Miniguru progress

[fruitsaladfingers]

I hope things are going well. Any updates worth mentioning?

[lowpoly]

I received new pcbs 2 days ago. Have to figure out which parts I need.

I also try to set up a "firmware group" here at the office. Not sure if this will work out.

[fruitsaladfingers]

Did the firmware group prove to be successful? I hope you've made progression on the ultimate keyboard. Please let us know if you've got any progress/updates!

[lowpoly]

Firmware Group: not yet. But I'm optimistic.

I had another case milled (the third one) which is waiting for finishing. Case will be matte black, hopefully powdercoated. The finishing has to happen this month as I have to move the workshop soon. This month or next. Still not sure when. I also had another batch of pcbs made. Now I have to figure out which smd parts are where. Not easy at all.

Speaking of a black case... I have to figure out which keycap color I want. Probably black, but, this is too tacky right?

http://www.alibaba.com/product-detail/M ... 0.0.fmnglA

Chromed plastic was never convincing I guess.

There seem to be more pbt keycap sets available on Alibaba these days. Last time I checked there were none.

[fruitsaladfingers]

I appreciate the update and efforts in getting it production ready and wish you the best.

I imagined myself using blank black keycaps, so I'm not particularly in favor of the set you linked. But I haven't seen them on a board. You'd have to alter the spacebar for any different keycaps though, right?

Thank you.

[fruitsaladfingers]

Hey lowpoly, just wondering how things are going - hopefully well . How's was the move? Are there any progress updates you can share? I understand you're busy and appreciate the work you put towards making this awesome board, I would love to own one.

[lowpoly]

Hey fruitsaladfingers,

the move is still in progress. Still working on finishing things for that. Many are not keyboard related.

During the last weeks I put together the "white case" prototype which will be a RevB. Firmware is working (with a fixed layout), bootloader not yet. The initial plan was to put RevC into this case but for various reasons this is not going to happen. Because RevB uses a different mouse switch I have to mill new buttons. Working on that right now. Unlike the very nice RevC and RevD buttons this is trial and error. With the new buttons, RevB should be mechanically finished. Only thing missing for a completed board is a working bootloader and loader app then.

This is what the host app looks like right now:



The "layout handles" were the last thing added here. In the upper right you can see how backspace is split into two keys by klicking the handle. Still a lot to go, but also a lot there already. For ex. when changing keys it will use the big Excel file I made with the different language layouts. This will also help with typing Macros. Based on Node Webkit the host app should run on every platform. "Drivers" will be the more difficult thing here.

Still building the firmware group, I've been not that successful yet.

I have another CNC-milled Rev D case. Going to sand it this week.

Next step: soldering Rev D pcbs (switch and CPU). And having the new case powder coated in black.

[pjmelon]

Mouse switches: currently one of these here:
http://www.mouser.de/Search/Refine.aspx ... ewSearch=1

A middle mouse button would kill the design but both mouse buttons are part of the matrix (= programmable) so you can put middle click on the right button if you need it all the time or use for ex. one of the keys in the lower right. Fn + right button should work too, which would probably be the best solution.

I am looking at a custom IBM model M and liked the switches that you used. I am having issues determining which of the 4 switches to chose. By the looks of it you didnt use the hinge lever variant. If this is the case how did you make the switch part?

[lowpoly]

URL says it's the 653-D2FS-F-N.

The switch part is milled from aluminum, the hinge socket in the case too. This requires some precision, not easy to do by hand.

[shreebles]

Awesome that there's still progress on this. I'll probably buy one once it's finished, however long that may take. It looks like a great work keyboard and a perfect notebook/tablet companion.

[lowpoly]

Yesterday I finished the Rev B version (the white one) electronics and mechanics.

Firmware is unfinished, the bootloader doesn't work yet. And the connection to the desktop interface isn't there. Pointing stick acceleration could be faster.

But writing works, as you can see here.

Not sure if I'll look at the bootloader issue because finishing Rev D is next which requires firmware changes anyway.

While this only is a small in-between step, it is also the first time I have a finished keyboard I can actually use.

Mouse buttons being part of the matrix seems to work too.

Edit:

[lowpoly]

Looking at that photo I find the different radiuses of the caps and the case corners not that good...

In Rev D this is solved.

And I'm not even sure this is a good thing. Because having it here in front of me I don't see it and the bigger case radiuses are actually nice.

[Menuhin]

Hey lowpoly, please keep up with the progress.
I hope I can see the finish product and join the group buy some day in the ... future 6 years maybe?

[lowpoly]

6 years? You are optimistic!

I have a new shop now but stuff is still in storage. Have to move it the next days and set it all up again. This will take weeks.

[usopia]

...so we can order the Miniguru in 6 weeks!?

[Phenix]

if its going into production - split spacebars? ISO and HHKB support?

[7bit]

Report from the Keyboard races at Circuit Board:

With only 624 of 3000 laps to go, the Miniguru Special of team Lowpoly is taking the lead from HyperMicro of team 7bit Racing. Since currenly all Keyboards are in the pits, there is no keyboard on track.

[lowpoly]

if its going into production - split spacebars? ISO and HHKB support?

If...

No split space bar, there's no space for two switches and stabilizers AND the mouse buttons. But there's the touch controller so it should be possible to detect what side of the space bar has been hit.

ISO and HHKB is already there, in the pcb too of course:

[lumogas]

How was that saying? "Good things come to those who wait?" Looks like it might be true in this case...

[fruitsaladfingers]

Hi lowpoly - is there anything worth updating us on? How was the move?

Will it be a merry Christmas?

[lowpoly]

The move is finally done since last Friday. Now I have to set up everything. That alone will take me way beyond Christmas.

I just started a new effort to build a firmware team. If this fails too I'll need outside help. Meaning one of you.

[Menuhin]

One quick question.

One my ThinkPad keyboard, I use the Trackpoint most for scrolling, even more than pointing. For scrolling, I hold the middle button when moving the Trackpoint up or down.
That comes back to my question: There is no middle button in the prototypes. How is Trackpoint scrolling implemented then?

[lowpoly]

You can program any keyboard button as a mouse button. Or press the Trackpoint harder for scrolling, the pressure is programmable.

[fruitsaladfingers]

Hey lowpoly,

Hope this finds you well. Since it's been a few months, do you have any progress to report on the miniguru? Excited to finally own this awesome board.

Thanks!

[Sigmoid]

No split space bar, there's no space for two switches and stabilizers AND the mouse buttons. But there's the touch controller so it should be possible to detect what side of the space bar has been hit.

How bout this? Split spacebar, and no dedicated mouse buttons. Fn+left space is left button, Fn+right space is right button...

[shreebles]

Short travel makes sense for mouse clicks. Mouse buttons ftw.

[lowpoly]

Short travel ftw!

No progress at all.

Moving the shop is finished. Some stuff left but that can wait.

I haven't been able to find someone for the firmware programming. I won't have time in the near future. Not sure how to proceed now. I could either completely open source the project, or partially, or find a coop with one of the other vendors. Kickstarter is no possibility right now because of time constraints.

Any opinions?

[jonlorusso]

completely open source the project

Any reason not to do this? This would be my vote.

[lowpoly]

The reason is that company money went into this. So as long as I have a chance to get some of that back, I have to try it.

A friend knows a guy who does firmware programming. Currently looking if he's interested.

[Daniel Beardsmore]

I find this idea concerning. It may be easy to get the first 90% of the firmware working, but it's all the tricky details that take the most work and cause the most problems for customers, and it seems illogical to start this process all over again when there will be no benefit from it.

The Miniguru is an excellent use case for my idea of a modular extensible firmware with an open, text-based configuration protocol (maybe through COM port emulation — think working with Cisco IOS but at an extremely rudimentary level that implementers can't get wrong) that is independent of platform, making the product able to be used under any OS or with any alternative configuration application designed for this protocol. (That is, layout changes are passed in using a simple protocol without any recompilation (that can also respond with the current configuration) and the different aspects of the firmware (key acquisition (from matrix or from protocol conversion), key remapping, and host communication) would be entirely segregated in replaceable modules.

However, it became clear that there is zero interest in this idea, so we'd be looking at another proprietary firmware and be back to square one. Unless this friend is working for free, you'd be spending more money to get firmware that already exists and is likely to be superior having already gone through an extensive development process and would share its future enhancements and bug fixes with yours (and vice versa).

If there is going to be new firmware, it seems silly for to be anything else other than a) open source, and b) platform-independent programmable and extensible. Anything else is plain madness.

Commercial firmware is a perpetual disappointment, with everyone and their dog thinking they can achieve something good yet never, ever, ever understanding what we actually want to achieve. Getting that one universal open firmware would allow anyone to make a keyboard with quality firmware. Software and firmware should be a solved problem: write once and enjoy everywhere, while it's clear that we will never agree on the switches or layout and that's what the Miniguru should stand for: the hardware.

I think that was a few more than 2¢.