Questions about the Unicomp 122

[Inops]

Hi all.
I'm interested in the Unicomp 122 keyboard (http://pckeyboard.com/page/FeaturedProducts/UB40B5A). I read the comments about Unicomp's response time to emails, so I thought it best to post it here first, as well as get some opinions.

The main reason for being interested in this keyboard -- other than the buckling springs, which the joy of I discovered on my Chemistry teacher's replica Model M keyboard -- is the large amount of function keys. However, I have queries because of this thread http://geekhack.org/index.php?topic=19024.0, which states that the PC/5250 version sends modified functional keys for the extra keys (e.g. Shift+F1 for F13). This somewhat defeats the purpose of the keyboard for me. The aforementioned thread mentions the "Emulator" version of the keyboard, which apparently sends individual scan codes for the extra keys. My question is there a USB version of the "Emulator" (I alternate between a dinky desktop and a laptop, so no PS/2 for me); and if there is, would this work pretty much out-of-the-box for MS-Windows and Linux? (I jump distros about every 6 months, but currently I'm using Linux Mint). Additional, does this particular model have easily removable keys: I swap Caps Lock with Left Ctrl, and would like to have this physically replicated on my keyboard.

I'm in the UK, does anyone have experiences with the postage time/possible taxes?

Regards, and thanks in advance,
Jordan.

[0100010]

I received this from Unicomp from a previous inquiry:

"The Emulator uses unique codes for every key. You will need a special keyboard driver to enable the unique key codes that aren't support by the operating system. The difference between the 3270 and 5250 versions is legending only. The keys function the same way."

Didn't ask about interface though, I figured I would need to add a Soarer converter.

[rootwyrm]

I received this from Unicomp from a previous inquiry:

"The Emulator uses unique codes for every key. You will need a special keyboard driver to enable the unique key codes that aren't support by the operating system. The difference between the 3270 and 5250 versions is legending only. The keys function the same way."

Didn't ask about interface though, I figured I would need to add a Soarer converter.

Nope. Read the email again.

You absolutely do not and actually cannot uses a converter with the Unicomp 122. Period. It is a standard USBHID or PS/2 that speaks standard scancodes except for selected keys. The TN5250 keys (F13-F24, etc) are not specified in the USBHID standard or default Microsoft PS/2 driver. The TN5250 emulation software is expected to provide the appropriate driver.

[0100010]

Cool, thanks rootwyrm.  Like I said - I didn't ask about interface format, only about unique scancodes - so their email had no reference there.  Good to know no need for a Soarer.  Now off to read more about this TN5250...

[rootwyrm]

Cool, thanks rootwyrm.  Like I said - I didn't ask about interface format, only about unique scancodes - so their email had no reference there.  Good to know no need for a Soarer.  Now off to read more about this TN5250...

Put on your "kill me now" hat. It's really, really bad. Suffice to say, the key part is the low-level (AKA kernel-mode) driver itself which is not so easily available. It may be possible to use AutoHotKey since it's only unique scancodes in addition to standard ones - specifically so the keyboard will work as a normal AND 3270/5250 - but I've not personally tested it.

As far as getting kernel-mode drivers that give you F13-F24 in Windows baseline? Fuhgeddaboutid. Ain't happening. KM drivers must be double-signed (manufacturer/vendor plus Microsoft) or you need to flip the 'untrusted drivers' switch at install. AFAIK most 5250 clients don't even provide the kernel-mode drivers, instead requiring the GPO administrator or AS/400 administrator to configure the clients for blind scancode-output mapping by hand on USBHID types.

[Inops]

Thanks for the replies, guys.
Right, so for me, the emulator is off the table? There's a jargon in the above posts, but that's what I fathomed. Am I right?

If so, I'm left with the PC/5250, not that much of a problem, the extra keys'll still come in useful with AutoHotKey (I've heard about AutoKey as an equivalent for Linux which I think I'll try).

So I'm guessing, because of the use of the standard driver, PC/5250 will work out of the box for Windows and Linux?

Are the keys easily swapped on that model? Also, what about the UK postage times and taxes?

Regards, Jordan.

[kps]

The TN5250 keys (F13-F24, etc) are not specified in the USBHID standard

The USB HID standard does in fact include F13–F24 as Usage IDs 104–115 of the Keyboard page (p56, USB HID Usage Tables). The left-side keys on typical IBM 122-key terminal keyboards (CrSel, ExSel, etc) are also included.

[rootwyrm]

The TN5250 keys (F13-F24, etc) are not specified in the USBHID standard

The USB HID standard does in fact include F13–F24 as Usage IDs 104–115 of the Keyboard page (p56, USB HID Usage Tables). The left-side keys on typical IBM 122-key terminal keyboards (CrSel, ExSel, etc) are also included.

Sorry, was late, brainfried, etc. Meant to say that they're not in the default Microsoft USBHID driver and not specified in WHQL USBHID. WHQL being essentially the bible for Windows regardless of what USBHID says. (It was a long weekend.)

[Inops]

Hello?

Anyone got any information or guidance to my questions?

Thanks.

[Soarer]

I'd say the best choice is one that outputs unique codes, and uses PS/2, but I don't know which unicomp model that is. On some there is/was a jumper that can change the behaviour, which might help.

A PS/2 version certainly can be used with my converter. (Off-the-shelf converters won't convert any of the keys other than the usual 104/5 keys).

Without a converter... On Windows, the PS/2 scan codes should mostly get through, so you could remap them with Autohotkey.  On Linux, you should also be able to find a way to use the extra codes.

Windows USB HID driver does convert most if not all of the codes produced  - e.g. F13 to F24, crsel, exsel, etc. - to scan codes (and has virtual key (VK_ *) definitions for them). It's just that no software on Windows uses them! So again, Autohotkey should be able to remap them.

[microsoft windows]

I'VE BEEN INTERESTED IN THOSE 122 KEY BOARDS FOR A WHILE TOO.  I THINK THOSE EXTRA FUNCTION KEYS WOULD COME IN HANDY FOR INPUTTING GREEK AND/OR OTHER SPECIAL CHARACTERS THAT I USE A LOT WHEN CREATING WORKSHEETS.

[wcass]

i know a few guys here that use a 122 with Soarer's converter in them (i'm typeing on one now). this one started life as a 1394100. i've remapped about 30 keys with no issues at all.
[attach=1

[rootwyrm]

Windows USB HID driver does convert most if not all of the codes produced  - e.g. F13 to F24, crsel, exsel, etc. - to scan codes (and has virtual key (VK_ *) definitions for them). It's just that no software on Windows uses them! So again, Autohotkey should be able to remap them.

I've never tested with USB, so I'll take your word for it. F13-F24 are blank in the WHQL docs I have for USB though, so basically per WHQL and Microsoft, may stop working or go away at any moment without any advance warning. The VK_* stems from TN5250, by the way, and is inherited from Windows CE of all places. So it's preferential to have a device specific driver as a future USBHID update may break functionality. My WHQL docs may or may not be out of date, but given the VK_ source.. well, has anyone tried with Windows 8? ;P

The obvious downside being that such a driver would require doing and passing WHQL driver suite since you'd need a signed KM for USBHID (since it'd need to be pre-login.) Plus the costs associated - $200+/yr for a CSC sufficient for 3rd Party Boot. (You'd need an SHA256 to cover both i386, x64 and ARM plus WHQL Signature is also required for ARM.)

[Soarer]

Yeah. Of course, there's a big difference between "what is required for a keyboard and/or driver for WHQL certification" and "what is supported by the Microsoft keyboard drivers". The first is a minimum, and doesn't imply that anything more (in the Microsoft driver) might be removed (and that's highly unlikely, IMO). But using a 3rd-party driver might lose you that functionality.

Windows Hardware Certification Requirements - says Win 8 on the page, but says Win 7 too, at least in the devices doc.
Key Support, Keyboard Scan Codes, and Windows - translate.pdf shows which HID codes will convert to scan codes - F13-F24 do, CrSel etc don't, but do produce VK_* events, IIRC.
named keys that you can specify in the KeyboardEvent.key attribute in Windows Internet Explorer 9 - includes VK_CRSEL etc!

Virtual keys are a logical extension for keys that don't have scan codes, so I can't see that mechanism going away, ever. Surface probably added a few more, if anything!

[Soarer]

i've remapped about 30 keys with no issues at all.

And you've remapped about 30 keys to nothing at all!

[zig_ziglar]

I'm thinking of getting one, but most of the above went over my head.

Do the extra keys on the modern Unicomp version work or not? Would they show up as additional key bind options in hotkey setups?

At work some of the software I use have F13-24 commands that we activate by pressing shift-F1 etc, being able to just press the key would be good, but I'd prefer for them to be detected by windows as unique keys like modern keyboards with macro keys.