deskthority

Hmm, okay, it lists pin for Nocke, but not for Nocken...

Ah, I didn't know that 'einem Nocken' could mean 'a pin'... but nor do any German-English dictionaries on the web, it seems :lol:

Now we have two plausible mis-translations, even if the original German was that text - we could probably guess at alternative original texts and find further meanings ...

Well, I think you're both wrong :D

Not that it matters now that we know that the identification means a clicky switch, and that's all we really care about, but still...

A cam is a rotating or sliding piece in a mechanical linkage used especially in transforming rotary motion into linear motion ...

7bit wrote:BTW: A color comparison between those greys whould be nice.

36%, 37%, 39% ?

Actually, I think yab8433408 is correct, and Cherry are wrong.

The 1994 "Keymodule MX" brochure has white (A) and grey 36 (B) down as "tactile" with "Movement Differential" (hysteresis). It looks like the only switches that have hysteresis are the clicky switches (with this being generated by the ...

The margin helps, and that I have all calls to seq_malloc in one part of the code, where the stack is probably close to it's maximum anyway. I played with tracking stack use by recording the lowest point in an ISR, but static analysis would be more reliable, I'll give that script a try. Whether ...

I have a very simple allocator, and an even simpler free ;)

#include "seq_malloc.h"

extern char __heap_start;
extern char __heap_end;

uint8_t* seq_malloc_heap_end = (uint8_t*)&__heap_end;
uint8_t* seq_brkval = (uint8_t*)&__heap_start;
#define seq_malloc_margin 32

int get_free_sram(void ...

Chorded Fn keys might be useful, since you don't have many keys conveniently placed to use as Fn keys on most beam spring 'boards!

The core requirement I wanted to support with my layers was something like 'a few layers, with perhaps left and right Fn keys for each layer'. So, my code only supports ...

Yeah, that - sneakiness ;) FN keys can't be used as modifiers in macro triggers directly yet - I might put that in when get round to revamping the trigger logic - until then there's a fairly easy workaround: in the FN layer, remap the key to some HID code you don't use elsewhere, and then trigger ...

Best dead cockroach ever!

The other way is the "bitmap" technique Soarer uses, and Hasu developed I think. I don't know any keyboards that use it. But our own inventions do: custom controllers and converters.

http://cl.jroo.me/z3/g/1/f/e/a.baa-Old-History-Teacher.jpg

dfj has (low-speed?) NKRO, 9 June 2010
dfj has ...

Nice legends!

Just look up the codes after '+' in the codes list, and remap them! Last three are F23, F24 and LANG_3... no idea why they come out as 'a'...

Thanks randku, glad you like it! Sounds like a really neat build :D

Muirium, everyone already knows you're nuts :P

You can wire an extra LED to the PD6 pin if you want, in the same way as the lock LEDs, to make an external copy of the onboard LED.

Switching needs to disconnect at least 3 USB ...

Hey 7bit - where are my replacement keys for Round 3?! Do I need to send my (unchanged) address to the bot?

Oh cool, that one sounds really easy to handle! It doesn't even really need an interrupt if it's scanned at 1kHz, I think.

I want to rewrite some of my USB descriptor handling to be more flexible, so perhaps I'll add a low-speed 6KRO mode at some point. Shame, because apart from KVMs and PS3s the current setup is very compatible with the main OSs. I haven't opened the source yet.

KVMs are fickle... at best they read the 6KRO report, and I guess many are only tested with low-speed USB keyboards (and mice). They certainly only ever present a 6KRO keyboard to the attached computers, and are unlikely to have media key support. They might get startled if the keyboard is swapped ...

Muirium wrote:From West Calder! I should pop down on my bike, if I had a need for these myself…

Say, Soarer, do you happen to know a little about USB MIDI? I've a 16x8 straight matrix PS/2 Tipro and I'm getting strange ideas…

Heh, neighbours! I don't know USB MIDI at all, sorry.

Really? Does it have a PCB on it? There's a few pins left, but not many that can generate an interrupt.

The good thing about a pot is that it wouldn't have to be translated into keypresses! The downside is that is takes over completely, and you can't set the volume any other way. Hmm.

Got a link to the encoder? The problem with them is debouncing - can't do that on an interrupt since it would swamp ...

Hmm, I dunno. It would be a lot easier to implement an absolute control driven by a potentiometer, if it would work well (HID supports absolute controls, but there might be niggles)! That's probably what I'd prefer to use, myself.

It should be possible though to use an external decoder circuit to ...

Unicomp only seem to have one variant of PS/2 122-key now, and I'm not sure which it is! If it's the one which can have its behaviour modified by a jumper inside then that's fine. What you don't want is one that sends combos (e.g. shift + F1) instead of F13 to F24 and which doesn't have that jumper.

That's why I always say "full-speed" instead of "USB 2"! :ugeek:

USB 1.0 is so rare that "USB 1" usually means USB 1.1. And TBH, I haven't bothered learning any details about 1.0 because of that :o

Anyway, when using low- or full-speed, it shouldn't matter whether the host is USB 1.1 or USB 2.0 ...

I doubt they're USB 1.0... probably USB 1.1, which has full-speed

Hasu's notes describe NKRO vs normal pretty well... describing how HID works is best left to the HID docs!

It is. Bitmap with full-speed USB is problem free!

IIRC, the chocs use bitmaps as well, but because of low-speed USB they have to split it over two keyboard interfaces, which is where the problem comes with OS-X.

I don't know the k65, but since the page lists Windows as a requirement I'd guess it works in a similar way to the Chocs, which use low-speed USB, and appear as two keyboards.

Appearing as two keyboards isn't via a hub (real or virtual) - each USB device can have multiple interfaces, and a keyboard ...

I dunno. I have the hardware, but it keeps slipping down my todo list.
hasu and darkness have code already!

Soarer wrote:Although, in the latest (unreleased) docs I have added a small section to the Config page describing how to use hid_listen for identifying which codes you need to remap

Now released

Minor doc update (v1.12) posted to the OP covering changes since v1.10, some polishing, updated thread contents pages, and search boxes for searching this thread and the GH thread.

Well, it's an idea that might be useful. Particularly for ifselect, since that's the only one that will help you work with multiple PS/2 keyboards with the same ID! But you can probably use the fact that later remappings override earlier ones to achieve the same effect...

remapblock
F1 F2 # remap ...