My background §
Even before I started pressing buttons for a living I was always a pretty heavy computer user. When I was about 18 I started experiencing symptoms of RSI. At around the same time I learnt of the Dvorak keyboard layout, and so I switched, hoping it would solve my problems. It did, at least at the time, and so for the last 17 years I’ve been a Dvorak user.
At the same time I’m really interested in human/machine interaction and ways to get the contents of the human brain into a machine as fluently as possible. As a combined result of these two factors, I worked my way through a few different physical keyboards, including the Microsoft Ergonomic series (which I struggled to make Windows 2000 recognize — go figure) and a Kinesis Advantage, but my keyboard of choice for many years now has been the Ergodox EZ. The Ergodox (and its ZSA comrade the Moonlander) is a solid keyboard, with a few features that I’ve come to realize are very important:
Split halves. Any keyboard with the halves joined together (that is narrower than shoulder-width, as is the case with every joined keyboard I’ve seen so far) causes the elbows to bend in and the wrists to bend out, which seems to be the primary trigger of wrist pain for me.
Tenting. Ditto for the elbows: last year I found myself struggling with pain and numbness in my arm that, on investigation, matched symptoms of early cubital tunnel syndrome. Tenting, rotating the halves of the keyboard so that the wrists sit in a slight angle, allows the elbows to rest in a more natural slightly-bent position, and since adopting it I haven’t had any further problems with my elbow.
Thumb clusters. The first symptom I got was what’s affectionately known as ‘emacs pinky’. In fact, on most keyboards (though sadly no longer on ThinkPads) the control key, positioned in the bottom-right corner, can and should be held with the palm — this makes some combinations more awkward, though no more awkward than with the little finger, and saves the weakest digit. In fact, I think it would be better named for vi — the escape key is perhaps the least ergonomic key on a standard keyboard. Some people like to put it in the place of caps-lock, which is of course entirely missing the point. The reason I picked up the Kinesis Advantage is that on these keyboards the thumb, which is highly underutilized in standard layouts, gets a dedicated cluster of keys, perfect for things like modifiers, since the opposable thumbs don’t limit the motion of the rest of the hand at all, no matter how it’s engaged.
Programmability. My customization is light compared to some people’s, but my keyboard usage is pretty idiosyncratic, so it’s important to me to be able to customize my keyboard when I need to. Particularly for keyboards with fewer keys that need heavier layer usage, I want to be able to set up the layers as I want. My most elaborate customization is that I am (from time to time) a hobbyist stenographer, and rather than having a separate device I find it convenient to be able to configure my keyboard to speak one of the steno protocols understood by Plover such as TX Bolt or GeminiPR.
Finger-oriented keyboarding §
With this in mind, I was pretty excited when I stumbled across the CharaChorder. The promise is simple: increase typing speed through a combination of minimizing finger movement and stenography-style chording (albeit with a theory designed to have a shallower learning curve, basically just mashing all the letters of the word at once). I signed up to the pre-order, and when I got it I immediately started playing around, practising a few hours per day on the (very good!) training software.
If you’ve ever done typing training, you’ll know that most of the point of the drills and forms that they teach you is to keep your fingers on the home row so that you can reach all the letters with a minimum of movement. The CharaChorder turns this on its head: on the CharaChorder there is only the home row. Where you would stretch up or down or sideways to reach a key on a traditional keyboard, on the CharaChorder you simply wiggle a joystick in that direction, never taking your finger off the key. There is in fact a second row of joysticks used for arrow keys or mouse emulation, but we’ll ignore them for the purposes of this since they don’t impact most of the typing experience.
I loved the compact form factor combined with the power of chording, and I was excited to develop a new steno theory based on the CharaChorder’s unique structure. Unfortunately, I was thwarted: the CC1 runs the CharaChorder OS, a closed-source project with only a restricted set of chording configurations, and the fact that it must take place on-device severely limits what can be done compared to the software-based solutions I was used to. The CC1 is great for people who need to travel and connect their device to many different computers, but that’s not my use-case. Furthermore ergonomics are decidedly not a priority for the CharaChorder, with (for example) the down presses being a chord of each of the cardinal directions, for a hefty (42 × 4 = 168) grams of force to activate, and some default chords requiring punishing scissoring and twisting motions that left my hands very sore after a practice session. The final straw for me was that a month or so into using the CC1 one of my joysticks broke, coming loose in its housing; CC were kind enough to send me a replacement half, but shortly after that arrived the same thing happened to a joystick on the remaining half from the first keyboard! I hear that the build quality has improved significantly on the CC2; please don’t be put off from trying it by this experience of mine! It remains a really nifty device, especially if you regularly need to transport your keyboard and use it with computers whose software you don’t control, or if you want to attach the keyboard to your trousers.
Thankfully, on the CharaChorder Discord I was introduced to the Svalboard Lightly.
The Svalboard Lightly §
The Svalboard, a descendent of the beloved DataHand, is based on the same principles as the CharaChorder: each finger gets a key ‘cluster’ and rather than moving the finger off the key to reach other keys as in traditional keyboards, the finger stays in its cluster and merely moves in a direction to press a key — including down, like a standard keyboard but mostly unlike the CharaChorder, in which the down press is too awkward to use as part of standard typing.
It throws away the usual mechanical keyboard spring-loaded switches in favour of something at once simpler and more high-tech: an infrared source fires a beam at the stem of the key, initially blocked by a ‘flag’ on the keystem, and triggers a keypress when the key is moved out of the way to allow the beam to hit a detector. Magnets — actual, permanent bits of magnetic neodymium — are used to replace the tactility of traditional switch springs, with pressing the key requiring breaking the magnets' hold in a rather satisfying way that feels similar to a traditional buckling spring but requires much less total force to activate.
This is in line with its philosophy: apart from the circuitry and the magnets, everything is 3D-printed to open schematics and held together with standard screws, ensuring that any piece that’s likely to break can be replaced, even if the board stops being produced (as happened to the DataHand).
The excellent §
The Svalboard runs QMK, a de-facto standard for customizable keyboard firmware that I was already used to from the Ergodox. Something that was new to me, though, ZSA do now provide Oryx, configuration software with a similar purpose; I’d just never tried it! was that the variant of QMK it uses (by default) is a custom spin of Vial, a fork of QMK enriched with dynamic layout editing via a GUI editor, or (for the Svalboard) the excellent Keybard, which replaces the Vial frontend with a much lighter-weight Web-based alternative.
I’d never bothered to get set up with a graphical configurator before, reasoning that I’m perfectly capable of writing a bit of C so I might as well avail myself of the full power of the firmware, but the qualitative difference it made, especially as I was still getting used to the board and ‘dialled in’ to make my everyday tasks comfortable, was phenomenal. The ability to open a Web page and tweak a key has led to a huge number of quality-of-life improvements that I could have made without it but never did because I didn’t want to interrupt my workflow to flash my keyboard, and then I promptly forgot about them.
That ties neatly into another aspect of the Svalboard I really appreciate: every part is minutely customizable. At build time several options are available for different key sizes and shapes, and at the time of writing there are three different pointing devices available which can be swapped out on either side of the Svalboard. But even after assembly every part of the board that the user’s hand touches can be shifted, rotated, raised, lowered, or detached entirely by loosening some screws. For the first few months I kept a screwdriver on my desk so I could make small adjustments as I noticed certain movements were more or less comfortable for me, and the software allows me to easily tweak my layout to put more common actions on more easily-reached keys.
Some caveats §
While overall I think it’s a fantastic device, it would be remiss of me to pretend that everything has been plain sailing with the Svalboard. I’m very happy with it and it’s now my sole everyday keyboard. That said, there are a couple of things that, while far from deal-breakers for me, definitely bear thinking about, and if you have the chance to test-drive a Svalboard you should pay attention to.
First, and most definitely least, I’ve had a couple of minor build-quality incidents. Notably, the magnets that hold the keys in place have escaped from their plastic housing three or four times during my ownership of my Sval. Thankfully, in sharp contrast to the CharaChorder, the hackable nature of the Svalboard makes this pretty easy to fix: some superglue and a steady hand has them better than new in a couple of minutes. But it’s never nice to have to stop work to re-glue my keyboard. Of course the Svalboard is also available as a kit for 3D printing around, so potentially I should have built it myself and then I wouldn’t have anyone else to blame!
Then, there’s the matter of the price. A premium mechanical keyboard like the ZSA Moonlander will typically set you back up to around USD$400. But the Svalboard comes in at easily twice that and even higher depending on respective configurations. For me, so far, the device is well worth the investment, but your mileage may vary. If you have some spare time and access to a 3D printer, the kit option is significantly cheaper, and seems like it could be fun.
The Svalboard demands correct configuration, and can be very uncomfortable to use until arranged right. Being able to comfortably and consistently press the keys is, in my experience, very much a function of how well set up the keyboard is, and so if you have a setup where it or you will be moving around a lot, you’ll find yourself spending a lot of time repositioning and reconfiguring the keyboard. I strongly suggest looking into a chair-mounted solution, which lets the keyboard remain statically positioned with respect to your body and arms. Relatedly, its many moving parts make it infeasible to transport around easily; Morgan sells a blessed travel case, but it’s a chunky thing, and doesn’t lend itself to throwing in a backpack, nor (due to its positioning sensitivity) would I really want to use a Svalboard on the go on a train or aeroplane. Likewise, the optical keypress sensors require active infrared beams, meaning they’re too power-hungry for cable-free usage.
It’s punishing to sloppy typing too. For those of us used to hammering away at slab keyboards all day, the lightness of the keys can make it feel like the keyboard is on a hair trigger. I often find myself making typos that wouldn’t have happened on a traditional board, because between having the impulse to press the key, moving my finger over to where the key is, and exerting the requisite force to register a keystroke, I have plenty of time to reconsider my actions. With the Svalboard, it can feel like I type the letter before I decide to, and then I’m left to clean up the resulting mess! I think this is significantly a matter of practice, and I’ve definitely seen it get better as I become more accustomed to the board, but it’s something I still fight with six months in.
Finally, in balance to its many advantages, there is one straight-up disadvantage to the Svalboard (and other DataHand-like layouts) compared to a slab keyboard when it comes to typing, and that’s repeat keypresses. Whereas with a traditional keyboard a repeat keypress involves making a potentially awkward movement to the desired key and then making several simple pressing motions, for the Svalboard repeating a key involves repeating the stretching/curling motion for every press, which can be much more difficult than just repeating the press. This is most evident in games that need you to hammer a particular key several times in quick succession, but can also be seen when typing words with certain double letters. There are workarounds available for this, though:
-
First, not all keys suffer from this problem: at least the centre keys of each cluster are at least as easy to hammer as on a traditional keyboard, if not even easier due to their super-light switches. I also find the inward and downward presses easy enough to repeat. If you can keep keycodes that need repeating on those keys that are comfortable to repeat, you’ll have no problem.
-
Alternatively, as a more general solution, QMK supports binding a ‘repeat’ key that can be used to repeat the previous keypress. If you bind that key to an easily-repeatable key you can use it to comfortably repeat any key on the keyboard. I’ve tried this a couple of times, but the combination of sacrificing a prime key with the mental overhead make it not worth it for me, compared to rearranging the keys that need to be repeated.
My Svalboard setup §
The above is my take on the Svalboard as a device; you can find many other reviews online, some of them in much greater depth, but these are the things I’ve noticed about it in my first half-year. The remainder of this article is about how I have my Svalboard set up, and how I use it day-to-day.
Mounting §
I have my Svalboard on a chair mount, simply attached to my chair arms with some SmallRig Magic Arms. I experimented with on-desk and under-desk mounting, but I found I tend to move my chair too much with respect to the keyboard, which puts me into poor typing position and messes up my typing for a bit before I figure out what the problem is and adjust back. By contrast, having the Svalboard mounted to a chair means my arms, so long as I keep them on the arm rests, are always positioned the same way on the Svalboard, and the only thing I have to worry about is my cats using the cables as chew-toys. Magnetic cable attachments mean that if a cat decides to dangle from your cables by the teeth the cable detaches, resulting in minor inconvenience and maybe a damaged cable rather than a thousand-dollar keyboard being forcibly ripped from its housing and dropped onto the floor. I settled on these units because they swivel vertically, but they’re all pretty interchangeable.
Base layout and learning to type again §
My Svalboard layout is available on GitHub in Keybard format, along with some documentation that is sometimes in-date.
When I learnt Dvorak I learnt to touch-type properly for the first time, and for whatever reason I’ve managed to almost fully maintain my QWERTY hunt-and-peck skills, with only a brief adjustment peried when faced with a QWERTY keyboard before I can begin typing at a reasonable speed. So, I reasoned, given that the Svalboard is physically quite different from the flat keyboards on which I normally type Dvorak, if I pick a base layout that’s sufficiently different to Dvorak anyway it and my Dvorak muscle memory won’t interfere with each other and I’ll keep both intact. And, since I’m learning a drastically different physical keyboard layout anyway, it won’t slow me down much. So I picked Hands Down Promethium, which was recommended to me for split keyboards, adapted to the Svalboard by Ira Cooper.
This was a mistake.
In fact, HDPm and Dvorak interfere with each other constantly, and I find myself making typos from one whenever I’m using the other. Six months later I’m up to about 80 WPM on the Svalboard now from my ~120 on Dvorak, but given the amount of interference from my Dvorak muscle memory I can only assume I’d be at full speed by now if I’d used an adapted Dvorak, and my flat-keyboard Dvorak would not have suffered as it now does. Particularly, since the home rows are roughly mirror images of each other, I frequently press with the right finger on the wrong hand. Perhaps it would have been easier if I’d mirrored the layout, but it’s too late now.
Nevertheless, I’ve found picking up a new keyboard layout as an adult to be a pretty educational endeavour. There’s a symbiosis with the Svalboard that rewards mindfulness: I will do a couple of rounds of monkeytype, realize that I’m consistently making a particular error because my keyboard is misconfigured in some way, get out the screwdriver and fiddle with it, then continue. But doing that requires a certain level of conscious introspection to determine what typos are due to keyboard misconfigurations vs brain slips vs physical mishaps (which occasionally happen randomly!). It’s an almost meditative state.
The main adaptation from HDPm is a vertical mirroring. This is because, at least for me, the Svalboard’s north keys are much harder to hit than its south keys, even though on slab keyboards conventional wisdom considers the top row to be lower-cost than the bottom row. Something I didn’t realize at the time is that I also find the inward lateral keys much easier to hit than the north keys, so I’d have liked to try mapping the bottom row of HDPm to the inward laterals of the Svalboard rather than the top row. Nevertheless, I find this layout to be pretty good overall, and I don’t think it would be worth the switch now I’m starting to get used to it.
On the fingers, after a long adjustment period, I’ve found that the easiest keys to hit are, in order:
-
the centre keys,
-
followed by the south keys,
-
then the inward keys,
-
then the north keys.
The outward keys are relatively uncomfortable, especially on the weaker fingers, and I reserve them for infrequently-used keys. The astute reader might have noticed a lack of an escape key: I actually type escape by pressing ⌦ Delete and ⌫ Backspace in unison. This took me a bit to get used to but now I really like it, and would consider a similar approach for other infrequently-used keys like ↵ Enter. Combos with the outward keys can use a slight roll of the wrist, and so are much more comfortable for these kinds of one-off actions than for typing letters.
On the thumbs, the easiest keys for me to hit are the inward ‘pad’ keys and the bottom ‘knuckle’ key, followed by the downward key. The lateral motions required to hit the ‘nail’ and ‘up’ keys are not difficult to do, but cause strain that makes my thumbs hurt at the end of the day if I assign them to anything that’s used too heavily. This is probably unique to my pathophysiology: you should experiment yourself.
Higher layers §
The Svalboard has only 60 keys, meaning that layering is inevitable to get a full complement of keystrokes. I have three ‘everyday’ layers, not counting shift: my base layer, a numbers/symbols layer, and a combined navigation/control layer. These are all ‘momentary’ layers, meaning that I switch between them by holding down a key, and I switch between them all on the fly while editing text. Layering is something that seems to put people off of keyboards with fewer keys, but while experimenting with the Ergodox (on which, for many keycodes, I had both a less-accessible key in the base layout and also a more-accessible key on a higher layer) I found it much more comfortable to have an easily-reachable layer switch than to have a dedicated key that requires a stretch and a mispositioning of the hands. I find three layers few enough to not cause me undue mental load, while still providing plenty of key space (none of my layers, even the base layer, is fully utilized at the time of writing).
I’ve settled on a hand-split setup in which software modifiers like ⇧ Shift, ⎈ Ctrl, and ⎇ Alt (very important since I’m an emacs user!) live on the left thumb, while layer modifiers live on the right hand. Having modifiers on the thumb allows me to have only one copy of each modifier, since the thumb can move independently of any other finger, while having the layer-switch keys (which are not detectable by software) on the right thumb allows me to bind the entirety of the left half of the keyboard when gaming. The ❖ Super key is on the right hand for the same reason: it isn’t usually bindable in games. The design of the individual clusters is subtle: on the Svalboard thumb cluster the pairings on the same side of the thumb cluster (e.g. ⎈ Ctrl+⎇ Alt on my layout) can be easily hit together, while those on opposite sides (e.g. ⎈ Ctrl+R can’t be). ⇧ Shift, therefore, lives in the middle of the cluster, where it can be easily hit with either side, while for the specific combinations of the modifiers on the right with R or ↵ Return I have dedicated combinations with other layers: for example, R on the control layer produces ⎈ Ctrl+R. On the right cluster I have my layout modifiers, and here I make use of a controversial feature of the Svalboard, inherited from the original DataHand: the middle key of the thumb cluster has a ‘deep press’ that is activated by pressing on it harder than usual. Obviously this can only occur while the usual key is held down. This is ideal for me, as I control my window manager using a combination of ❖ Super with various control and function keys, so while the middle press usually enables my control/navigation layer, by pressing it more firmly I can activate ❖ Super as well and control my window manager. Unfortunately, I sometimes need to use ❖ Super with keys from the base layout, so I also need a bare ❖ Super key that doesn’t involve switching to the control layer.
The symbol layer has symmetric pairs of brackets as well as the standard numbers and symbols. Worthy of note is that I have a ‘programmer Dvorak’–like layout in which symbols are prioritized over numbers: as a programmer I type symbols a lot more often than numbers, and privileging small numbers over large numbers encodes Benford’s law. It’s notable that, due to the abundance of key space given by having an entire layer dedicated to numbers and symbols rather than just the top row and some peripheral keys, I’m able to represent the entire set of standard keyboard symbols and numbers here without shifting. Insofar as needing an extra layer is a negative for the accessibility of these keys, I find that this advantage more than offsets it. I also keep some infrequently-used keys here, such as my ⎄ Compose key, which I use only for symbols too rare to have their own dedicated key. I use composition for occasional letters with diacritics, emoji, mathematical characters, or non-ASCII punctuation like quotation marks or dashes. It is important for clarity for quotation marks to be properly aligned: the typewriter ‘straight quotes’ arose only out of lack of glyph space, and should be avoided where more capable encodings are available. I do find the ‘straight apostrophe’ useful for semantically distinguishing the apostrophe, whose chirality is unimportant, from the single quote, against Unicode recommendations.
The control layer has, most importantly, two ‘WordStar diamonds’ for textual navigation: one on the left produces standard arrow keys, while one on the right produces ‘big-step’ versions of the same (⇞ Page Up, ⇟ Page Down, ⇤ Home, and ⇥ End). These diamonds differ from the standard QWERTY WASD in that on the Svalboard (at least for me) the ‘south’ direction is much more comfortable than the ‘north’ direction, but both are trumped by the ‘centre’ direction, so both ← Left and → Right are on the centre keys along with ↑ Up, and instead of ↑ Up being on a row above ↓ Down is on a ‘row’ below, with a southward keypress. Also on this layer I have function keys, matched in position to the number keys on the symbol layer, which I use with ❖ Super to switch workspaces, and media control keys, as well as the key to switch to my steno layer.
Stenography §
I haven’t focused too much on stenography yet, since I’m still getting
up to speed with the base layout, but with help from
Ira Cooper and
Greg Millam I did add support to the
Svalboard and to Keybard for generating and configuring steno keycodes
using QMK’s steno
support, so the Svalboard works out of the box with Plover — connect
it to the Svalboard’s emulated /dev/ttyACM device, set up
stenography keys on your Svalboard, and you should be good to go.
A big reason I bought the Svalboard in the first place is because the super-light tactile keys with low travel, combined with its full N-key rollover, should be excellent for steno, and so far my experiments have borne this out. There are two main caveats I’ve found that I’ve had to somewhat mitigate in my steno layout. Neither seems insurmountable, but I need to experiment with different options to figure out how to make something that’s sufficiently expressive but still comfortable.
-
The Svalboard has only one ‘column’ per finger, while the stenotype traditionally has an extra column on the outside for the ＊ Star modifier on the left, which is easily emulated with the outwards lateral, and the –D –Z terminals on the right, which are not. Fundamentally, the steno layout allows the little finger to press up to four keys at once while the Svalboard only allows combos of up to three. Thankfully, some combinations of the four (like –T+–D+–S or –T+–Z) are inadmissible, so if we make sure that we can type each of the adjacent two-key combinations and then the four-key combination we should be okay.
-
The centre/south combination, which seemed like the most obvious mapping of the stenotype layout onto the Svalboard, is pretty difficult to do consistently due to the angles of the keys involved. I have two workarounds for this, and I’m not yet sure which I prefer: either using the centre key as a steno combo key, e.g. PW–, or eschewing the centre keys entirely for the inwards lateral key as the upper row of the stenotype. Each has its own advantages and disadvantages: using the laterals requires tilting either the keyboard or the wrist to make the neutral position (with fingers on both ‘rows’) natural, while using combination keys exacerbates the –D/–Z problem: the rightmost cluster would need 6 keys to represent it uniformly with combinations, but it has only five.
Either way, it’s clear to me that I at least need combo keys for the vowels: the only two thumb-cluster keys that can be combined comfortably enough for steno are the inward keys, and constantly performing that lateral motion makes my thumbs hurt pretty quickly. I’ve chosen the thumb knuckle and pad keys for the steno inner and outer vowels, as the easiest thumb keys for me to hit, with the centre key for their combination.
Pointing §
The Svalboard comes (optionally) with real pointing devices on each hand, and the default firmware supports a ‘mouse layer’ on the highest layer (15) that is automatically switched to when a pointer event is detected, and switched away from when a press of a key not on the mouse layer is detected. The default configuration has the mouse layer time out after inactivity, as well; I found this very unpredictable, and it led to me making a lot of mode errors. It can be configured or disabled with a special keycode that can be bound in Vial or Keybard. I use two trackballs, one for pointing and one for scrolling, but I’m looking forward to switching at least one of them to a trackpad as soon as the trackpad gets smooth-scrolling support, since the ability to use trackpad gestures like tap-for-click frees up a bunch of key space (and maybe removes the need for mouse mode at all?). When gaming, to avoid accidental mouse-mode switches, I usually remove the left trackball altogether.
Retrospective §
Overall, I find the Svalboard ticks all my boxes for a keyboard, and if you can afford the money and the time investment I’d definitely recommend giving it a go. Typing is a lot more comfortable for me now, and there are no downsides big enough to make me want to switch back.
It took me six months to achieve 80 WPM, but I think it would be a lot faster if adapting an existing layout that you’re familiar with. I don’t recommend switching keyboard layouts at the same time, unless you wanted to switch layout anyway. Even though it can be fun, it will likely only hinder your adoption of the keyboard, and doesn’t come with any particular benefits. Benefits, that is, over switching layouts before or after switching keyboards — the ergonomic advantages of modern layouts, especially if you’re a QWERTY user, are well-documented, and I won’t repeat them here, though they are probably somewhat diminished for a keyboard that reduces all finger motion anyway. I’ve noticed a logarithmic pattern to my speed improvements: I hit 20 WPM within the first week, and 40–50 WPM in the first month, which is a bit painful but good enough to get things done with. By the second or third month I no longer felt like my keyboard was imposing a large additional mental load on me and instead I felt like it was just a bottleneck through which I have to squeeze my thoughts — which has been my experience with all keyboards, even when I was typing at 120 WPM.
Though I’m still quite slow, and the Svalboard doesn’t try to sell itself on improving typing speed, I see steady improvement in my typing speeds as I use it more, and I don’t see any reason I shouldn’t be able to match or exceed my typing speeds on slab keyboards while enjoying significantly better typing comfort. My hands feel much less tired at the end of a day, though it’s important to note that this is only the case after I’d spent quite some time adjusting the keyboard to my typing habits and my typing habits to the keyboard. Particularly, figuring out my typing posture and keyboard mounting setup made a world of difference.
In addition to the practical aspects, and despite the earlier warning, it’s been pretty fun getting back into the world of crazy keyboard layout optimizations, and I’ve discovered a lot about how I learn physical skills and learnt to pay attention to aspects of how I use computers that I would never have considered.
Finally, this post wouldn’t be complete without a shout-out to the Svalboard Discord community, who have helped me a lot on my journey both with hardware and software.
Happy typing!