We Kana believe it! Raspberry Pi Foundation launches Japanese keyboard The Raspberry Pi foundation has created its first made-for-Asia peripheral: a Japanese keyboard. The device retains the three USB type-A ports and a single micro-USB port to plug into the Pi itself found on other official Pi keyboards, but has a new moulding to handle the 83 keys needed to offer input for Japan’s three …
Probably trying to keep the commercial ball rolling by producing new stuff. However the one thing they've never done yet is produce adequate documentation. This might be partly explained by the declared approach to designing this keyboard - reverse engineer someone else's kit and then proceed by trial and error. This is not robust engineering practice.
Try getting documentation in English for a Japanese keyboard. The foundation could not find any.
Try reading any English documentation for any Japanese product. In Japanese culture, the boss is the boss because he is the best at everything (just like Kim Jong-un) - including writing English. Just imagine your nearest PHB being involved in the documentation at all - then imagine the chaos when he translates it into Greek. I have had colleagues thoroughly confused because they cannot recognise the most popular Japanese->English translation errors.
SAMBA was made by packet sniffing - far more reliable than reverse engineering or Microsoft's documentation. Linux has some handy tools for working out what a keyboard is sending to the kernel. I would expect them to be able to get good results because they had feedback from people who type in Japanese.
There's code all over the Internet for virtual keyboards, handling scancodes, keyboard layouts, modifiers, the symbols produced, etc. If nothing else, Linux itself has handling for them, not to mention Linux GUIs, not to mention virtual keyboard software that runs on those GUIs that produces the scancodes and tells you which keys are modifiers under which circumstances, etc. right in the code.
Hell, my secondary school 20 years ago offered a course in Chinese Wordprocessing with pretty much everything you'd need to know from a non-technical basis, and coupled with a basic knowledge of scancodes you could sort it out in a few hours, let alone having a native Japanese speaker that you can ask to get that kind of documentatio / translate that for you.
What this suggests is they got someone to buy a Japanese keyboard, press every key while running a scancode monitoring tool, and then did that for a few hours recording the results. Possibly the cheapest/worst way to arrive at such a thing, because it's unlikely to catch all the unusual / little used modifiers and then you're hardcoding everything.
Honestly, it's 2020. Buy a Japanese USB keyboard encoder IC (literally pence) and then hook it up to the specified keys that will be in its datasheet. Likely it'll literally just specify the row/column on the keyboard matrix that it expects each key too.
Just checked hut1_12v2.pdf. Footnote 28 for the table you mentioned shows the sort of problem you should expect for this sort of thing. Some hardware precedes the standard. Some OS behaviour does not match the standard. Hardware gets changed to fool the OS into doing what the user wants. Actually doing what it says in the standard causes / to appear when other keyboards produce ¥ (when the user actually wants ¥).
Sometimes it is easier to do what everyone else does than to replace all existing Japanese keyboards and installed operating systems with something standards compliant.
I agree that it seems rather odd that RPi are re-inventing this keyboard themselves (that worked so well with the USB-C power connector…) instead of just re-badging an existing compact Japanese keyboard (but make it white and red), such rare things as they must be…
(But I was really just commenting to say a big ツ for the linguistic jokes in the headings! ツ )
(Yes, I'm sure that was some very bad inverse-mojibake on my part, sorry, sorry.)
As for the keyboard itself - good god, it's HORRIBLE! I want - nay, *NEED* - a keyboard with proper keys, y'know, blocks about half an inch deep that actually move under yer actual fingers as you attempt to type. And all the same size, not function keys that are three times thinner than everything else "we're not really here, ignore us, ignore us!" I'll continue using my good old Dell 3200 skull-crushers.
I too downloaded those documents from usb.org when writing a USB keyboard driver, but I still needed to plug a physical Japanese keyboard in and read the actual physical codes being sent to actually get the driver actually working. A nice historical link, the Japanese keyboard layout is one of the few bitmap layouts, aka BBC layout. Such joy to press Shift-6 and get & just as god intended. :)
And Yen is commonly aliased to backslash, not to divide.
Here's me who always thought that packet sniffing is reverse engineering.
Sure, reverse engineering can be applied to code - you take the assembler code and build the corresponding functions and procedures.
But taking what is going through the wire can allow to reverse engineer the procedures that are sending the data, then you have what you need to build the application that replicates that.
There is a big difference in law. Software licences can include "do not dissemble/reverse compile/run through a debugger/boil a kid in its mother's milk/run in an emulated environment". Such conditions can actually be legal and disqualify your license making further use of the software copyright infringement (which is now a criminal rather than civil offence). Either the lawyers did not know about pack sniffing or thought that it would be better not to mention something they could not legally prevent to avoid giving people ideas. The distinction was important for SAMBA back when Microsoft were looking at criminalising NAS that did not have a Windows license.
For a bunch of techies in the UK sourcing a keyboard is trivial but the Pi is also intended to be used by the thoroughly computer illiterate who have severe supply chain problems and cannot read any language you have heard of. At some level of computer illiteracy, getting something from the manufacturer with a "this will work - no problem" statement is really useful. Being able to buy the whole kit from from a single source is close to essential when the nearest road is a day's hike away. Having the hardware precisely match the pictures prevents the need for handling technical support calls in an unrecognisable language.
The current Pi kits will remain in production until January 2026. That is a difficult promise to make unless you have all the tech and tools needed to make the whole kit.
If you're looking for somewhere where it's really hard to find electronics and the road is a day's walk away, Japan might not be it. Japanese is spoken elsewhere, but not so much that you'd expect to stock many Japanese keyboards for the general public there. If they want to create keyboards for languages that aren't well represented at the moment, maybe they should focus on those which haven't been well-established in computing for five decades. There are languages covering millions that fit that bill.
My assumption, given the usual way these things happen, they’ve had some direct feedback and have created this keyboard in response to market demand. It is a neat little keyboard that includes a built in USB hub, which then enables more USB devices to be plugged in.
If the market for the Pi itself is large in Japan then it is a compliment to their community there to support their culture directly.
Pi documentation is published in some nice books that you can buy or have for free as pdf and numerous magazine publications some of them also free in electronic form, and my Japanese car has a wonderful printed manual that I’ve never needed to open.
> Being able to buy the whole kit from from a single source
I've just completed a little project involving a number of Pis as servers and workstations. Being able to buy the keyboards and mice together was administratively easier, technically easier, and in the end, aesthetically easier. They're good enough little devices, and the spare USB ports will reduce a handful of issues as the project continues, as the Pis themselves could be "hidden", while still allowing USB sticks etc to be plugged in.
Confused... I thought that USB keyboard packets contained a 16bit value for a key (along with an up/down flag). Further I thought that the values were pretty well defined. Also hiring a Japanese-speaking temp/intern feels like a good idea for this project.
Still congrats. Onto Chinese, Korean, Thai keyboards next!
Japanese actually has three syllabaries, not two since Romaji (Roman) characters are commonly used in signage and the like in Japan these days. Japanese kanji based on Chinese characters are not the only form of words though -- combinations of kanji, called jukugo can also represent individual words.
It's complicated.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
