the doc would be helpful.    The AT is from wcass, so that one is mainly some tweaks to his.

FWIW, they are attached.

Rknize, how would you describe typing on a F-122 (makes it sound like a fighter jet, LOL) -- awesome?

[fohat.digs]

The large Model Fs are perhaps not quite as loud as the small ones, just because of greater mass of the entire assembly.

The 122s and XTs have the metal back plate, while the ATs are plastic all around. The sound is as different as the feel, in comparison to the M.

Sound is not my big thing, it is feel. I have flossed almost all of my buckling springs, and it is a big help in the "annoyance" department.

While the Model Ms are already less resonant than Fs, with the one-piece plastic barrel (aka chimney) assembly vs the heavy F front plate, the metallic resonance of the Fs is smoothed out very nicely by the floss, without being unnecessarily deadened.

The 122 and XT have 3 different steel plates (front barrel plate, the back plate of the switch board, and the case back) so there is a lot of mass and deep resonance working for you already.

The single metal back plate of the M is isolated from the rest of the assembly by the rubber mat, and buried in the interior of the case.

For me, the deep underlying resonance remaining, with the upper register attenuated by the floss, is the perfect combination for sound.

[Soarer]

rknize - Thanks for posting configs Is it OK if I include them in the next release zip?

Re. LANG_4, I won't add a specific hint, but I will try to make clearer the method for finding out what the codes are... it has been mentioned before, but since hid_listen output is detailed on the Trouble page, it's not obvious that it has 'regular' uses as well!

That would be great.  Even dropping a hint in the comment column of the codes page would help.  It's a common enough scenario, I think.

For single oddities like that key, I'm not sure that's the best way to do it - perplexed user would still have to scour the list. I'm thinking of replacing the single line "See the HID code list for HID and META values." at the end of the Config page into a section (perhaps called 'Finding Codes') which would mostly be a rewrite of this post.

Perhaps some "cook books" for common projects would be helpful as well.  It would have saved me multiple look ups in multiple keycode tables on the Internet if I just knew which switch translated to which soarerese HID code for the mapping table. 

An index to the example configs would be useful, even if it says little more than which are examples for 122-key etc. I'd love to have time to produce some keymap diagrams like John Elliot's, but when I have any time to work on this there always seems to be something more important/interesting to do!

Feel free to include them.  I left some comments in the remapblocks.

Thanks

[Soarer]

Here's a draft of the new section, which will go just after the Overview section on the Config page.

Initial Translation

For most keys it's fairly obvious which codes are produced after the fixed initial translation. However, for some keys on a 122-key keyboard, and the normally unused extra keys on various keyboards, the codes don't follow an obvious pattern.

To find out which code a key is producing after the fixed initial translation, you can use hid_listen to examine the Diagnostic Output.

When you press a key, a set of codes is sent by the converter to hid_listen. For example, pressing the centre key in the 122-key's cursor block results in:

Code: [Select]

r62 +93 d93 rE2 -93 u93
Note the code prefixed by '+', then look it up in the HID code list. For the example, looking up 93 reveals that the identifier to use for remapping or triggering a macro would be LANG_4.

[rknize]

Rknize, how would you describe typing on a F-122 (makes it sound like a fighter jet, LOL) -- awesome?

As a long-time Model M user (bought my first one new in 1994), the Model F is the ultimate typing feel of you like tactile switches.  Nothing I have experienced so far compares.  I have tried a few...see my sig.

Stock, they are extremely annoying to listen to.  The spring ring combined with the steel plates makes the whole keyboard sing.  It starts bugging me almost immediately, to say nothing about those around me.  Model M's ring too, but you mainly only hear the spring because the steel plate is damped.

The first thing I did to my AT when I got it was a floss mod.  I don't care for the floss mod on Model Ms...I feel it swamps the feel of the switch too much and adds more weight to a switch that borders on being a bit too heavy for me.  However, on the F it works wonderfully.  The F already has a lighter feel and extremely sharp snap on actuation (no plastic membranes or rubber mats).  The floss mod, when done just right, stops the ring without affecting the feel enough to matter.  It's still plenty loud.  I can't use it at work.  But it's quiet enough that my wife doesn't complain too much.  At work I use a grease-modded Model M (the one I bought in '94 in fact) and an SSK on occasion.

[rknize]

Here's a draft of the new section, which will go just after the Overview section on the Config page.

Initial Translation

For most keys it's fairly obvious which codes are produced after the fixed initial translation. However, for some keys on a 122-key keyboard, and the normally unused extra keys on various keyboards, the codes don't follow an obvious pattern.

To find out which code a key is producing after the fixed initial translation, you can use hid_listen to examine the Diagnostic Output.

When you press a key, a set of codes is sent by the converter to hid_listen. For example, pressing the centre key in the 122-key's cursor block results in:

Code: [Select]

r62 +93 d93 rE2 -93 u93
Note the code prefixed by '+', then look it up in the HID code list. For the example, looking up 93 reveals that the identifier to use for remapping or triggering a macro would be LANG_4.

Awesome.  What really threw me was that the codes at the HID level for set 3 were actually the translated "PC" values and not the original key codes.  I suppose this is just a legacy thing.

[rknize]

I posted my M-122 configuration in this earlier post, and it's also in the newest distribution from Soarer in a file named halvar.sc

http://geekhack.org/index.php?topic=17458.msg658161#msg658161

I put a lot of comments in it, also about LANG_4, numpad and the other keys on terminal keyboards, so I'm sorry that you had to find it all out for yourself again.   

Halvar

Ah yes, I see it now.  I guess I didn't look because the file name didn't give any hints about what keyboard it was for.  Oh well...it was a learning experience, at any rate.  Sorry, I haven't followed this thread as closely as I probably should have.

Perhaps a naming convention or a README index to all the configs would be helpful in that regard.

[Soarer]

Awesome.  What really threw me was that the codes at the HID level for set 3 were actually the translated "PC" values and not the original key codes.  I suppose this is just a legacy thing.

The initial translation to HID tries to make a common ground ready for configs. That means a lot of configs can be written without caring what kind of keyboard is plugged in and therefore without caring what kind of scan codes the keyboard is actually producing (e..g. Colemak remappings).

Ah yes, I see it now.  I guess I didn't look because the file name didn't give any hints about what keyboard it was for.  Oh well...it was a learning experience, at any rate.  Sorry, I haven't followed this thread as closely as I probably should have.

Heh, no-one is saying you should have known, just (annoyingly) that you could! But in this case, what would you search for? Even I'm having trouble finding the posts where I've covered looking with hid_listen to find codes, and I'm cheating by using words from the answer, not the question!

[Krogenar]

rkinize, fohat, thanks for your feedback on the Model F -- I've got a Model F in the mail and I'm looking forward to trying it out. As far as sound goes, I usually don't care about the sound on a Model M, so I'm hoping flossing won't be necessary, I actually like them loud -- but maybe the *pInG* sound will annoy me. Thanks also to Halvar and rkinize for the Soarer codes for 122-Model M's and F's.

[AKIMbO]

F122's have a sharper, more shrill ping sound when compared to AT F and XT boards.  For me, I can't stand the sound of a F122 without floss mod.  The stock sound of an AT F is just right imo.

[rknize]

It's a very personal choice.  If you like loud keyboards, the F won't disappoint.

[rknize]

Heh, no-one is saying you should have known, just (annoyingly) that you could! But in this case, what would you search for? Even I'm having trouble finding the posts where I've covered looking with hid_listen to find codes, and I'm cheating by using words from the answer, not the question!

Yeah, I know what you mean.  The docs are a great reference once you know what questions to ask.  People learn things differently.  I learn the most quickly by working backwards from an example.  Unfortunately, I missed Halver's.  Most folks are going to have a specific project in mind when using this.  I imagine that IBM XT and terminal keyboard owners are probably some of your biggest users.  It may be worthwhile to provide walk-throughs for those and provide boilerplate configs for those projects' FAQs?