In previous article, I focused exclusively on one side — the front face. I found an interesting design, and the next step was to collect a list of essential functions and create a 3D model. However, at that stage, there were still some uncertainties. For example, how fast you could adapt to the layout, does the typing feel comfortable enough for long-time usage, and object ergonomics in general? My buddy Boris helped me to clarify some of them by building and configuring a working prototype of the keyboard.

After a few weeks of using it, my fingers eventually began to adapt muscle memory. However, I quickly realized that I had focused on compactness too much, which led to some miscalculations of the key sizes. They turned out quite small, even for my relatively small hands, and they were placed too close together, causing me to often press two keys at once. After recalculating and conducting a few quick tests, I found what I believe to be the perfect balance. Simultaneously, I read about the standardization of different input tools and found some alignments with my measures. I want to highlight a particular book, that became a real treasure and valuable source of information, called Human Engineering Guide for Equipment Designers by Wesley E. Woodson and Donald W. Conover. It’s a fascinating collection of guidelines and advice for designing various objects, including spaceship dashboards, keyboards, etc.

The triangular keys for the arrows, despite my friend’s skepticism, turned out to be fairly comfortable. To clarify, I wasn’t using all of them all the time, so their shape remains an open question. As older keyboards demonstrate, it’s been an open question for quite a while.

I think the best solution might be to forget about the triangular shape and simply allow the user in some way to adjust them as they prefer within these four spots. But I’ll come back to this later because I don’t consider it a big issue. Of course, objectively, this is nоt the most comfortable one, but let’s admit — it looks cool.

With the new key sizes, everything seems to be in the right place again. Boris’s prototype used a solid PCB with mechanical switches, which I found amazing. All that remained was to add various ports, an SD card reader, a battery, a cup holder…, and it’s done. But I couldn’t get rid of my desire for compactness, so I had to make sacrifices and find workarounds, relying on what was available on the market. I wish I could fit everything in, but the puzzle just wouldn’t come together until one day. After several attempts, I decided to “rip off the band-aid” and, for example, replace USB-A with a smaller USB-C, change the SD card reader to a tiny Micro-SD .

Now, I’m on the doorstep of building the first model. There’s a mountain of technical details ahead that I’m afraid to even think about. But I’ll try to dive into them and share the results in a future article.

I hope you still find this project interesting to follow. Special thanks to Sasha Shestakovych and his team for these fantastic renders. Also, a big thanks to everyone who’s asking questions, advising, and supporting me.

How I designed a keyboard was originally published in Bootcamp on Medium, where people are continuing the conversation by highlighting and responding to this story.

Previously, I was searching for a better character placement and keyboard layout. At that point, I tried to save similarities with popular keyboards.

I’m happy with the result and didn’t want to stop on that. My new experiments show some pretty interesting ideas. So here’s the second part of my experiments with the keyboard layout.

I don’t want to limit myself, except only try a more symmetrical look. First, let’s see what happens if all of the keys will be the same size.

Here’s the first issue. The bottom row is the simplest one to fix because there’s the space key. We’re using it a lot, so it should stay pretty large and fill that gap on the right. I didn’t bother with enter and shift keys. I’m not sure how often basic computer users like myself are using this shift key, and do they at all, so I decide to delete it and make the enter key bigger, like a cliche alarm button.

For sure it’s not the best way of filling the area, but let’s continue for now. Next dilemma I faced was whether to leave the arrows on three keys or split them up into four. Big brother helped me with this one. Around that time I accidentally came across a pill box from the middle of the last century, designed by Fred Troller.

Visually it’s a perfect match. This design inspired me to use triangular elements as the shape of arrow keys. I’ll use it for a new look, and for now, I placed them as a group of four instead of that big enter key. This combination has the most sense among others. As a result, the enter key became a long, horizontal space-alike key.

Now it’s time to use triangular arrow keys and clean up all proportions, colors, typography, small details like tactile hints, etc. Of course, everything was designed with taking into account my native Ukrainian language. Here’s my final design.

And as a dessert let’s talk about a virtual copy of this keyboard. Virtual keyboards usually work as an additional input source. Keep in mind that the basic symbols are kind of coordinates for others. It helps you to navigate between symbols with shift and option keys. In my opinion, a virtual copy should always highlight these coordinates, so you could keep working with the keyboard, without switching to the mouse every time you need some specific symbol like “∆” for example. At this point, I was searching for a good placement of these highlights and found out that using the bottom left corners is the best way because of the location of the shift and option keys down there relative to the rest of the letters.

Imagine a virtual assistant, opened on top of all windows, which allows you to keep the focus on the text and make it easier to find the wanted character.

For me, it’s a very unusual concept. To be honest, I have no idea how it’s gonna go, but I want to give it a shot. Of course, we can argue about its usefulness, yet in vain since only long-term testing of a physical prototype will give an objective answer. It takes some time to build and test it out, and I’m looking forward to doing it. For sure I’ll publish all the details after that. So feel free to follow my profile if you’re interested. Thank you.

Imagine a keyboard with intuitive character placement. Searching for some specific typographic symbols, ligatures, or formula element is not a big deal. You don’t need to guess where this or that punctuation mark has moved after changing the language. I’m using the bilingual keyboard and dealing with all these pains almost every day. That’s the first part of the story of how I found better character placement and basically redesigned the entire keyboard.

Let’s start with the layout. There’s a detail that I was sure about until I began the research. You can notice a small offset of the Q, A, Z rows on your keyboard by 1/2, 1/4, and 1/2 of the key size respectively relative to the previous row. This detail can be seen in all standards (ISO, ANSI, ABNT, JIS) and is usually explained by the convenience of typing as if it’s easier for a person to reach different symbols. This surprised me because people’s hands are so different in proportions and sizes. How can it be suitable for everyone? Maybe it was meant you could easily get used to it? I haven’t found any complete research on this topic. Please share a link if you come across something worth paying attention to.

It make sense that a modern keyboard is built on the basis of typewriters that were popular previous two centuries. Most of them are built according to the same principle, so all of them have the same row offset. I suspect it’s because each key pedal is stacked not in the usual order —5—T—G—B—6—Y—H—N, but —5—V—T—G—6—B—Y—H instead. I can’t see any comfort here besides simple and elegant construction.

This mechanical offset is understandable. Same offset on the early computer keyboards as well, because of a user’s muscle memory. Modern models don’t face the same problems. You can easily find an example with symmetric offset, mirrored offset, horizontal offset, or even without it. Are they uncomfortable to use? No, that’s just an old vestige. I think you can safely remove it and nothing bad happened. After experimenting, I found the 1/3 of the key size fits best. Thus, the visually new keyboard will not differ much and will keep its compactness. But there remains a pretty large shift key from the right and a free key-size space at the top corner. Although it’s not necessary to focus on these issues at this moment because the solutions will occur later.

Now, it’s time for the character placement. I’ve divided all of them into three groups: primary (opens with a key), secondary (opens with shift+key), and optional (opens with option/alternatice+key or option/alternatice+shift+key). Here is a diagram for better understanding.

I won’t touch most of the primary ones. Among the secondary ones, punctuation marks were moved the most, but one more key had to be added to achieve a universal solution for the two languages. It perfectly matched the shift issue, and thus solves it. Moreover, you can find a similar solution in the ABNT and JIS standards. Optional characters placement was guided by the following principles:

1. Select a separate row for currency, for example, the upper left corner had the most space;
2. The first letter of the name of the Greek alphabet letter matches the base letter, for example “∆” (delta) is placed on “D”;
3. The first letter of the character name matches the base letter, for example, “❧” (‎hedera) is placed on “H”;
4. The higher rank of some character, for example “«” (\+option+shift) is a higher rank of “‹” (\+option);
5. An option/alternative of some character, for example “‰” (5+option+shift) is an option of “%” (5+shift);
6. The character is visually similar to the letter, for example “√” (‎square root) is placed on “V”;
7. Fitts a free space, such as “×” and “÷”;
8. Letter stylization, such as the “¸” element on the bottom of another letter (Ç), but more on that a little later.

I’ve decided to delete the entire row of functional keys. They unite with primary numbers, with the escape key at the beginning of the row, because the original characters have been moved. In turn, the function key became another kind of navigation, along with shift and option. At this point, I was constantly switching between my languages to don’t miss anything, and after several dozen cycles of changes, I ended up with this placement.

Looks like nothing special, but I’m happy with the result. It would be great to make a real prototype and test it out, but I don’t have that opportunity yet. In the process, I realized that a well-configured automatic text replacement will allow you to quickly stylize letters, and create ligatures or various characters. For example, replace “AE” with “Æ”, “O/” with “Ø”, or “ffl” with “ffl”, and so on. In fact, you can already do this in your keyboard settings, at least in Mac OS.

Anyway, my experiments with layout continue, and I will publish all new findings in the second part. In the meantime, I will be glad to hear your thoughts on this topic. Thank you.

Your keyboard can be much better was originally published in Bootcamp on Medium, where people are continuing the conversation by highlighting and responding to this story.

In stores, you can often encounter a seemingly minor problem — the checkout line. This nuisance can make you uncomfortable or even affect your decision. Time is a valuable resource. That’s why, when there is a risk of wasting it or not getting to your appointment on time, a person may cut back on purchases. In this case, both the customer and the seller lose. The first gets into an unsettling situation. The second risks losing profit, or a potential customer.

In my opinion, the process of registration and payment of goods at the checkout has long been out of date. Human aspect is crucial here. Let’s look at the process diagram.

Cashiers are responsible for most of the steps. The process can be greatly accelerated by replacing a person with a computer.

We’ll face only two key challenges: quick item registration and leave the customers behind. The items should be registered at different distances, in any position, taking obstacles into account. Considering this, the best option would be using an RFID tag.

RFID is a contactless data transmission technology due to energy condensation on an antenna attached to the chip. This smart label is fairly small and carries up to 2KB of information. It has no power source, so its activation requires a reader antenna and a computer to analyse the data. Depending on the RFID antenna power, the tag can be easily recognized at up to several meters’ distance.

The store can be turned into a network with a smart database engine. The readers’ antennas will play the role of receivers, and the RFID tags — of the signals to be recognized. In this case, the signal is a unique identification code assigned to a specific item in the database. So, a tag affixed to a product makes it visible and familiar to the computer. A low power antenna, attached to each shelf, should record the items availability.

Each antenna can be added to the database as a separate data group with a list of items assigned to it. So, the computer will understand the position of items. Such a system can be expanded using antennas of different powers.

An item with tags should be registered at three points:

• On the shelf, to understand the current placement of goods in the store;
• On the exit line, to create an order and cost estimating;
• At the cash desk, for additional identification of the items, order confirmation and cost estimating confirmation.

Depending on the interior and the distances between the shelves, a low-power antenna is sufficient at the first point to read the tags within a radius of 20 cm. At the other two points, one or a pair of parallel, high-power antennas is required to read the tags within a radius of 150 cm. These antennas can be decorated or completely hidden. They will be protected and invisible to the customers.

All necessary devices are available in various online stores. The number of labels required is equal to the number of items in the store. Low power antennas should be equalized with the total length of shelves, in meters. The number of high-power antennas for cash desks may be less than the number of current checkout lanes. In order for the system to work properly, it surely requires software and a database. It is also necessary to consider the possibility of re-planning of shopping stalls and checkout lanes.

Now we have freedom of imagination. For example, you can make various applications that, having access to information about the items slotting in the store, could be able to build the optimal path between the rows of shopping stalls.

In any case, the lack of queues will positively affect the customers and make for the store profit.

Feel free to comment, if you want to collaborate on such project. Also, thanks Marina Veremeienko for these awesome illustrations!

As you know, the sound is a wave, vibration of the air. So-called analog signal. It is stored on your device in a digital format, as information with some size. The audio channel is the flow of such information from the source (microphone, application, web page) through the processor, to the target (speaker). Each stream reserves a certain amount of memory of the processor. At the endpoint, the speaker turns this information back to vibration of air.

As the element of an audio player interface, volume controls visualize the level of vibration at the target point (speaker). For this, we use mostly Mute, Zero, Quiet, Loud states.

It’s important to understand that Mute, Zero, and Quiet — are completely different states of the audio channel. Mute is the indicator of the absence of any sounds. It means there is no audio channel at all. Zero and Quiet — indicates the audio channel is present here, and the volume at the target point is 0% and ≥1% respectively. All these states are often mixed or used incorrectly.

Here will be appropriate to notice that there’s no need of using the Mute at all. This shortcut is not able to turn off the audio channel at the source point. So by pressing it the reserved memory on the processor would not be released. We got Stop or Pause for this.

Now let’s take a look to one of the widespread volume control layouts. It consists of Zero, Quiet and Loud volume states with a progress bar placed between the last two. The progress bar percentage gradation goes from 0 to 100%.

This logic is not quite correct because the progress bar gradation must correspond to the neighboring icons. In this example, it must go from ≥1 to 100%. This is the first important fix. The second one is to use the percentage gradation of the whole layout. So if Zero state is 0% and Loud is 100%, the Quiet state will be somewhere between them (in this example ~17%). Therefore, the progress bar will correspond to graduation from ~17 to 100%.

Such layout works correctly. Now you can simply remove the Quiet state. There are two reasons to do so. Avoid the visual misunderstanding of the progress bar gradation and make whole interface clearer by using fewer objects. So we would see Zero and Loud states with the progress bar between them. Now the progress bar will correspond to the gradation from 0 to 100%.

In my opinion, this is better, optimized version. Here we have full volume control with the least number of objects. Of course, if the progress bar is absent (keyboard shortcuts), you need to leave three volume states Zero, Quiet and Loud. Last two states will correspond to the decrease and increase of the volume respectively.

And little advise additionally. Please use icons to indicate the directions of volume states. Also, make these icons active in your interface. And please don’t use modality for the progress bar because it’s only one object.

Thank you. Feel free to share with me your thoughts and ideas about this.