For some reason or another, I keep gravitating towards these Asus Ally vs Steam Deck videos that are making the rounds over YouTube:

The talk about controls came up and it made me think of what I'm doing for Tetent. I've also been annoyed that I've got everything except the (in my opinion) seldom used joystick button. 

I'm taking advantage of the fact that, unlike with Tetent, you can't actually use the joystick AND buttons at the same time, and I'm putting the joystick press on the end of the Tetrinsic active area:

I've also corrected the position of the XYAB buttons. If you were looking at Tetent from the top, the layout shows the left side of Tetrinsics. For TrueGame, the left side would have left, up, right, down instead of X, A, B, Y. I can't find any actual numbers, but I'd imagine that the order of most to least used of the 4 buttons is A, B, X, Y. Like with Shift, there's some force between position 4 and 5, and 5 is "spring loaded" as if it was a button perpendicular to the sliding axis.

Replace Joystick Button, Analog X and Analog Y with Left Click, Cursor X, Cursor Y respectively and there's some kind of Steam Deck trackpad alternative.

So I've recently looked into my main-display and 5G celluar options for #Leti: the Tetent + PC handheld concept that doesn't yet have a project code. Now, I just looked at what buy-it-now options there are, especially 5G enabled options, in the portable x86 market (PDAs, laptops, tablets, UMPCs) and it was painful to see. 1080p 14"+ screens (so low PPD), costs £1,500+ yet doesn't have a 2TB option, 16GB soldered RAM... STILL! In 2023! It doesn't seem like much has changed since 2020.

Leti is expected to be around the size of the steam deck, and if I can find a case similar to the steam deck one, it would be a more portable and convenient solution than any laptop, tablet and not-actually-pocketable-UMPCs. Someone at The Verge just wrote about the elastic strap that allows such portable convenience.

I think this realisation promotes Tetent from "Is this too big to fail?" to "It's too big to fail.", which basically means it's an un-shelvable / un-cancellable project (unless something better shows up). I've tried my best to try and sidestip Tetent/Tetrinsic entirely (because I'd rather not spend X hours a day for Y months mining for a solution), but I cannot offer myself any ideas for an acceptable alterntative solution.

• If Tetent is plugged straight into Teti, I can see myself opting for the 15.6" monitor instead of the 2.9" 1440p most of the time.
• If I'm using Tetent remotely, there doesn't seem to be a 120Hz remote desktop client. Even if there was, I'd still require a good internet connection, which is unlikely.
• I don't run Linux applications.

Thus, it seems that Tetent without the 1440px, Tetent Taic and Tetent UMPC are the 3 versions that sound like it'll be worthwile to invest time into.

[17 May: I've now made this idea into its own project]

[21 April]

I really should get to a MVP Tetent out of the door ASAP so that Reality(TM) can throw me under the bus ASAP and tell me why my solution doesn't work, even when mitigating as many possible negative scenarios as I can.

Why? 

It seems I've passed the proverbial event horizon; I'm losing time if I work on Tetent, and losing time if not working on Tetent, simply because I'm thinking of it and wondering how I could make it better or adapt it for a new usecase. 

In terms of making it better, yesterday I talked about the plan for the Tetent Taic, a photovoltaic variant of Tetent that replaces all the LCD screens. For possible usecases, today I'm thinking of a UMPC.

First, some history

I owned a GX1 Pro for... about a week, which I called Aeti (pronounced "Aey... Tee-yai") because of the powerful NVidia RTX GPU hype (eg 2080Ti, 3060Ti) and it was hopefully going to replace the second-hand Sharp HC-4100 (named A4100) that I wrote electronic notes on (instead of the little A7 booklets I was using previously, hence the whole "A" thing). Yes, the technological difference of Windows CE + 8MB of RAM + mono-LCD to Windows 10, a 1920x1200px truecolour screen, Thunderbolt 4 enabled, 4C8T, 16GB RAM device was very nice to think about. 

Unfortunately, there were small issues that compounded into sending it back, namely that a) the WiFi connectivity was directional, b) the screen didn't bend 180 degrees (aka, to my face when held, but to my chest) and c) there was only 16GB of RAM.

This was in the time where I was struggling to obtain parts for #Teti [gd0022] and was faced with a decision: Wait it out, or sell off the components and double down on portability, skipping 16" laptops and going all the way down to this 7" device. While I returned it, I still loved the thing and the possiblities, though it doesn't look like OneNetbook is coming out with a second edition now that their hands are full with the OneXPlayer. 

The idea

You know how I started building Teti "because I can't make a 2 in 1 laptop"? Well, a UMPC in a Tetent form factor could be even better than the laptop form factor for me. 

I've been occasionally wondering if there was a solution where Tetrinsic was integrated into a traditional laptop, but this was before Tetent Concept3. The idea for the laptop was to have each half spring loaded (like a pen) and pop up out of the chassis to an ergonomic angle.

Now, the idea is to create something like #Tetent TwySize [gd0040] but using an x86 CPU.

Computing hardware

Framework motherboards now exist:

At this point, they're essentially a suite of (relatively) high performance single-board computers that seem to continuously get the latest chips. 

When I was researching the GX1, I notieced that the 1135G7 and 1165G7 (the GX1 Pro has the similar 1160G7) were rather close in benchmarks, and that the former was the value buy, if given the option. Considering that I ended up upgrading from an 7500U 2C4T laptop (Deti) to a 11600K 6C12T desktop (Teti) and that desktop only feels 33% faster in general despite being 100% faster in single-core benchmarks, I'd still say that the £299 1135 is the better choice, especially considering the DDR4 to DDR5 price difference (see image below). A nice thing is that the 1135G7 can take up to 64GB of RAM, just like I've got 64GB in Teti! Many previous laptop chips seemed to max out at 16GB for some reason.

The price (and the difference) is much lower from 3rd party vendors:

Anyway, it's not exactly small though. It took a while to understand, but looking at the 2D schematic, the board is like 104 x 233mm, though a redditor has measured it and it's more like 114 x 233.

[Edit 5] It turns out that the listing itself has this information, and it's 233 x 112 x 6.8mm. [/E5]

[12 May: Edit 2] The other option that makes the most sense is AMD's brand new, hex-core 7640U, which can support up to 256GB for those... erm... gigapixel processing algorithms or something? The site says shipping late Q3, so I'll assume "sometime in October" which is early Q4.

Like I said, I uprgaded from a dual core 7500U to a hex-core 11600K, and there's actually not that many multithread applications I run (and if I do, the task is usually short anyway, like batch resizing 10 images). Rendering 1MP Fusion360 images is the most multithreaded thing I do. Thus, I recommend saving the £250 difference to the 7840U if just after the 2 additional cores. Looking at benchmarks, an 11600K and 7640U are effectively equivalent in performance.

The 7640U has 8 RDNA3 iGPU CU's and the 7840U has 12, and it does seem like they scale well from these fresh-off-the-press benchmarks (66% the CU's = 64% the performance).

[/Edit 2]

The screens

This is what I actually came here to talk about. I was going to start this log, mentioning that I'm planning to take into accout the possibility that somebody would want to have a dual-screen Tetent solution. The Toshiba chip has 2 I2C addresses, and many SBCs support dual monitors. I'd expect they'd use larger (albeit 60Hz) screens instead of the 2.9" square.  It'll still be an exercise for the reader to actually design a UMPC / cyberdeck, but I was intrigued about possible components for such a solution.

After all my research, I've found 3 candidates:

1. LS055R1SX04 
   1. 5.5" 1440p: 97.1 Pixels Per Degree (PPD) from 26cm
   2. IPS
   3. £17
2. BB055FHM-T00-6P01 (guesstimate, since it's not mentioned in listings) 
   1. 5.5" 1080p: 72.8 PPD 
   2. OLED
   3. Touchscreen
   4. £31
3. LS060R1SX02 
   1. 6" 1440p: 89.3 PPD
   2. IPS
   3. £23

Like a normal 2-monitor setup, the idea is to have the screens side-by-side (to look like a 32:9 with a line in the middle), mounted on the speaker side. It swivels out 90 degrees (similar to how I rendered it for the #Tetent TwySize [gd0040]). 

My 1st recommendation would be panel 2 as it's overall the best solution. The second one would be panel 3 for more screen real estate and text sharpness. Remember, Tetent is naturally held a few centimetres closer than a normal phone, and desktop OSes like Windows show more content than iOS / Android.

[15 May: Edit 3] 

I've scanned Panellook... and found nothing, but I scanned my mental archive of AliExpress products and remembered another candidate, the H381DLN01.2 (not .0, which is maxed out at 100nits). It's a 3.81", 1080x1200px, 76.1 PPD, OLED, 90Hz panel that's £27.50 on Aliexpress, and the plan would be to have 4 in total as a kind of mini video wall. I didn't see this because I set my search query to min 250 nits on Panelook and this screen is 100 in their database, but the listings say that it can go up to 360 nits on some and 400 on others. 

The benefits are that it's the only high refresh option, it's OLED, the active viewing area takes up most of the available vertical space, a low-poly curved monitor is possible and there shouldn't be any screen orientation issues. 

The drawbacks is that there would now be 3 black bars across the active area, and it's near 2X the cost of the next most expensive option.

It's possible to see those black bars as a feature instead of a bug though if you see them as content seperators. All my productivity workflows -- other than Fusion360 -- would either fit just fine in a single 1080x1200 block of screen (e.g. MS Word), or has a resizable, window-in-window/panel user interface (e.g. Visual Studio) where I could just arrange the panels such that the digital boundary occurs at the physical boundary.

If I was going to make this UMPC, this is the option I'd try first since I'd 1) very much like to get out of this 60Hz lifestyle and 2) the low poly curved mini video wall sounds both cool and productive.

[16 May: Edit 4] 

Perhaps, the best compromise would actually be a single, centered 5.5" OLED and dual 3.8" side panels. 

• It's unlikely that someone would game with such relatively large bezels cutting though the scene:

• The main reason I want high refresh are for moving-text situations, such as scrolling a research article / file explorer / AliExpress page. These are also situations well suited to be placed in a side panel screen.
• Due to the stubbornness of YouTube, 60fps is the top limit for video content. Thus, wether I'm creating or consuming content, 60Hz would be the most I'd need for a screen. That content will also be comfortably centered, instead of on the left or right half of the device.
• They're both OLED and of similar size (Vertical of 68.3mm center for the 5.5 and 72 for the 3.8), so the sizes of UI elements should only look slightly smaller on the side panels.
• 3 panels is cheaper than 4 (or 5, since 4 is £128 + 20% VAT, and 5 is £137 + 0% VAT due to the >£135 limit thing). Including VAT, it's £98.
• [Edit 5] Still, a 32G/2TB 7640U 2x1080x1200 + 1920x1080 OLED system would cost around £1,000, which is about the price I paid for the 16G/512GB 1160g7 1920x1200 IPS GX1 Pro.

[19 May: Edit 6] 

I've just found out about the LS063R1SX01 for £34, which is a 6.3" 1440p IPS screen. It has a PPD of 85.3 and has a horizontal FOV of 30 degrees, which I found to be ideal from #T^2 TyMist [gd0138]. It can be configured with or without the backlight, so a possible sub-project could be a DIY local-dimming LCD. A pixel pitch of 2mm will result in 2730 local dimming zones, and probably could obtain true HDR and daylight-readable brightness levels.

Ok... is that all the screens?

No, actually! For the low low price of £99, you can actually get a 5.7" 1440p OLED for an inky black, 93.8 PPD experience! It seems to be the AMS567JD09, which has a PenTile instead of RGB, so I think the actual resolution would be around 1080p levels still.

Wait... if PenTile is the reason why the blacks on my 2021 Realme Q3 Pro look blue-tinted in any sort of ambient light -- making it look worse than my 2012 PSVita -- I want to MAX avoid it; I do NOT consider that Q3 screen "OLED" as the ambient black looks only marginally darker than an IPS screen in most lighting conditions. 

Places a drop of water on the screens of the Realme and PSVita to be able to see the sub-pixel layouts...

Now I've got to make sure that the 1080p OLED screen I found is RG/BG or RGB. This site seems to be a good resource for finding panels that I might actually be able to find on AliExpress.

As you can see, the PenTile configuration is more common than RGB stripe when it comes to this size range. The AliExpress listings I found don't specify what the screen is, but the y have a squigly and bendy cable so it's probably RG/BG unfortunately. 

That T601F2160OM-01 is something I'd really like! Have you seen my brand image? I will totally run the side screens at 1080x1080px if it meant that I could have a 1:1 / 2:1 / 1:1 screen setup. Unfortunately it seems out of stock at the moment:

It's not like any of the Framework mainboards I want are going to be in stock anytime soon either, so it's fine for now...

Oh wait I just got a message from the seller:

The display is rare in the market now, if you need 5pcs+, 
we will help you find it.

 Hmmm... 2pcs maybe, since it's good to have a spare, but 5 is a bit high... there's also the BOE BO559FF2KM on Panelook, which seems like the same thing.

As for the H546DLB01.1 5.5" OLED that isn't PenTile, I was wondering why it was slightly more expensive than the £30 one I found when I was looking for screen options.

Well now I know.

Still... I'd really like that 2160x1080px edition. There's some in stock somewhere...

!!! It's those guys?!

Nohhhhh is it these ultra expensive flexible ones? Panelook doesn't mention anything to do with flexibility in its database. I think the only telltale sign was "Pixel Format: Sub-Pixel Rendering" instead of "RGB" or "PenTile" If I was spending that kind of monopoly money, I'd at least get more pixels:

[/Edits 3 - 6]

Dual-screening, with a small-enough SBC, should mean that the smallest dimensions could be 85x270mm for 5.5" screens and 85x305mm for 6" screens.

Considering the Framework mainboard size, and opting for the OLED screens, I'd assume the weight would mean that it's more comfortable to be able to rest the device on something else while still being able to type. Thus, a non-centred approach that holds Tetent high above the surface is optimal.

For this Framework based solution, it's likely that additional ports / (micro) SD card slot(s) would be found in that dark grey bar at the bottom, likely coming off the sides. Ideally, 2 TB4 ports will be exposed (supported by both intel and AMD) and one of the Framework USBC ports will be used for a power delivery/SD Card/Microphone hub, leaving one USBC port free for extensions.

The battery

In both cases, it's assumed that 3 x 5Ah 70x90mm cells are installed in the screen section, providing 55.5Wh of capacity (essentially the same as the stock Framework laptop). There are some 6Ah cells out there too, allowing for 66.6Wh. I was hoping that there'd be enough space for the thumbs to actually be able to use the edges of the screens, but that 15mm bezel is looking quite large when modelled. It is also assumed that the bottom reflective LCDs found in the Tetent Concept3 model are ommited.

[16 May: Edit 5]

Looking at listings, a single 357090 cell is 64g. That means that the 3cell battery would be 192g, which is slightly lighter than Framework's 217g battery with dimensions 239mm x 93mm x 6mm. However, even though this is a handheld device, I'd rather increase the weight by an additional 192g to allow the UMPC to go twice as long on a charge. An ultra-mobile PC isn't all that mobile if I'm tethered to a socket and isn't much of a PC if it has no power. The 6cell battery size will likely be about 216mm x 92mm x 7.2mm.

I'm aiming to be at or under 800g, which is the weight of one of my touchscreen portable monitors. My estimate totalled 772g, so there's hope!

[17 May: I've now made this idea into its own project]

[Edit 1] 

I've been staring at the design more, and I think I'll be going with the white bezels. The reason is because there's still much more white visible from Tetent itself than the bezels from the first person perspective, and the bezels themselves are much less obvious from the 3rd person perspective, allowing the design to have a centred, black square when all screens are off.

[Edit 1: End]

I thought the monitor integration design looked like a star, and so I thought the Tetrinsic background would work perfectly for the render. Now I'm playing Sonic: Reach For The Stars, agreeing with the lyrics.

Internal features and software features are still quite likely to increase, but externally, I don't think there will be any visible changes from new features. I'm hoping the BOM is around the £300 range. 

Speaking of internals, I've done some research on USB PCB creation, electromagnetic radiation and the pros/cons of 2 vs 4 layer PCB's, and I'll be playing it safe with the 4 layer PCB for both Tetrinsic and Tetent boards. 

I'm also going to take advantage of each Tetrinsic having USB capabilities and just wiring one Tetrinsic to 1 port/device, and then having a selector similar to input selection menus on TVs. Right now, the selection would be [USB Port 1, USB Port 2, Internal SBC]. I'd need some way to toggle between using the port for Tetrinsic I/O and using it for SBC I/O.

I'd also like to mention that the FOV of the 1440px display is low enough that these cheap magnifying glasses (with led lighs) don't cause any annoying chromatic abberations of the content. The focal distance, which is a tad too short for laptop/desktop viewing distances, is right in the zone needed for this handheld mini-monitor. Additionally, since I don't have to look down occasionally for a mouse/keyboard, the potential chance of eye fatigue (from constantly changing focus) is reduced.

The only main concern now is: Will the solution fail, due to ergonomics? 

I consider everything else from PCB creation to firmware programming to the mental overhead using this device as "Skill Issue". It is only if Tetent is damaging to my fingers where the solution will fail. Unfortunately, in these 15 months of research, I've noticed that the tolerance for an ergonomic solution is quite low. I guess that's somewhat expectable, since higher bandwidth electronic signals are usually less tolerant to PCB design imperfections too.

adjusts glasses

The bezel is larger due to the LCD chin, and the dark grey was chosen as a nicer, mellow colour (compared to white or black) to stare at for hours. 

I do acknowledge that black looks better from a purely aesthetic position as well as matches with Teti's 4K monitors:

I'm not letting this feature die off so easily.

- Officer of Feature Investment 

Impressive speed, but I've had to make 2 alterations. The black bezel will be chosen and the nauseating rounded corners were replaced with small filleted chamfers.

With the grey, there were 4 different colours, when a 3 colour palette would look sharper and more uniform:

And, as you said, black bezels match with the monitors for Teti.

- From the Aesthetics Compliance Division

Dear Feature Investment Officer,

The mutant mould "design tweak" (pictured below) to include the 1440px screen (and a second USBC) best be a late April Fools joke and you're about to say sike soon.

There's no way I'm signing it off when it originally looked as simplistic and moden as this:

"It's a 3rd person perspective; from my view, nothing's changed." is not an excuse. This may be an amateur project, but will adhere to professional design standards.

2 centered USBC ports and a centred logo, or another solution that obtains XY design symmetry, will be acceptable.

- From the Aesthetics Compliance Division

The first thing that's changed is swapping the location of H and L. The reason is because of this article of why the creator of Workman had issues with Colemak, talking about bigram frequency.

From the fingers I have, I get force feedback but not positional feedback. I also expect that less movement translates to less mental overhead. This is the reason I'm trying to keep similar characters under the same position, and why it's 6 zones 7 levels instead of the other way around. 

With this small change, something (that sounds like a common phrase) such as "they were" only require the fingers to be in zone 1 or 5. If not using Tetent TestCut, this would be 2 chords: "they we" and "re".

Other changes

I'm also thinking of splitting up the weights into blocks of 2, since when the finger is planning to go though multiple levels, it could be difficult to keep track of the haptic vibrations. Additionally, the top level is kind of light, so spacing things out could prevent mis-clicks into level 2.

1. 40g
2. 52g
3. 60g
4. 72g
5. 80g
6. 92g
7. 100g

This changes the haptic of, say, Ctrl, from BZBZBZBZBZBZ to BZ-BZBZ-BZBZ-BZ. I'm running the mental simulations now and it's a substantial improvement for knowing what level I'm on. 

I'm also swapping Windows and Enter. There's not a lot of keyboard shortcuts that use the Win key, but I can imagine pressing the enter key much more, especially in an order sensitive manner.

Thus, the updated Tetent TrueType layout:

Simulating typing on a keyboard that doesn't exist:

My goal was to cycle between fingers on every character, but sometimes I did get a chord in, such as "ll" and "e,". 

The errors are because I still somehow hit the wrong physical key when marking "I've pressed this key on the virtual Tetent". 

So far, I haven't felt any layout changes I'd want to make. Shift 1st was a major assist, as I just had to drag Finger5 back and boom I could get a capital G.

So I've been staring at the current Tetent model for over a week, wondering if there's anything that I'd want to add. The main question is if whatever I want to add actually requires a Tetrinsic change. This idea isn't that and concerns my possible use cases for Tetent. I was thinking of possible tasks I could get done if there was sometihng like an RPi 0 installed, but most of my tasks are still better done on my main PC, which currently is Teti.

I've got about 60 x 65mm of otherwise empty surface area on the back, and I know about the 2.9" 120Hz 1440*1440px screens that are like £17 on AliExpress (driver sold seperately). I also know that a cheap, small (ie sub 10") USBC monitor is hard to come by, with most being 13.3" / 15.6". 

At 120 pixels per degree (PPD), the same as Teti's 4K portable monitors from 62cm away, this 1440px screen would be 25cm away from my eyes, which just so happens to also be the ergonomic default position for my arms if held to face my face. However, that distance increases to 50cm if my arms are resting on my lap.

The screen brightness would be fine for indoor and low-light usage, but is unlikely to be outdoor visible. There's no plug sockets outside for Teti either though.

Anyway, the idea is to have the screen on a hinge and then USBC video on over the feed, which is exactly the same amount of pixels as 120Hz 1080p FHD. This is likely to be an extension project after I've got a basic Tetent that works.

I've been using Teti with a custom 1440*1440px resolution and no scaling for a handful of hours now, and it's actually quite usable, even at my usual viewing distance of around 68cm. The only main change was to set the taskbar to auto hide.

The peak bandwidth is 6.3Gbit, so I should only be limited by the time it takes to create a driver PCB and not the accessiblity of chips and/or their datasheets. Alternatively, and perhaps more ideally, there's support for 1080p 120Hz over USB 3.0; I know that the cheaper Windows devices especially usually have USB 3.0 these days but don't have video through USBC, and I doubt anyone would be gaming (anything intensive) on a 2.9" 1:1 aspect ratio screen. Hopefully, there's 12mbit of bandwidth spared for actually communicating with Tetent.

I'm also looking into seeing what remote desktop solutions are out there, mainly looking for a 120Hz capable solution and mainly because Me In The Past decided that USBC video out was a feature I could skip when buying my Realme Q3 Pro. I don't think the bandwidth would allow for such a solution though.

The rear now starts to look cluttered. The rounded corners make me think of the iPads, but not a fan of them when looking at them for more than a few minutes, instead preferring sharp ones:

There's also likely to be some... geometric challenges when it comes to folding it out:

[The Next Day] I'm going through with it, mainly for a subproject of fun inside a project for survival. (Tetent is my way out of an "All Routes Lead To Doom" situation.)

I've scanned Panelook and the LS029B3SX04 is indeed the best choice. The cool thing about this one which isn't seemingly supported by the 1-year-older LS029B3SX02 is that the duty cycle is designed to be under 100%. To balance maximum brightness while cutting down motion blur, I'm planning to have a duty of 50%. This is similar to the Sharp "240Hz" smartphone strategy, which actually just blackens the display every other frame.

It turns out that the full datasheet for the TC358870XBG is over 300 pages, which is the chip I'm planning to use since any SBC I could potentially use as a remote desktop client (RPI 0, CB1) only outputs HDMI, as well as the fact that there is HDMI USBC-alt mode. I never found the TC358860XBG (eDP to MIPI) datasheet of such a size, meaning that my custom display driving PCB options may have actually been limited to 7.2Gbps this entire time. I'm planning to use a HDMI switch to toggle between the thin client and USBC video, hopefully for 60Hz video wirelessly and 120Hz low persistence when plugged in.

The Toshiba chip is programmed though I2C, so I can use the Tetrinsics ESP32-S3 MINI for that without much issue. The PCB that drives the screen is also likely to be responsible for battery charging + power delivery (since I'm finding it hard to find a 15W li-ion charger with USB-breakout) and for the speaker DAC (since space is getting quite tight).

I'm also planning to use gestures to switch desktops, similar to what I thought up for #T^2 TyMist [gd0138].

More useful, but will never look as good as the non-1440px display design, especially when all screens are turned off.

I just wanted to do one last research round of less known / more unique input devices before I stop looking and focus on creating Tetent. 

Devices for content creators

Suprisingly, the Orbital2 is a device that costs 2X a SpaceMouse, which was already considered a high end device for 6 degree navigation. I don't think this device can even do more than 3, but does have 8 capacitive buttons. I thought, perhaps, it's the software integration that commands the high price, but many reviews say that its subpar. 3Dconnexion should look into something like this since it looks like they're leaving money on the table.

[Mar 30 Edit] WOUGH!!! 3DConnexion did create a spacemouse with 8 buttons, made between 2003 and 2009. It's called the SpaceTraveller and honestly it looks modern even today. Cost $199 back in the day.

The current SpaceMouse Compact seems like a downgrade in comparison to this.

[/Mar 30 Edit]

The XENCELABS Quick Keys is a nice looking device. This looks to be easy inspiration for an editing layout for Tetent.

It's a shame that there doesn't exist a trackball with the quantity of buttons that the Shuttle Pro V2 has. This + trackball was what I was looking for when looking for a replacement to my 12-thumbkey mouse. I don't really have the deskspace for a macro pad + trackball.

Loupedeck Products

I first found the Loupedeck+, but it turns out this company makes a few non-typical devices. I started this search to see if Tetent's £250 - 300 expected price would still have a market if it was a commercial product. 

The Loupedeck+ is targetted towards photo / video editors. There's many rotary encoders on this, which is to allow the user to input changes into multiple values at the same time (such as highlights and shadows of an image), but the thing that notably got my attention was actually the arrow keys in the bottom right of the device, implying that quick non-mouse navigation is important. This device costs a tad over £200.

At over twice the price, there's the Loupedeck CT which seems to have LED backlit buttons and LCD backlit dial + buttons. The video below looks to be a good insight to what functionality a power user would like from a dedicated editing input device:

Trackball Remote (and other handhelds)

So, inspired to look to see if there has ever been a trackball that has had a decent number of buttons, and I stumbled on this beautiful, modern looking InterAct SV-2020 remote:

The product actually looks better than the image on the box:

I can see that this was made in a time when USB wasn't widespread. Since there aren't any current or new input devices I know about, I had a feeling that the past would have unseen, interesting finds. I really like this one and it's a shame the idea died off in the past. Now I'm struggling to find a trackball with more than 4 buttons (remember 2 of them have to be used for left / right click).

[Oct 07 Edit]

Yesterday, I discovered the Elecom Relacom handheld trackball on Amazon:

It's one of the most expensive peripherals I've bought (I think only the Airberries have it beat, and I bought my spacemouse used) and its the modern equivalent of the remote above. It doesn't have enough buttons to act as a handheld macropad, but I just got it in the mail today and there's enough for the essentials while keeping everything ambidextrous (thus I've got 2 duplicate buttons)

Looking in the manual, they expect about 4 months on a set of AAA batteries. 

The only drawbacks I've got so far is that I don't like that I have to lift my thumb to cover any reasonable distance and that there's friction, meaning that when I'm trying to move the ball a slight amount, it won't move when the force is lower than the restoring force and will glide like ice if above. Standard physics, really.

They're both kind of one in the same problem, because a faster cursor speed would reduce the first issue and increase the second issue, and vice versa. Maybe my dying mouse (which is the reason why I bought it btw) doesn't have this issue because the minimum force to slide is higher? Perhaps it's the higher mass resulting in higher acceleration times after the force exerted exceeds the frictional restoring force? Perhaps it's a skill issue because this is the first trackball I've been able to use?

Oh, and the scroll wheel has much stronger dedents than I'm used to. 

[Oct 15 Edit]

Unfortunately, the mouse isn't what I'd call programable, but instead software remapped. What I mean by this, is that I can't just program the trackball on my main PC and then have the same keybinds present on any other PC.

These are the keybinds that are on my Intel N100 laptop, which uses MS Edge as the default browser. 

I found that this binding resulted in less mental overhead when trying to zoom in/out of webpages and allowed me to easily pan in Fusion360 (using the Tinkercad orbit option). 

Those bottom buttons 11 and 12 were kind of hard to press often, so I put less used commands on them.

I'm likely sending it back because my hand often blocks an otherwise direct-line-of-sight signal to #Teti [gd0022], I keep wanting to hold the device at angles that means that the trackball isn't resting on the bearing balls (causing it to jiggle around) and my thumb is overworked having to take a flight from the trackball to the 5-button area.

[/Oct 07 and Oct 15 Edits]

Another device from the past is the iGrip. Ideally, I'd like to be typing 50+ wpm on Tetent in about 2 hours, not 2 months as claimed with the iGrip.

IFYOO GTP01, a handheld device that either comes in pink/white or black/yellow-orange. I'm very much more of a fan of the white keys, and I'd like to imagine a milky green edition. This one seems quite interesting because of the trackpad design and the rear buttons, as well as the assymetrical grip

Cube Keyboard

I first found this when I first designed #Tetent TestCut [gd0139], but the Super Cuber X7 is a device that has many of the things that I'm trying to put into Tetent, such as a soundspeaker, alarms and a cursor moving input. I don't have plans for FM radio or a flashlight though.

AR / VR input devices?

The good news about AR and VR is that the keyboard and mouse is unsuitable as an input method, thus research into new input devices is underway. I thought this would be a good place to see if there has been any developments, but I found this 2019 article instead. 

The author first mentions some options that they know don't work, then says "What is left? Thinking out of the box." I actually had hope, thinking that the author found something when he was explaining why all the aformentioned input strategies fail. He hasn't, but has provided a list of things that such an input device needs:

• Input while standing, without the need of hand-rest in front
• Natural 3D pointer input (think mouse for three dimensions)
• Full 104+ key size of input variety without jumping through hoops
• Good haptic feedback on keys for blind touch typing

Conveniently enough, Tetent gets full marks. My initial requirement considerations for use with the Pimax 12K (before I found out that 35PPD is very low resolution) and later the #T^2 TyMist [gd0138] seemed to have paid off here. I've now changed the project description to "An input device for quickly typing text / equations, gaming, drawing and cursor / 6DoF movement."