The Balthazar laptop: An all-European RISC-V Free Hardware computer The Balthazar project is designing an all-Free Hardware laptop based around RISC-V and several existing standards. Balthazar is an interesting hardware project which we were told about at the Open Source Policy Summit in Brussels last week. It's a project to build a standards-based, fully open source RISC-V laptop "for all …
Why do projects like this end up with weird looking designs? OLPC was the same.
Why do they think making something that looks either like a child's toy, or something rendered by someone the first time they tried Blender, is good? Surely it reduces the appeal of the entire project?
Why does every do-good, free-as-in-freedom hardware project* seem to use people they believe are "egonomic keyboard designers" -- people who, based on their results, are terrible at designing good keyboards?
Forget the general aesthetics of the unit -- look at their keyboard, which is a critical system interface (magnify the image as needed):
https://regmedia.co.uk/2023/02/09/balthazar.jpg
Two tiny little buttons instead of a space bar?!
*I am in favor of such projects.
[Author here]
> Two tiny little buttons instead of a space bar?!
I think I can answer that one, at least.
Space bars are tricky. You need supports underneath so that when the user presses at one end, the whole thing doesn't seesaw -- you want it to depress evenly to trigger the mechanism, but that means when pressed off-centre, it needs to be pulled down in the middle.
The MNT Reform has the same issue, and by using 2 large keys the size of Shift or Backspace, you can use the standard key mechanism.
This is quite common on Japanese keyboard layouts.
Also, for people addicted to horrible old Unix editors, you can put Alt or Ctrl keys in the middle where it's easy to press with the thumbs, rather than the weak little fingers which share a tendon with the ring fingers.
I don't like it, but it _is_ a thing.
As Michael W Lucas wrote in his book "Ed Mastery"
..."ed(1) is the standard Unix text editor. Dennis Ritchie, co-creator of Unix, declared it so"...
"A few are so lost as to devote themselves to turgid editors meant for mere users, such as vim and Emacs. This way lies not only appalling sysadmin skills, but an absence of moral fiber."
The non-manface edition has an mature BSD daemon on the cover - pity its not also a reg icon :(
Mastering ed isn't exactly difficult. It's not like it is either complicated or complex. No need to buy a book from some hack in the midwest, instead try info ed.
Still, it's one step up from toggling config file changes in on the front panel ... which is pretty much all it was built for.
Learn it. It's useful in rare (these days) emergencies, and can be quite handy in script files, if nothing else. But choose another editor for actual writing.
Only people who make mistakes actually need an editor:
$ cat > testfile.txt
Hello World!
^D
$
Do it once, get it right the first time, no need for silly editors. Some may argue about the whole needle and butterfly thing ( https://xkcd.com/378/ ) but they are just bat$#1+ crazy, that takes far too much upfront planning and patience.
I agree, space bars are a slightly difficult design compared to smaller keys. However, they're a problem that have been solved since forever at this point.
I don't think you can frame the choice as the result of a technical challenge.
The only other reason for this design that I can think of would be cost saving but I cannot see how it could be, the entire keyboard can't be looked at as some form of cost saving - manufacturers pump out millions of laptop keyboards all the time - using a custom layout like this is likely to cost more to make than going for something much closer to a standard layout. I can't help thinking it is designed in the way it is simply to be "different".
[Author here]
> I don't think you can frame the choice as the result of a technical challenge.
I disagree: I think you can, _if_ you are hand-making the keyboards from off-the-shelf key-switches and 3D printed keys.
But if you _can_ source COTS keyboards, then yes, that's a better option. I don't think it's that easy, though. It ought to be, but it sadly isn't. Laptop hardware is not standardised at all.
If you think that keyboard is bad, checkout the MNT Reform. Same idea - open source hardware and the most godawful keyboard ever devised in the history of keyboards.
https://shop.mntre.com/products/mnt-reform
You don't just get one space bar, you get two stumpy space bars with two alt keys in between. It's just bizarre and ugly. I don't know if this was someone's idea of the perfect keyboard, a limitation of sourcing parts or something else.
"the most godawful keyboard ever devised in the history of keyboards."
Never used a Commodor PET, did you. How about the Mattel Aquarius? The Sinclair ZX81? The TI-99/4? The Sinclair 2068 (these last two AKA the "Wot? No backspace?" keyboards).
And of course IBM's very own hono(u)rable mention, that of the IBM PCjr ...
And of course there are atrocities like the Big Keys LX, various roll-up rubber keyboards and so-called "projection" keyboards. Also don't forget every single on-screen keyboard ever inflicted upon the public as punishment for not doing their due diligence and so getting suckered into spending money on a device without a keyboard, even though it clearly needs one for proper operation.
In fairness to Commodore, only the PET2001 keyboard was horrible.
A dishonourable mention goes to Sharp for the MZ80K, which had a keyboard very much like the aforementioned 2001. The original Tandy Color Computer keyboard was rough, as was the membrane abomination on the Atari 400. The Oric 1 was grim too, I don't think these were ever sold in the US, so you were lucky there...
Hmm. I think I'd take the PET's chiclet keyboard over any of the membrane keyboards. At least the PET keys had some travel.
Also, it was easy to get an add-on conventional keyboard for the PET 2001. It had a ribbon cable which you snaked into the front of the case, between the top and bottom halves, so it sat in front of the case. The ribbon connector was a pass-through so both keyboards would work, which made for some entertaining two-person operation, NCIS-style.
The reason the Atari 400's keyboard is so limited is that (unlike its more expensive sibling, the Atari 800) it was intended primarily as a gaming console. Originally it wasn't going to include a keyboard at all.
However, when Atari saw "Star Raiders"- which they predicted, correctly, would be a "killer app"- they realised a keyboard would be required to play it, and included one.
I don't think the keyboard was ever intended to be used for heavy-duty typing or extended use, just the occasional keypress.
Hi,
thanks for your feedback. The original design was constrained by the sizes of Choc keycaps available at the time (1U and 1.5U). Also, a spacebar traditionally takes up a lot of real estate on the keyboard, so we opted for a slightly higher number of switches in the available space.
Together with FKcaps we have now invested in additional keycap designs and updated the keyboard to have the traditional typewriter stagger that most people are used to. The spacebar is still split in 3, but with a longer 2U bar at the center, where a percentage of people are used to hit with their thumb.
Here is a KiCAD render of the new part (without legends): https://mastodon.social/@mntmn/109841319638929298
My latest gripe about laptop keyboards is that so many of them now have the power button in the top row of keys, and somewhere in the middle, to boot. Likely an errant click would trigger sleep mode or an "are you sure?" display, but could all be avoided by not fixing what's broken, not shuffling important buttons around to be iNnOvAtIvE...
> something rendered by someone the first time they tried Blender
You might be on to something. The picture does look like a bad render actually: All edges are razor sharp, the material is strange-looking, and the lighting is unnatural (no shadows, the whole thing is glowing). Last but not least, if you look at those strange black circles and the square (???), they look blurry, like a low-definition texture.
I'd vote somebody doodled it in SketchUp...
[Author here]
> I'm SO disappointed!
In my defence, they're pretty cheap now. I bought a £150 Core i5 T420, liked it, so bought a Core i7 one. I have a quad-core W520 which is lovely but about as portable as a brick. And most recently, I bought a mint i7 X220 with docking station to replace my well-travelled but battered i5 one.
sadly, my x220s have started to give up ghost. Battery replacements are hard to get, or rather, 'it's hard to get battery replacement that's not a horrible, fake shit'. Also, one screen has already packed up and it proven near-impossible to get a _true_ high-resolution IPS replacement, only horrible tfn ones are available, unless you're prepared paying up to 100 quid or even more, which becomes silly. Truth is, I just love it for the keyboard (not great, but good) and number of ports which by today's standards is obscene. And the fact, I can swap sdd in 5 sec and have a different machine.
This post has been deleted by its author
From the project Web site
"Balthazar as a name for a laptop draws its associations from the Three Wise Man known as a Magus from Babylon and a Professor Balthazar, a 1970's cartoon character..."
And here's me thinking it was named after van der Pol. Hope they bundle a maxima or reduce port.
If it costs extra to swap out the keyboard that comes with it, somehow purchase a Lenovo replacement part, and use an adapter, to have a standard keyboard, this is a sure way to get a vast number of people to dismiss the unit out of hand. Now, of course, if people want as cheap and powerful a laptop as possible that is maximally conformant with existing standards, so they can easily buy lots of softwere for it... they're going to get a Windows laptop with an x86-64 processor.
So I suppose they're figuring that this product is for the adventurous. And if it's aimed at children in schools, perhaps the idea is not to endanger their health with a non-ergonomic keyboard.
However, I am being slightly unfair. If I were in charge of such a project, I would also design in a bizarre and unusual keyboard! However, it would be one more like a standard typewriter keyboard, rather than less so. I would design a keyboard like that of the Tandy Model 100 portable computer - with the same basic 44 keys for printable characters other than space as many office typewriters, including the Selectric I and II. The characters on the missing keys? I would use the CTRL key with keys other than letters to allow them to be typed. As well, all the control characters that aren't achievable by CTRL-A through CTRL-Z would be assigned to the control shifts of the number keys; ctrl-zero being NUL, and ctrl-3 through ctrl-7 being ESC through US.
Possibly DEL could be ctrl-backspace.
And to make things even more typewriter-like, one would have to use ctrl rather than shift on the comma and period to get less than and greater than.
This way, both shift keys, backspace, and enter would all be handily in reach for touch-typists!
"It's probably just me, but I do wish there was an open hardware laptop for the high end."
I can explain why there isn't. The problem with all these completely open projects is that things don't work. Every one that shows up has a few missing features in hardware or software that someone has to get around. Take Pine64's PineBook Pro. I want one of these. It looks cool, it's low-power, reasonably fast processor. Yet even though it's existed for years, there are still lots of notes about which software will just not run and which software has to be compiled with some patch on an anonymous git repository, that even when using the manufacturer's own docking station, not all the ports function correctly, and hardware-specific operating systems with unclear maintenance. I still want one, but I can't pretend that I could give this to anyone nontechnical and expect them to use it. This project will have similar problems.
For something I expect to tinker with, a cheap device can be explained away as a toy that doesn't have to work and can require more manual maintenance. If I'm spending a lot on it, I expect more things to work. Theoretically, a company could charge more for something and use that money to do the work that the cheap open platforms didn't, and Liam's suggestion of the Framework laptop is such a project, but it's worth keeping in mind that that's just a good laptop with relatively open hardware, not a ground up redesign. That project is a good one, and I'm happily typing this message on the one I bought, but it's not in the same realm as projects like these designed for completely open hardware and software.
From Ken Hagan,
"Nobody's stopping them. If their hardware needs better software support, I'd say the onus is on them to commission that software."
--------------------------------------------------------
Congratulations.
This same, exact attitude is precisely why the Pine64 organization is in trouble.
What is troubling is that you are, by no stretch of the imagination, in the minority when it comes to being accepting of the fact that it is completely acceptable for an organization to "...put a half-baked project out there...", demand good money for it, and then disavow any further responsibility for its maintenance and upkeep.
You are perceptive enough, one would hope, to note--and realize exactly why--the word, "project", is used, as opposed to the word, "product".
If it's not too much trouble--and what's been written so far indicates that "...too much trouble..." very much includes information you'd rather not read or have--, I suggest you force yourself to read the following, including following all links, and, most definitely, make absolutely certain that you read ALL comments from respondents to Pine64's blog.
Consider this, the next time you decide it's quite appropriate to be an apologist for Pine64's unique business model...which has worked for them for years, but seems to be approaching its end--precisely because of that "unique business model".
-------------------------------------------------------------------------
"More crucial than what you do know. or what you do not know, is what you do not want to know."--Eric Hoffer
"Thing about Pine is that they don't do software, like at all. They depend on the community, so unless some user decides to fix the issues nobody will."
Yes, that is the issue. For me, that's acceptable if I'm using the device for entertainment, and for me playing with tech and fixing things can sometimes be entertainment, but I tend not to spend much money on things for that. I can't buy such a device for work where its reliability is important, and I can't suggest anyone else do so.
Open source software is in the state it is because I feel comfortable giving someone a Linux installation on normal hardware and expecting that they will be able to use it (maybe after a tutorial) without it breaking on them. Maybe it won't have all the applications they want, or they'll have to get used to LibreOffice having its options in menus instead of a ribbon, but it's not going to fail outright. I don't have that confidence in open hardware projects given the history they've had, and I alone don't have the ability to fix that. The community's efforts in getting there are slow and while I support them, I can't and don't count on them. This may change in years to come, for example I'm happier giving someone a Raspberry Pi as a production machine than I would have been when those were newer, but it's not in that state yet.
Importantly, it's not just Pine64 doing this. It's all the open hardware projects based on nonstandard hardware. Even the Raspberry Pi has been missing hardware video acceleration in things like Firefox, and they're by far the best out there. Every open hardware device in my experience requires its own custom operating systems with weird patches, none of which I'm familiar with because I didn't build them, and maintaining that is possible if you do a lot of research, but not really what I plan to spend a lot of time doing.
Weird designs like these always seem to be 'designed' for (and probably by) the 20 or so ultra-hardcore emacsheads in the whole world whose fingers never ever leave the keyboard and who vehemently hate mice and trackpads (and who therefore use somewhat masochistically quirky 'pointerless' window managers that are equally difficult for normal folk to get the hang of; not that keyboard-friendly WMs are a bad thing, I do hasten to add, but hostile learning curves really aren't so great), and this one looks no different. The grudging (and I'm sure awkward as hell to use) placement of the touchpad above (not below) the keyboard seems to reinforce this suspicion. I'm only a little surprised that they haven't also gone for Dvorak keyboard layout by default to complete the stereotype…
(If all of that works for you, then great, but you're not really going to get many takers at all among any sort of (relatively) "normal" [Linux] computer users, I suspect.)
No. Most of these kind of keybr0ad "designs" make EMACS harder, not easier.
They are nod "designed", per se, rather they are Marketing Decisions, made at The Highest Level (whatever that means, but you can always hear The Caps in their voices), usually by idiots who would have difficulty arranging food on a plate, much less making ergonomic design choices for computer I/O.
Consider 1982's near ubiquitous Key Tronic Corporation 101/102 key keyboard that we are all familiar with. It was based on the Knight keyboard, which was specifically built around the EMACS keybindings (with more than a little help from the Stanford Keyboard). See also the Space Cadet keyboard and the Symbolics keyboard.
I would love to have a Space Cadet keyboard, but if such could be found at a sane price, would probably require custom hardware to interface it to a "normal" PC. At the same time, the SCK is not perfect; they screwed up the positions of the function keys, and put weird icons on them (pointing fingers and roman numerals).
"meh, why all those keys ? use a stenograph keyboard."
I have various chorded keyboards. They work fine, but the Model M still gets ASCII from brain to computer faster and more accuratly.
"better even : only a 1 and 0 key. use binary encoded ascii."
I lived through the days of entering in code by toggling front panel switches. Trust me, you don't want to go back there.
I'm happy to see more activity in the OSHW computer space and it's great to see the MNT Reform mentioned. However, we also offer an OSHW FPGA processor module (Kintex-7 based) which we demonstrated running Linux on RISC-V in September 2022: https://mntre.com/media/reform_md/2022-09-29-rkx7-showcase.html
Yet Another F***ed-up keyboard! You cannot comfortably use Control+Alt+Shift with a single hand on that keyboard. That keyboard has no [Insert] key, no [System Request] (aka "Secure Attention") key, no right-side [Control] key, no [Print Screen] key, no [Scroll Lock] key, and they've stuffed a [Delete] key between the Z key and the left-side [Shift] key.
Whomever(s) designed it had the "If I don't use a particular key (or key-combination), then nobody else uses it, and I can get rid of it" mentality.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
