<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Sergey Kyune on Medium]]></title>
        <description><![CDATA[Stories by Sergey Kyune on Medium]]></description>
        <link>https://medium.com/@sergekyune?source=rss-1fe7ac8fe1c5------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*x9wvO1G0NgCvca_e.</url>
            <title>Stories by Sergey Kyune on Medium</title>
            <link>https://medium.com/@sergekyune?source=rss-1fe7ac8fe1c5------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 19:37:40 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@sergekyune/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[APIs design tips]]></title>
            <link>https://medium.com/@sergekyune/apis-design-tips-547e99d0e095?source=rss-1fe7ac8fe1c5------2</link>
            <guid isPermaLink="false">https://medium.com/p/547e99d0e095</guid>
            <category><![CDATA[services]]></category>
            <category><![CDATA[software-architecture]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[api]]></category>
            <dc:creator><![CDATA[Sergey Kyune]]></dc:creator>
            <pubDate>Fri, 29 Jul 2022 11:56:08 GMT</pubDate>
            <atom:updated>2022-07-29T11:56:08.083Z</atom:updated>
            <content:encoded><![CDATA[<h3>Designing APIs</h3><p>Author note: <em>I wrote this article in 2020, but because of some turbulent life events didn’t publish it back then and completely forgot about it until I accidentally came across this draft in July 2022 and thought I’d make it public.</em></p><h4>Who is this article for?</h4><p>This text is mostly targeting people who are new to the IT industry and to API design in particular: students, recent university graduates, interns, young specialists, and people shifting career paths. In the text, I will be explaining some basic API design concepts, so if you have already worked in the industry for a while then most of the content should not be new to you.</p><p>For more experienced readers I suggest to rather use this article as a <strong><em>checklist </em></strong>of topics that you should not forget to think about while kick-starting a new service. I like mind maps and prefer graphs to flat structures. To prepare this article I used the knowledge graph that I initially created on <strong>kumu.io</strong>. You can check it out here:</p><p><a href="https://sergey-pic.kumu.io/how-to-design-apis">How to design APIs</a></p><p>^ Walk through the nodes of the graph and see if you covered all the topics in your design.</p><h3>What we will cover</h3><ul><li>Business purpose &amp; endpoints granularity</li><li>Syntax &amp; request/responses structures</li><li>Idempotency &amp; duplicates handling</li><li>Async communication</li><li>Pagination</li><li>SLIs/SLOs</li><li>Versioning</li></ul><h3>Tip: define the purpose of your service</h3><p>Start by defining the purpose of your service. Answer the questions:</p><ul><li>Who are your API customers?</li><li>Which exact problem of your customers will be solved by the API?</li><li>Which features will the API provide? How are these features helping to solve your customer problems?</li><li>How do your features translate into your API endpoints and domain model behind the endpoints?</li></ul><h4>The granularity of endpoints &amp; entities</h4><p>Make sure that each endpoint has a clear purpose. Try not to create multifunctional endpoints. Complex endpoints will likely create problems in the future when you will want to introduce changes.</p><p>Keep your endpoints as simple as possible. That will allow you to introduce updates for one endpoint without touching the other ones.</p><h3>Tip: think about the syntax</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*PA9kfBWe9I488XCl" /><figcaption>Photo by <a href="https://unsplash.com/@mbaumi?utm_source=medium&amp;utm_medium=referral">Mika Baumeister</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Identify which language and standard to use.</p><p>Probably you will want to select JSON over XML. It’s less verbose, is easier to read, and can be used in JavaScript ‘as is.</p><p>You might want to stick to XML if some of your clients have legacy restrictions or if it’s easier to represent the domain model of your service using tags (e.g. your service works with some sort of mark-up).</p><h4>REST</h4><p>(<a href="https://www.codecademy.com/articles/what-is-rest">Definition</a>)</p><p>Consider making your API comply with REST standards. A lot of people confuse REST API with any API that uses JSON. It’s a good idea to make yourself familiar with <a href="https://martinfowler.com/articles/richardsonMaturityModel.html">Richardson Maturity Model</a>.</p><p>REST might suit you well if a lot of interactions with your API can be represented as CRUD operations.</p><h4>GraphQL</h4><p>GraphQL can offer an interesting alternative to REST.</p><p>In GraphQL the client defines the response structure. It helps to address the common REST problems of over-fetching and under-fetching of the data.</p><p>GraphQL API scheme is strictly typed.</p><p>GraphQL adds some extra complexity to the server implementation, so it might be not the best choice for a small simple API. It might be, though, a good choice for those who want to build an orchestration API, that hides underneath a layer of various services. Want to learn more about it? Check <a href="https://www.braintreepayments.com/blog/graphql-at-braintree/">how Braintree uses GraphQL</a>.</p><h4>Types &amp; formats</h4><p>You might want to make your API contract <strong>more</strong> or <strong>less</strong> strict in terms of field formatting. Examples of your choices might be:</p><ul><li><em>String</em> <strong>vs</strong> <em>Enum</em></li><li><em>Float</em> <strong>vs</strong> <em>Integer</em></li><li><em>Signed number</em> <strong>vs</strong> <em>Unsigned number</em></li><li><em>String of undefined length</em> <strong>vs</strong> <em>String with limited length</em></li></ul><p>And so on.</p><p>The main advantage of choosing any stricter formatting is that your clients will avoid a lot of potential errors while using your API. From the negatives — if you will want to change some strictly formatted field in the future, then the update will likely require introducing a breaking change in your endpoint contract.</p><p>Taking into account the things said above, it’s a good idea to make stricter choices for the fields that you are more sure about. That applies to the fields that are already part of a known standard or you do know very well which values will be used in the field.</p><p>If you can’t predict very well which kind of values will be used in the argument, it will be wiser to keep your constraints relaxed, collect some data, and later on introduce the necessary changes based on the newly acquired knowledge.</p><h3>Tip: think about Idempotency</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*55AHs7jqrsz1P5G_" /><figcaption>Photo by <a href="https://unsplash.com/@noahdavis?utm_source=medium&amp;utm_medium=referral">Noah Näf</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Decide to which degree your service should be idempotent. (If it’s even possible for your service at all).</p><p>Think about the most common scenario: how will you identify and handle duplicate requests?</p><p>Implementing <em>real</em> idempotency — when the same request always produces the same result — might lead you to an expensive implementation. This implementation though can simplify the life of your clients. For example, it will make it easier to handle retries.</p><p><strong>Useful link: </strong><a href="https://multithreaded.stitchfix.com/blog/2017/06/26/patterns-of-soa-idempotency-key/">SOA Patterns: Idempotency Key</a></p><h4>Working with IDs</h4><p>Define what your identifiers should look like. A safer solution is to use UUIDs unless you need your identifiers to be ordered or you have to encode some extra information inside them.</p><p>In general, try to avoid adding encoded extra knowledge into your identifiers. Separation of concerns can save you from headaches in the future when you will need to change something about this extra information.</p><p>Having numeric ordered identifiers can increase the chances of data duplication between various clients, but it can become a benefit in case you expect people to pronounce them over the phone.</p><p>It might be fine to use numbers instead of UUIDs, just make sure you don’t expose in the API your internal system ids. (E.g. database primary keys).</p><p>And do not introduce new IDs for the sake of having them. In situations with a lot of interconnected services, it’s often better to have fewer IDs than more — to simplify request tracing and debugging.</p><p>If you can dictate the format of IDs to your client it might be OK to use only client-generated IDs and not to produce them inside your service at all.</p><h4>Handling duplicates</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*e4SnImHJmiem1ZFG" /><figcaption>Photo by <a href="https://unsplash.com/@frankiefoto?utm_source=medium&amp;utm_medium=referral">frank mckenna</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Well-defined IDs are crucial to handle duplicate requests.</p><p>For handling critical transactions the suggestion is to have a pair of unique IDs per transaction generated on both client and server sides. If you will decide to generate your IDs only on the server, you might run into the risk of not being able to easily resolve network timeouts. On the other, hand client IDs generation process is out of your service control.</p><p>Which kind of entities needs an identifier? Let’s say your API request creates an order. You might consider having a separate ‘request ID’ for every API call. Another option is to force the client to provide an ‘order ID’ in the request and require it to be unique.</p><p>Both approaches are fine, but make sure not to mix them together. It will most certainly confuse some of your clients.</p><h3>Tip: decide how to handle async operations</h3><p>If your API offers some operations that take significant time to be completed, then it’s highly likely that you will want to introduce asynchronous processing.</p><p>The main approaches to facilitate async behavior in HTTP-based APIs are:</p><p><strong>Notifications</strong> — proactive updates from the server to the client. Can be implemented via:</p><ul><li><em>Webhooks</em> — requires the client to host an HTTP endpoint where updates will be submitted;</li><li><em>Queues</em> — using a transport alternative to HTTP like Kafka or RabbitMQ.</li></ul><p><strong>Polling</strong> — passively offering to the customer to repeatedly check for the status of the operation. Can be achieved by:</p><ul><li><em>The status endpoint</em> is offered on the server for the clients;</li><li><em>The same processing endpoint guarantees</em> <em>idempotency</em>, so it can be repeatedly called by the client with the same unique identifier until it returns the result.</li></ul><h4>Webhooks</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*PPTaqbA6IlVyxxOe" /><figcaption>Photo by <a href="https://unsplash.com/@ah360?utm_source=medium&amp;utm_medium=referral">stephen momot</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>If you decided to offer your clients notifications through webhooks then they will have to host HTTP endpoints to receive updates. URLs for these endpoints can be configured in two ways:</p><ul><li><em>Statically</em> — agree on upfront with the client where to send the notifications;</li><li><em>Dynamically</em> — the notification URL is passed from the client to the server for every request.</li></ul><p>Dynamic configuration is a bit more complex to implement, but it has its own advantages. If you and your clients will be managing simultaneously multiple environments (e.g. development, staging, and prod), then dynamic URLs will simplify the environment switches during the tests for your clients.</p><h3>Tip: implement pagination</h3><p>If you have endpoints that return lists of entities, consider implementing pagination in the early stage of your API development.</p><p>It will save your clients from potential performance issues as the amount of data grows. In most cases, you will be choosing between <strong>cursor</strong> or <strong>offset-based</strong> pagination implementations.</p><p><strong>Cursor pagination</strong></p><ul><li>Guarantees data accuracy;</li><li>Harder to implement.</li></ul><p><strong>Offset pagination</strong></p><ul><li>Easy to implement;</li><li>Clients might miss some data.</li></ul><p><strong>Useful link: </strong><a href="https://nordicapis.com/everything-you-need-to-know-about-api-pagination/">Types of API pagination</a></p><h3>Tip: define SLIs/SLOs/SLAs</h3><blockquote><em>“If you want to improve anything — you need to be able to measure it.”</em></blockquote><p>That topic deserves its own book and there is a very good one written — <a href="https://landing.google.com/sre/sre-book/toc/">Google SRE Book</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*hyLJCDftBlTcTR5O" /><figcaption>Photo by <a href="https://unsplash.com/@markusspiske?utm_source=medium&amp;utm_medium=referral">Markus Spiske</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Firstly, define your <strong>SLIs (Service Level Indicators)</strong> — a set of main metrics, that you will use to track how your service is performing. You can include into your SLIs widely used <strong>latency</strong> and <strong>uptime</strong> metrics to track the technical health of your service. On top of that, you will probably have some metrics that will indicate how your service is doing from the business perspective.</p><p>For each of your metrics or even for some combinations of those you will want to achieve some target values — those target levels are your <strong>SLOs (Service Level Objectives)</strong>. The values can be set based on the competition research on the market, come from your experience, or reflect your ambition. What’s important is that they should satisfy your customers&#39; needs. SLOs turn out to be very handy when you want to prioritize new feature developments vs clean-ups and foundation groundwork. If you <a href="https://medium.com/kudos-engineering/managing-reliability-with-slos-and-error-budgets-37346665abf6">track your SLO budget</a>, it gives you a good picture of available room for innovation.</p><p>Talk to your customers, collect feedback and agree on <strong>SLAs (Service Level Agreements)</strong>. It’s important that you and your customers state upfront what happens if SLOs are violated.</p><h3>Tip: have a versioning policy</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*I4nX2PpNKOUu-Jy2" /><figcaption>Photo by <a href="https://unsplash.com/@oowgnuj?utm_source=medium&amp;utm_medium=referral">Jungwoo Hong</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>It’s a good idea to start thinking about versioning your API in the early stage of your design. Try walking with your team through the following questions:</p><ul><li>What will be the naming format for your versions? Will it be reflected in your URL paths? If so, how exactly?</li><li>In which cases will you introduce new versions?</li><li>Will you version the endpoints separately or the API as a whole?</li><li>Will you have major and minor versions? What will be the difference?</li><li>To which extent will you support backward compatibility between your versions? Will you allow introducing breaking changes? Which changes do you consider “breaking”?</li><li>For how long do you plan to maintain older versions? What will be your deprecation / decommissioning strategy?</li></ul><p>You might not know some answers at the early design phase and that’s OK. You will learn as your service will evolve.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=547e99d0e095" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What is vocal range? This is why you should know yours.]]></title>
            <link>https://medium.com/@sergekyune/what-is-vocal-range-this-is-why-you-should-know-yours-d17ed1a19a90?source=rss-1fe7ac8fe1c5------2</link>
            <guid isPermaLink="false">https://medium.com/p/d17ed1a19a90</guid>
            <category><![CDATA[vocals]]></category>
            <category><![CDATA[repertoire]]></category>
            <category><![CDATA[singing]]></category>
            <category><![CDATA[vocal-range]]></category>
            <category><![CDATA[voices]]></category>
            <dc:creator><![CDATA[Sergey Kyune]]></dc:creator>
            <pubDate>Sun, 14 Jun 2020 19:04:44 GMT</pubDate>
            <atom:updated>2020-07-25T13:46:46.901Z</atom:updated>
            <content:encoded><![CDATA[<p>In this post I will cover:</p><ul><li>What is vocal range</li><li>Why is it useful to know your vocal range</li><li>How to find your vocal range</li><li>What voice types are and which of them your voice belongs to</li><li>What does it all have to do with <a href="https://www.singingcarrots.com/">Singing Carrots</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*PQbtv3RbLglCSTMZ" /><figcaption>Photo by <a href="https://unsplash.com/@charlesdeluvio?utm_source=medium&amp;utm_medium=referral">Charles Deluvio</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3>What is vocal range?</h3><blockquote>Your<strong> vocal range </strong>is a set of notes, that you can comfortably sing.</blockquote><p>I would like to emphasize the importance of the word “<em>comfortable” </em>here. The notes that you can hit with significant effort and with reduced sound quality (e.g. you loose in loudness, pitch stability, pressure control… ) are considered to be in your <strong>vocal range extension </strong>and are not part of your range itself<strong>.</strong></p><p>Ranges are normally identified by their boundaries. The vocal range is not an exception. On paper, the vocal range might look like this: <strong>(C3-C5). </strong>All it means is that the lowest note of the range is C3 and the highest note is C5.</p><figure><img alt="Notes from vocal range (C3-C5) on the piano. Green — notes from the range, yellow — notes from the extension" src="https://cdn-images-1.medium.com/max/1024/0*RRSLCDFUWTXozcOp.png" /><figcaption>Notes from vocal range (C3-C5) on the piano. Green — notes from the range, yellow — notes from the extension</figcaption></figure><h3>Why is it useful to know your vocal range?</h3><p>Once you find your vocal range you can:</p><h4>📐 Measure your progress.</h4><p>“If you can’t measure it, you can’t improve it” — says the wisdom. If you plan to develop your voice and learn how to sing higher notes, you should know where your limit currently lies. Only then can you exercise and track your progress.</p><blockquote>“If you can’t measure it, you can’t improve it.” (P. Ducker)</blockquote><h4>🧑‍🎤 Find songs that you can sing.</h4><p>Just like your voice, any solo song has a vocal range too. It’s similarly defined by the lowest and the highest pitches performed in the song. If you know your voice range and the range of the song, then you can easily check if you will be able to perform the piece. This knowledge turns handy if you are building up your singing repertoire or just thinking of how not to sound too embarrassing during an upcoming karaoke night.</p><p>You might be wondering how you can quickly check the range of a song you like. Or, maybe, even better, how you can get a list of songs that are within your vocal range? For that, you can use <a href="https://www.singingcarrots.com/">Singing Carrots</a>, a website with a database of vocal ranges for more than 70 000 songs.</p><h4>🔧 Adjust songs to fit your voice via transposition.</h4><p>Quite often people want to sing a song that is outside of their vocal range. The solution to this problem is called <strong>transposition</strong>. Musicians can change the key in which they perform the song by shifting all the notes of the song by a certain number of tones up or down.</p><p><strong><em>Example. </em></strong><em>Let’s say you would like to sing “Yesterday” by “The Beatles” and your vocal range is </em><strong><em>(E3-E5)</em></strong><em>.</em></p><p><em>“Yesterday” song’s vocal range is </em><strong><em>(D4-F5)</em></strong><em> and it is originally performed in the key of “</em><strong><em>F major”</em></strong><em>. The difference between your highest note </em><strong><em>E5</em></strong><em> and the song’s highest note </em><strong><em>F5</em></strong><em> is </em><strong><em>one</em></strong><em> semitone.</em></p><p><em>So, if we transpose the song </em><strong><em>one</em></strong><em> semitone down into the key of “</em><strong><em>E major” </em></strong><em>the song’s vocal range will become (</em><strong><em>C#4-E5</em></strong><em>) and you will be ready to go.</em></p><p><a href="https://www.singingcarrots.com/">Singing Carrots</a> is a service that can do all these calculations for you. All you need to do is to provide your vocal range and search for a song you want to sing. The website will outline all the possible transposition options for you.</p><h4>🌅 Make your warm-ups effective and safe.</h4><p>Vocal warm-up might be one of the most important parts of singing training. Done properly it helps to develop your voice and to protect your body from injuries. Unfortunately, if you do not know how to gradually warm up your voice, you might harm yourself rather badly. This is not much of a concern for those people who are working with a vocal coach. A qualified teacher can easily tailor your warm-ups to the specifics of your body. But what should you do if you want to practice on your own?</p><p>Warm-ups might vary greatly from one teacher to another, but conceptually a lot of coaches are following this common pattern:</p><ol><li>Start with some exercises that make use of controlled air pressure (humming, lip trills, singing through the straw, etc… )</li><li>Pick a simple note sequence and sing it through with just one vowel starting from the <strong>middle of your vocal range</strong>. Then you gradually move the sequence towards the lower and the higher boundaries of your range. After doing a couple of repetitions around the boundaries of your range, the note sequence and/or the vowel is changed and the exercise repeats.</li><li>The exercises following these two will vary widely depending on the particular vocal techniques you want to tackle.</li></ol><blockquote>In order to perform the step #2, you need to know where the middle and the boundaries of your vocal range are.</blockquote><p>Your life will be simpler if you play a musical instrument and can accompany yourself. Alternatively, there are plenty of mobile apps that you can use. Just make sure you do not blindly follow their patterns. Always adjust the exercises to your vocal range, so you can start from the pitches that are easier for your voice. Singing carrots is also offering <a href="https://www.singingcarrots.com/warm-ups"><em>interactive vocal warm ups.</em></a></p><h3>How to find your vocal range?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*8JirY1n0hqP6xv8Q" /><figcaption>Photo by <a href="https://unsplash.com/@charlesdeluvio?utm_source=medium&amp;utm_medium=referral">Charles Deluvio</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>🎹 <strong>If you play piano, guitar, or any other musical instrument</strong> with a relatively wide range of notes, then you can simply play the notes and sing along to see how low and how high you can get. Tip: start from C4 (the middle C). It’s highly likely C4 is inside your range. From there you can move up and down.</p><p>🙌 <strong>If you do not play a musical instrument </strong>or you do not have one at hand, then you can roughly estimate your vocal range by following the instructions in this video:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Fm_ZJ42ecvoM%3Fstart%3D1%26feature%3Doembed%26start%3D1&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Dm_ZJ42ecvoM&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fm_ZJ42ecvoM%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/d76787eafc80f4fa374d3698cb5911f3/href">https://medium.com/media/d76787eafc80f4fa374d3698cb5911f3/href</a></iframe><p>It’s not very precise for a couple of reasons:</p><ul><li>You might end up picking the wrong octave (e.g. E2 instead of E3)</li><li>Some people have ranges lower than E2 and higher than C6.</li><li>If your range starts with something like F#2, you will end up with G3</li></ul><p>🎙 <strong>If you want to get more precise results and have a microphone on your device </strong>you can use the interactive <a href="https://www.singingcarrots.com/tuner">Vocal Range Finder from Singing Carrots.</a></p><h3>What voice types are and which of them your voice belongs to</h3><p>This article would not be complete without a few extra words about voice types. Classical music defines the following vocal ranges with special names:</p><p><strong>Female</strong></p><ul><li>Soprano (C4-C6)</li><li>Mezzo-Soprano (A3-A5)</li><li>Contralto (F3-F5)</li></ul><p><strong>Male</strong></p><ul><li>Countertenor (E3-E5)</li><li>Tenor (C3-C5)</li><li>Baritone (A2-A4)</li><li>Bass (E2-E4)</li></ul><p>For instance, if you can cover the notes from A3 to A5 your voice can probably qualify as a Mezzo-Soprano. Congratulations! What can you do with this badge?</p><p>A lot of classical music pieces are marked with the voice type that is supposed to sing it. The knowledge of your voice type can help you to c<strong>hoose proper parts in choirs, classical pieces, musicals, etc. </strong>🏛</p><p>I hope you found this article useful. Regards, and enjoy your singing!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d17ed1a19a90" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Handshake  —  cryptocurrency without blockchain. Part 1: Introduction.]]></title>
            <link>https://medium.com/@sergekyune/handshake-cryptocurrency-without-blockchain-part-1-introduction-cac25a6847a6?source=rss-1fe7ac8fe1c5------2</link>
            <guid isPermaLink="false">https://medium.com/p/cac25a6847a6</guid>
            <category><![CDATA[cryptocurrency]]></category>
            <category><![CDATA[bitcoin]]></category>
            <category><![CDATA[crypto]]></category>
            <category><![CDATA[distributed-systems]]></category>
            <category><![CDATA[payments]]></category>
            <dc:creator><![CDATA[Sergey Kyune]]></dc:creator>
            <pubDate>Tue, 12 Nov 2019 22:19:29 GMT</pubDate>
            <atom:updated>2019-11-12T22:19:29.454Z</atom:updated>
            <content:encoded><![CDATA[<h3>Handshake — cryptocurrency without blockchain. Part 1: Introduction.</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*f8k-8dqlMC4VA4B6QhA8RA.png" /><figcaption>made with <a href="https://pablo.buffer.com/">Pablo Buffer</a></figcaption></figure><p>In these series of articles I’ll make an attempt to design a distributed payments system that could potentially replace existing blockchain based solutions. This new solution might as well coexist with the ones that are based on blockchain in order to help them evolve.</p><p>It’s a theoretical exercise based on some assumptions related to human behavioural psychology and the distribution of the existing financial flows. I’m not an expert in either of this fields so please do not throw rotten tomatoes at me. I hope you find it entertaining and, who knows, maybe some of the ideas listed below can at the end of the day become fruitful.</p><p>This first article will cover the basics of the system. To make this read productive you have to be familiar with basic blockchain and cryptocurrency concepts. There are plenty of <a href="https://lmgtfy.com/?q=introduction+to+blockchain&amp;s=d&amp;t=w">online resources</a> explaining the topic.</p><p>Before we jump into the implementation details, let’s make a pause and reason a bit. Do we really need to get rid of blockchain in the first place? What are the problems with Bitcoin and Co.?</p><ul><li><strong>Blockchain wastes tonnes of electricity due to its proof of work mechanism. </strong>Proof of Work is a concept that is used in the blockchain in order to achieve consensus between the nodes of the network. It forces all the machines in the network spend resources on solving artificially created mathematical puzzles. As the network becomes bigger, the puzzles become more complex requiring more computational resources and burning more energy. <a href="https://digiconomist.net/bitcoin-energy-consumption">Some numbers.</a></li><li><strong>Blockchain has privacy concerns</strong>. Everyone in blockchain network knows everything about everybody. You might not know who the owner of the wallet is, but you can see how much money an account has and track down the history of all the transactions that happened in the entire network since the beginning of time.</li><li><strong>Blockchain is slow. </strong>If you want to perform a transaction in blockchain — you need to put it into a block. The block needs to get signed and replicated across the network. That takes time <em>by design</em> to prevent double spending.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3rVpa_1Wcv2t3KL6Bxy-gQ.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@macu_ic?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Macu ic</a> on <a href="https://unsplash.com/s/photos/pillars?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><h3>Basis for the new solution</h3><p>To present the alternative design we will need to introduce two core principle ideas based on which our system will operate.</p><blockquote>Principle one: “Financial trust between two individuals can be measured in numbers”</blockquote><p>Imagine that Bob asks Alex to borrow him $10. Alex and Bob are good friends and the amount of money is relatively small. Alex is not worried that Bob will not return the money. She simply gives Bob the bill.</p><p>Now, let’s say, Bob asks Alex to lend him $100.000. What will Alex do? Depending on the situation various scenarios can occur. For example, Alex can ask for a written guarantee or involve some third party authority to verify the loan. Alex might also simply refuse. Who knows, maybe Bob is a good friend, but his financial situation is not. Maybe, with his income, managing such a big loan would be physically impossible.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*brLeIUmAGDWJj9EDgkeuPA.jpeg" /><figcaption>Bob tries to borrow money from Alex</figcaption></figure><p>For Alex $10 is a small enough amount that she can easily lend it to Bob. On the other hand $100.000 are too much. What if we ask Alex a question:</p><blockquote><em>What is the </em><strong><em>maximum</em></strong><em> amount of money that she is willing to lend to Bob without any guarantees?</em></blockquote><p>If Bob can not (or is not willing to) pay his debt back, then Alex will have to accept the loss. Imagine after thinking awhile Alex gave us an answer: $2300. Let’s call this number — <strong><em>credibility score</em></strong>.</p><p>Of course, one day the opposite situation can occur. Alex might want to borrow some money from Bob. Bob from his side can decide that he feels comfortable to lend Alex up to $8888. That’s how much he trusts her. <em>That’s Bob’s credibility score for Alex.</em></p><p>If we take Alex’s credibility score for Bob and Bob’s credibility score for Alex, we can use this two numbers to measure the financial relationship between the two friends.</p><p>Different pairs of selected people might assign each other completely different credibility scores. A mother might assign her daughter a credibility score of billions of dollars because she trusts her with her life. Two strangers, who’ve just met on the street, might assign each other zeros because they don’t know each other at all.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5FHaD3zedPdYhUyErtWvUg.jpeg" /><figcaption>Different pairs of financial relationships</figcaption></figure><p>If we extrapolate these examples to all humanity we can build a global bi-directional graph of financial trust between all the humans across the globe. People will be nodes on the graph and the edges will represent the credibility relationships with the weights defined by the people themselves.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8C4RQtksSUeB0Ouh-ZFqrw.jpeg" /><figcaption>Global credibility network graph</figcaption></figure><p>According to <a href="https://en.wikipedia.org/wiki/Six_degrees_of_separation">s<em>ix degrees of separation</em></a> theory we should be able to fairly easily find a path in this graph between any two individuals in the world. We will use this credibility graph as a basis for our payments network.</p><blockquote>Personal trust lies at the very foundation of this payments network idea. That’s why I decided to name it — <strong>Handshake</strong> — symbol of agreement between people who trust each other.</blockquote><blockquote>Principle two: “Anyone can run a bank”</blockquote><p>To introduce this principle let’s look at one more example. Eve decided to buy a second hand umbrella from Beth for $20. After each of them looked at the global credibility graph they realized that they both know Alex and they both trust her for more than $100. Alex happened to trust both Beth and Eve for more than $100 as well. So Eve and Beth asked Alex to facilitate a payment transaction for them.</p><p>Alex opens her notebook and starts making notes for her freshly created bank called <em>A-bank</em>.</p><p>Eve asks Alex to open for her an account in A-bank and to deposit into the account $50. She doesn’t want to go into a burden of physically giving Alex the money, so she assures her that she actually has $50 on her HSBS account. If Alex asks to actually present the money Eve will give it to her upon the request. Alex trusts Eve for more than $50 so she agrees to write down a deposit transaction of $50 on her account in A-Bank.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/811/1*pdaezSZoiVY3ImX7utmsaA.jpeg" /><figcaption>Initial accounts and deposits</figcaption></figure><p>Eve, not to forget that she owes Alex $50, now takes her notebook as well and starts making notes for Eve’s <em>E-bank</em>. She opens account for Alex in E-bank and deposits there $50 to note down her debt.</p><p>Finally, to facilitate a payment from Eve to Beth, Alex opens for Beth an account in A-Bank and makes a record for a transaction with the amount of $20 from Eve to Beth in A-bank. As for Beth it’s not urgent to cash the money out she leaves them on her account in A-bank until the day comes.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*s2ik_oSBF_mVFLMd0rLMXQ.jpeg" /><figcaption>Eve sends $20 to Beth</figcaption></figure><p>We’ve just went through the steps of how the very first payment transaction could have happened in our distributed system.</p><h3>More complex cases</h3><p>Couple of days passed and Beth found herself in need to pay for a haircut that cost $15. Hairdresser Martin happened to be a cuisine of Eve and had a credibility relationship with Eve exceeding $1000. As to the credibility graph the shortest path from node Beth to node Martin is:</p><blockquote>Beth — &gt; Alex — &gt; Eve — &gt; Martin</blockquote><p>Beth already has a bank account in A-Bank. If Martin opens an account in E-Bank then we can treat a payment from Beth to Martin as a regular bank-to-bank money transfer. Banks already figured out how this process works and we don’t have to invent anything here.</p><p>The transaction will get recorded in the following way: In A-Bank $15 will be moved from Beth’s account to Eve’s account. In E-Bank $15 first will be deposited to Alex’s account and then straight away moved to Martins account. Voila. Martin got his payment.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*77Mw8yjTs-XUAN_-tbeWaQ.jpeg" /><figcaption>Bank-to-bank transfer from Beth to Martin</figcaption></figure><p>I guess you get the idea what will happen if transaction needs to bypass more than 2 nodes. We will just iterate a bank-to-bank transfer process multiple times between each pair of the nodes in the shortest route.</p><h4>Potential problems &amp; limitations</h4><p>Thanks for bearing with me until here. It’s the end of Part one of these series of articles. In the next article(s) I will dig into the following topics:</p><ul><li>What is the incentive for people to use this network?</li><li>What to do if transaction amount exceeds the credibility weight of the graph edge?</li><li>What will be the process of dealing with potential cases of fraud?</li><li>How will the system scale if the flow of money is not cyclic, but has rather a determined constant direction?</li><li>Will everyone in the graph always have enough credibility to be able to pay?</li><li>How the system will take care of the privacy of user accounts and transactions?</li><li>Will it be able to scale performance wise?</li><li>Can legal entities connect to the system as nodes of the credibility graph?</li><li>What about government and taxes?</li><li>Is it all legal?</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cac25a6847a6" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>