<?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 Bosun Oyeleke on Medium]]></title>
        <description><![CDATA[Stories by Bosun Oyeleke on Medium]]></description>
        <link>https://medium.com/@modeleke61?source=rss-b67a1b0e065f------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*SlJ6IQBTtuEpWqic</url>
            <title>Stories by Bosun Oyeleke on Medium</title>
            <link>https://medium.com/@modeleke61?source=rss-b67a1b0e065f------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 03:12:37 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@modeleke61/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[I’m Building a Self-Custodial Lightning Wallet for Feature Phones]]></title>
            <link>https://medium.com/@modeleke61/im-building-a-self-custodial-lightning-wallet-for-feature-phones-00ecdc59d640?source=rss-b67a1b0e065f------2</link>
            <guid isPermaLink="false">https://medium.com/p/00ecdc59d640</guid>
            <dc:creator><![CDATA[Bosun Oyeleke]]></dc:creator>
            <pubDate>Wed, 16 Sep 2026 16:48:22 GMT</pubDate>
            <atom:updated>2026-09-16T16:48:22.193Z</atom:updated>
            <content:encoded><![CDATA[<p>Feature phones can already use Bitcoin Lightning.</p><p>Machankura proved that.</p><p>So when I started building Brave Bitcoin, I wasn’t trying to prove that you could send Bitcoin over USSD. That problem had already been solved.</p><p>I was interested in a different question:</p><blockquote><strong><em>Can I build a feature-phone Lightning wallet where the user actually controls the wallet?</em></strong></blockquote><p>That distinction ended up taking me much deeper into Bitcoin infrastructure than I expected.</p><h3>The idea behind Brave Bitcoin</h3><p>I wanted to build something for people who don’t necessarily have a smartphone.</p><p>The interface would be familiar:</p><pre>Feature phone<br>      ↓<br>     USSD<br>      ↓<br>Brave Bitcoin<br>      ↓<br>  Lightning</pre><ul><li>Dial a USSD code.</li><li>Choose an option.</li><li>Send sats.</li><li>Receive sats.</li><li>Check your balance.</li><li>No app store.</li><li>No smartphone.</li><li>No QR-code scanner.</li></ul><p>But there was one principle I wanted to keep from the beginning:</p><p><strong>Brave Bitcoin shouldn’t own the user’s Bitcoin.</strong></p><p>I didn’t want to build a system where users send their sats to a company-controlled wallet and the company keeps an internal balance for them.</p><p>I wanted the USSD system to be an interface to a wallet whose keys ultimately belong to the user.</p><p>That immediately creates a much harder engineering problem.</p><h3>My first attempt was LND</h3><p>I started with a fairly conventional architecture.</p><pre>Feature phone<br>      │<br>     USSD<br>      │<br>      ▼<br>   Flask API<br>      │<br>      ▼<br>     LND<br>      │<br>      ▼<br>Lightning Network</pre><p>The USSD side was Python and Flask.</p><p>The Lightning side was LND.</p><p>The idea was straightforward: receive a USSD request, perform the appropriate wallet operation, and return the result to the phone.</p><p>The prototype got me far enough to understand that the concept was technically possible.</p><p>Then I ran into something much less exciting than Lightning:</p><p><strong>hardware.</strong></p><h3>My hardware became part of the architecture</h3><p>I wasn’t running this on a giant server.</p><p>The machine I was using had limited resources, and I was already running Bitcoin infrastructure on it.</p><p>LND added another significant workload.</p><p>At some point I had to decide whether the answer was:</p><blockquote><em>Buy more hardware.</em></blockquote><p>or:</p><blockquote><em>Find a lighter architecture.</em></blockquote><p>I chose the second.</p><p>That led me to Rust.</p><p>More specifically, to <strong>LDK Node</strong>.</p><h3>Enter LDK Node</h3><p>The attraction was that I could build the Lightning component into a Rust application rather than treating Lightning as another heavyweight daemon that had to sit beside everything else.</p><p>So I created:</p><pre>~/brave-ln</pre><p>and started building around LDK Node.</p><p>The architecture became:</p><pre>Bitcoin infrastructure<br>        │<br>        ▼<br>     LDK Node<br>        │<br>        ▼<br>Lightning Network<br>        │<br>        ▼<br>   Brave Bitcoin<br>        │<br>        ▼<br>       USSD</pre><p>This looked much more promising.</p><p>Then Bitcoin Core reminded me that every dependency has a cost.</p><h3>The Bitcoin Core problem</h3><p>I already had Bitcoin Core running.</p><p>So my first instinct was to connect LDK Node directly to it through Bitcoin Core’s RPC interface. That sounds perfectly reasonable.</p><p>And technically, it was. The problem was performance.</p><p>My Bitcoin node was running on constrained hardware with the blockchain data on an external drive. Some RPC operations were taking dramatically longer than I expected.</p><p>I started measuring them. For example, estimatesmartfee was taking around a minute and a half.getblockchaininfo was also taking around a minute and a half. And then I tested getblockhash.</p><p>The response took roughly: <strong>nine minutes. </strong>Nine Minutes for an RPC call.</p><p>That was when the architecture started making more sense. The Lightning node wasn’t necessarily the problem. <strong>My chain backend was the bottleneck.</strong></p><h3>LDK started complaining</h3><p>LDK Node eventually failed during startup with:</p><pre>FeerateEstimationUpdateTimeout</pre><p>At first, I treated this like another Lightning problem.</p><p>But the more I investigated, the more obvious it became that LDK was waiting for information from the chain backend.</p><p>Bitcoin Core could answer. It was just answering very slowly.</p><p>This was one of those moments where debugging changed the way I understood the system.</p><p>I wasn’t simply building:</p><blockquote><em>“a Lightning wallet.”</em></blockquote><p>I was building a chain of dependencies:</p><pre>Lightning application<br>        ↓<br>      LDK<br>        ↓<br>   chain backend<br>        ↓<br>   Bitcoin Core<br>        ↓<br>      disk</pre><p>If one layer is slow enough, the layers above it don’t care why.</p><p>They just time out.</p><h3>So I changed the chain source:</h3><p>Rather than keep fighting the slow RPC path, I switched LDK Node to an Esplora backend for this stage of the experiment.</p><p>The configuration became:</p><pre>builder.set_chain_source_esplora(<br>    &quot;https://blockstream.info/api&quot;.to_string(),<br>    None,<br>);</pre><p>And finally, LDK Node started.</p><p>That was a pretty satisfying moment.</p><p>I got:</p><pre>Brave Bitcoin Lightning Node</pre><p>with a real mainnet node identity.</p><p>The node reported its mainnet chain state and its wallet synchronization state. I could generate an actual Bitcoin funding address. The node was running.</p><h3>This was no longer just a prototype</h3><p>At this point, I had something real.</p><p>My node generated a mainnet funding address:</p><pre>bc1qc74snakrq7e9zqu9vsrhvd5qtqelwqc9xeu3sp</pre><p>It had its own Lightning node identity:</p><pre>03d906282552b291b45442075a86a01dc02da7a648a080c8daf8c83721d7cc2dae</pre><p>And it was listening for Lightning connections. The next step was obvious. I needed peers.</p><h3>Connecting to the Lightning Network</h3><p>I configured real Lightning peers.</p><p>One of them was ACINQ.</p><p>The application reported:</p><pre>Connecting to ACINQ...<br>ACINQ connected.</pre><p>Then I connected to another test peer:</p><pre>Connecting to 50k test peer...<br>50k test peer connected.</pre><p>That was another milestone. The node wasn’t just sitting there starting and stopping. It was participating in the Lightning peer-to-peer network.</p><p>The architecture at that point looked roughly like:</p><pre>                Bitcoin Mainnet<br>                       │<br>                       ▼<br>                   LDK Node<br>                       │<br>              ┌────────┴────────┐<br>              │                 │<br>            ACINQ          Test Peer<br>              │                 │<br>              └────────┬────────┘<br>                       │<br>                 Lightning Network</pre><p>And that was the point where I realized: <strong>Okay. This thing is becoming real.</strong></p><h3>But I still didn’t have a wallet for other people</h3><p>This is an important distinction. I now had a working mainnet Lightning node. But that doesn’t mean I had finished Brave Bitcoin. And it certainly doesn’t mean I had solved the self-custody problem.</p><p>A Lightning node that I control is still a wallet that I control. If I simply created accounts for thousands of users inside my own node and kept their balances there, I’d be moving toward a custodial architecture.</p><p>That isn’t what I want Brave Bitcoin to be.</p><h3>The self-custody problem</h3><p>This is where the project becomes much more interesting.</p><p>Imagine a user with a basic feature phone.</p><p>They dial:</p><pre>*123#</pre><p>and see:</p><pre>BRAVE BITCOIN</pre><pre>1. Balance<br>2. Send<br>3. Receive<br>4. Pay<br>5. Backup Wallet</pre><p>That’s easy enough.</p><p>But now select:</p><pre>5. Backup Wallet</pre><p>A smartphone wallet can show a recovery phrase.</p><p>It can display QR codes.</p><p>It can let you copy text.</p><p>A feature phone can’t do most of those things.</p><p>So how do you give someone control of their keys through a tiny USSD interface?</p><p>That became one of the hardest questions in the entire project.</p><h3>USSD is only the interface</h3><p>The architecture I actually want looks more like this:</p><pre>              USER<br>                │<br>                ▼<br>          Feature Phone<br>                │<br>               USSD<br>                │<br>                ▼<br>         Brave Bitcoin<br>           Interface<br>                │<br>                ▼<br>       User-controlled wallet<br>                │<br>                ▼<br>        Lightning Network</pre><p>The critical part is the middle.</p><p><strong>Brave Bitcoin should not become the wallet owner.</strong></p><p>The service should provide the interface and infrastructure.</p><p>The user should ultimately control the keys.</p><p>That means wallet creation, backup, recovery, signing, and Lightning channel recovery all have to be thought through differently.</p><h3>The recovery phrase problem</h3><p>One idea I’m exploring is using a standard recovery phrase and presenting it through USSD in manageable chunks.</p><p>Something like:</p><pre>BACKUP WALLET</pre><pre>1. Word 1: ...<br>2. Word 2: ...<br>3. Word 3: ...<br>4. Word 4: ...</pre><pre>NEXT</pre><p>Then:</p><pre>5. Word 5: ...<br>6. Word 6: ...<br>7. Word 7: ...<br>8. Word 8: ...</pre><p>The user writes the words down.</p><p>Then Brave Bitcoin can perform a backup verification step by asking the user to provide selected words.</p><p>The important thing is that the recovery phrase should not become something that Brave Bitcoin secretly stores on behalf of the user.</p><p>The user needs to be able to recover their wallet independently.</p><p>And Lightning adds another complication:</p><p><strong>a seed isn’t necessarily the entire recovery story for an active Lightning wallet.</strong></p><p>Channel state matters too.</p><p>So even after solving seed backup, there is still work to do around Lightning-specific recovery.</p><h3>The next experiment: real sats</h3><p>So far, most of my work has been infrastructure.</p><p>The next step is much simpler to explain:</p><p><strong>put real Bitcoin through it.</strong></p><p>I don’t want to start with a huge amount. The plan is to work with a small amount of sats and establish a small Lightning channel.</p><p>Then:</p><pre>Fund node<br>    ↓<br>Open channel<br>    ↓<br>Obtain inbound liquidity<br>    ↓<br>Create invoice<br>    ↓<br>Pay invoice<br>    ↓<br>Receive sats<br>    ↓<br>Send sats back</pre><p>That will be a much more meaningful milestone than simply seeing:</p><pre>Node is running.</pre><p>A running node is infrastructure. A completed real payment is a working payment system.</p><h3>What I’ve actually accomplished</h3><p>I think it’s worth being precise here.</p><p>At this point I can honestly say that I have:</p><ul><li>built a USSD Bitcoin wallet prototype;</li><li>experimented with LND;</li><li>encountered resource constraints with LND;</li><li>moved the Lightning implementation to Rust and LDK Node;</li><li>attempted to use Bitcoin Core as the chain backend;</li><li>diagnosed severe Bitcoin Core RPC latency on my hardware;</li><li>encountered LDK’s FeerateEstimationUpdateTimeout;</li><li>switched the LDK chain source to Esplora;</li><li>successfully started an LDK Node on Bitcoin mainnet;</li><li>generated a real Bitcoin mainnet funding address;</li><li>established a persistent Lightning node identity;</li><li>connected the node to real Lightning peers.</li></ul><p>What I <strong>haven’t</strong> done yet is equally important:</p><ul><li>I haven’t finished the production USSD wallet.</li><li>I haven’t opened the first production Lightning channel.</li><li>I haven’t completed the first real payment through the entire USSD stack.</li><li>And I haven’t completely solved the feature-phone self-custody and Lightning recovery experience.</li><li>So I’m not calling this a finished wallet.</li></ul><p>I’m calling it an experiment that has reached a real milestone.</p><h3>Why not just use an existing solution?</h3><p>Because existing solutions have already demonstrated that the basic problem can be solved.</p><p>That’s actually what makes the experiment interesting.</p><p>Machankura showed that:</p><p><strong>Feature phone + USSD + Lightning is possible.</strong></p><p>I’m exploring a different question:</p><p><strong>What does the architecture look like if self-custody is treated as a fundamental requirement?</strong></p><p>That changes the engineering decisions.</p><p>It changes where keys live.</p><p>It changes how wallet recovery works.</p><p>It changes how the backend should be designed.</p><p>And it changes what Brave Bitcoin can and cannot do for the user.</p><h3>The business question comes later</h3><p>There is also an interesting business model hiding behind the infrastructure.</p><p>If Brave Bitcoin doesn’t custody user funds, it doesn’t necessarily mean the service can’t make money.</p><p>The business could potentially charge for infrastructure and services:</p><ul><li>USSD access;</li><li>merchant APIs;</li><li>payment processing infrastructure;</li><li>Lightning liquidity services;</li><li>developer APIs;</li><li>enterprise deployments;</li><li>premium merchant functionality.</li></ul><p>The important distinction is:</p><blockquote><strong><em>Charge for the service, not for taking possession of the user’s Bitcoin.</em></strong></blockquote><p>But that’s a problem for another stage of the project.</p><p>Right now I need to make the thing work.</p><h3>What I’ve learned so far</h3><p>The biggest lesson has been that building a Bitcoin application isn’t just about writing the application.</p><p>Every layer underneath it matters.</p><p>My original problem looked like:</p><pre>&quot;How do I build a Lightning wallet?&quot;</pre><p>It turned into:</p><ul><li>How much RAM does my Lightning implementation need?</li><li>How fast is my Bitcoin backend?</li><li>How does LDK obtain fee information?</li><li>How does it synchronize?</li><li>How does it find peers?</li><li>How does it recover?</li><li>Where do the user&#39;s keys live?</li><li>How does a feature-phone user back them up?</li><li>How does Lightning channel recovery work?</li></ul><p>The deeper I go, the less this feels like simply building a wallet.</p><p>It’s becoming an exercise in <strong>Bitcoin infrastructure, distributed systems and human-computer interaction</strong>.</p><h3>Where Brave Bitcoin is now</h3><p>The project started with a feature phone.</p><ul><li>It led me into Bitcoin Core.</li><li>Then LND.</li><li>Then Rust.</li><li>Then LDK Node.</li><li>Then Bitcoin RPC performance.</li><li>Then Esplora.</li><li>Then Lightning peer connections.</li></ul><p>And now I’m at the point where the next step is real sats.</p><p>The final architecture isn’t finished. The product isn’t finished.</p><p>But the experiment has moved from:</p><blockquote><strong><em>“Could this work?”</em></strong></blockquote><p>to:</p><blockquote><strong><em>“How do I make this actually work?”</em></strong></blockquote><p>And that’s a much more interesting place to be.</p><p>The original question is still sitting there:</p><blockquote><strong><em>Can I build a genuinely self-custodial Lightning experience that someone can use from a feature phone?</em></strong></blockquote><p>I don’t know the final answer yet. But now I have a mainnet Lightning node, real peers, a USSD prototype, and a lot more questions than I started with.</p><p>So I’m going to keep building.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=00ecdc59d640" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Distributed Intelligent LLM Gateway with Adaptive Routing.]]></title>
            <link>https://medium.com/@modeleke61/distributed-intelligent-llm-gateway-with-adaptive-routing-13027b1818f1?source=rss-b67a1b0e065f------2</link>
            <guid isPermaLink="false">https://medium.com/p/13027b1818f1</guid>
            <category><![CDATA[grafana]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[docker-swarm]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[prometheus-monitoring]]></category>
            <dc:creator><![CDATA[Bosun Oyeleke]]></dc:creator>
            <pubDate>Sun, 21 Jun 2026 03:49:26 GMT</pubDate>
            <atom:updated>2026-06-29T02:38:07.636Z</atom:updated>
            <content:encoded><![CDATA[<p><em>From zero to a multi-replica, observable, self-learning LLM router, here’s what I built and what I learned.</em></p><p>Two weeks ago, I was reading Anthropic’s Inference team job description. They talked about intelligent request routing, observability at scale, adaptive systems, and serving models to millions of users.I didn’t have that experience on my resume. So I decided to build it. In just two days, I created AI Gateway, a distributed, intelligent LLM inference system that dynamically routes requests to the best available model.</p><h4>What I Built:</h4><ul><li>Smart Adaptive Router: Analyses prompt complexity and real-time model performance (latency + tokens) to choose the optimal model (from 3B to 20B).</li><li>Distributed Architecture: 3 gateway replicas running behind Nginx, with Redis for shared state.</li><li>Full Observability: Prometheus metrics, beautiful Grafana dashboards, and OpenTelemetry tracing.</li><li>Self-Learning System: The router gets smarter the more it runs.</li></ul><p>The system supports five models and automatically routes:</p><ul><li>Simple queries → small &amp; fast models</li><li>Coding tasks → specialised coder models</li><li>Heavy reasoning → larger models (with a fallback logic)</li></ul><h4><strong>Stack</strong></h4><ul><li>Python + Flask (evolved from an initial Rust + Axum version)</li><li>Docker Swarm + 3 replicas</li><li>Redis for shared model statistics</li><li>Nginx load balancer</li><li>Prometheus + Grafana + OpenTelemetry</li></ul><h4>Why This Matters:</h4><p>Companies like Anthropic don’t just need people who can call APIs. They need engineers who understand the hard parts of inference: routing strategy, performance trade-offs, observability, reliability at scale, and continuous optimization.This project forced me to think about all of those things. Key Learnings</p><ul><li>Prompt complexity scoring is surprisingly effective</li><li>Real-time performance feedback loops make routing significantly better</li><li>Observability isn’t optional, it’s the only way to improve a system like this</li><li>Docker Swarm makes horizontal scaling much easier than I expected</li></ul><h4>Screenshots</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4rA8ZS2uoY52GrCVdyaxqw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JWAk1-iEUZ6pNdgPVdLKdw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/812/1*wx6EwDXdQbSoMECNtYJNcA.png" /><figcaption>Grafana Main Dashboard with manually configured panels.</figcaption></figure><h4>Panels and the short description:</h4><ol><li>Model Usage Distribution &amp; LLM Requests</li></ol><ul><li>Shows which models are being used most frequently.</li><li>Helps identify if the router is properly balancing load or over-relying on one model.</li></ul><p>2. Smart Routing Distribution</p><ul><li>Displays the percentage of requests routed through different paths (fast_path, reasoning_path, code_path, premium_path).</li><li>Validates that the intelligent router is making logical decisions based on prompt complexity.</li></ul><p>3. Average Latency per Model</p><ul><li>Gauge showing average response time for each model.</li><li>Critical for spotting under-performing models (e.g., when gpt-oss:20b becomes too slow).</li></ul><p>4. Adaptive Routing Score</p><ul><li>Real-time performance score for each model (higher = better).</li><li>Combines latency and tokens/sec. The router uses this to make better decisions over time.</li></ul><p>5. Tokens Per Request</p><ul><li>Shows how many tokens each model is generating on average.</li><li>Useful for understanding model behaviour and cost/efficiency.</li></ul><p>6. Prompt Complexity Over Time</p><ul><li>Line graph showing how complex the incoming prompts are.</li><li>Helps correlate routing decisions with prompt difficulty.</li></ul><p>7. Model Selections</p><ul><li>Bar chart showing how many times each model was chosen.</li><li>Great for monitoring routing fairness and load distribution.</li></ul><p>8. Resource Usage (CPU &amp; Memory)</p><ul><li>Monitors host resource consumption during inference.</li><li>Important when running larger models like 20B.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=13027b1818f1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How I Set Up UmbrelOS on Raspberry Pi 5 (And How You Can Too).]]></title>
            <link>https://medium.com/@modeleke61/how-i-set-up-umbrelos-on-raspberry-pi-5-and-how-you-can-too-f2f5346cd56a?source=rss-b67a1b0e065f------2</link>
            <guid isPermaLink="false">https://medium.com/p/f2f5346cd56a</guid>
            <category><![CDATA[bitcoin]]></category>
            <category><![CDATA[docker-compose]]></category>
            <category><![CDATA[lightning-network]]></category>
            <category><![CDATA[lightning-node]]></category>
            <category><![CDATA[raspberry-pi-5]]></category>
            <dc:creator><![CDATA[Bosun Oyeleke]]></dc:creator>
            <pubDate>Wed, 06 Aug 2025 22:51:17 GMT</pubDate>
            <atom:updated>2025-08-06T22:51:17.965Z</atom:updated>
            <content:encoded><![CDATA[<h3>🚀 Introduction</h3><p>Setting up your own Bitcoin and Lightning node has never been easier, or more important. With tools like <strong>UmbrelOS</strong>, you can take back control of your financial data and be part of the decentralised future. In this post, I’ll walk you through how I set up <strong>UmbrelOS on a Raspberry Pi 5</strong>, using it as a full Bitcoin node and personal server, and how <strong>you can do it too</strong>.</p><h3>🧰 What You’ll Need</h3><p>Before diving in, here’s what I used:</p><ul><li>🖥️ Raspberry Pi 5 (4GB or 8GB RAM)</li><li>⚡ Official Raspberry Pi 5 USB-C Power Supply (or equivalent, min 5V/5A)</li><li>💽 SSD (1TB or more)</li><li>🔌 USB-2.0 to USB-C cable (for SSD)</li><li>🌐 Ethernet cable (recommended) or Wi-Fi</li><li>🧱 MicroSD card (16GB+) — <em>for initial setup</em></li><li>MicroSD Reader to format the SD card.</li><li>🖼️ Optional: TFT or HDMI display with keyboard/mouse</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YCr1F0B9wK5mf0qqzRznfQ.png" /><figcaption>Bill of Components</figcaption></figure><h3>Step-by-Step Setup</h3><h4>1. 🔥 Flash Raspbian OS</h4><ul><li>Download the Raspberry Pi Imager</li><li>Select the 64bit OS Lite version and <strong>MicroSD card</strong></li></ul><h4>2. 🧠 Update Raspberry Pi Bootloader</h4><p>Proceed to the advanced settings:</p><ul><li>Set <strong>hostname</strong> (e.g., umbrelpi)</li><li>Enable <strong>SSH</strong></li><li>Set a <strong>username/password</strong></li><li>Optionally configure <strong>Wi-Fi</strong> (but Ethernet is more reliable)up.</li></ul><h4>3. 🧵 Connect Everything</h4><ul><li>Plug in your SSD, list it with the command below and mount it additionally:</li></ul><pre>lsblk #lists the volumes attached, here output looks like this /dev/sda1, etc.<br><br><br>sudo mkdir -p /mnt/data #create the mount point for the SSD<br>sudo mount /dev/sda1 /mnt/data #mount the SSD <br><br>sudo systemctl stop umbrel #stop umbrel from running<br>sudo mv ~/umbrel /mnt/data/ #move umbrel folder to SSD<br><br>sudo nano /etc/fstab #Auto-mounting SSD on boot.<br>/etc/fstab: put the line below into the file and save., Ctrl + X.<br><br>/dev/sda1 /mnt/data ext4 defaults 0 2<br><br><br>sude reboot</pre><ul><li>Connect power, Ethernet or Wi-Fi (though wired is better)</li><li>Optional: connect a monitor and keyboard to observe logs</li></ul><h3><strong>Where the Real Magic Works</strong></h3><ol><li>Install 🐳 Docker on Raspberry Pi 5:</li></ol><pre>curl -fsSL https://get.docker.com -o get-docker.sh<br>sudo sh get-docker.sh</pre><p>Just in-case any errors related to docker-compose, the issue here is in <strong>scripts/start, </strong>what’s going on? The start script is looking for docker-compose which is the v1 but we have the v2 installed, so use the command below to</p><pre>sudo tee /usr/local/bin/docker-compose &gt; /dev/null &lt;&lt; &#39;EOF&#39;<br>#!/bin/bash<br>docker compose &quot;$@&quot;<br>EOF<br><br>sudo chmod +x /usr/local/bin/docker-compose</pre><p>2. UmbrelOS: Install this via the command line</p><pre>git clone https://github.com/getumbrel/umbrel #clone the OS<br>cd umbrel #change directory</pre><p>3. Start Bitcoin Knots Node:</p><pre>sudo ./scripts/start #start the umbrelOS</pre><h3><strong>Bitcoin Knots Node</strong></h3><p><strong>Bitcoin Knots</strong> is a full node software, just like <strong>Bitcoin Core</strong>, but with <strong>more advanced features and improvements</strong>. It’s fully compatible with the Bitcoin network.</p><p>At this point you’d be able to connect to your node, if you had no errors after using the script from the previous section.</p><h4>Option A: SSH</h4><p>If you setup SSH according to instructions, connecting to your pi should be easy via the terminal:</p><pre>ssh username@umbrelpi.local #username being the one you set when flashing the SD card</pre><h4>Option B: Chrome Browser</h4><p>For this option, you have to be on the same network as your Pi. Assuming you’re on the same Wi-Fi network, you’d have to figure out the IP address for the Pi.</p><p>You can get the IP address via the terminal using</p><pre>hostname -I #this first IP address of this output, is what you need.</pre><p>Take the ip address and paste into your address bar, you should be able to access the UmbrelOS dashboard or <a href="http://umbrel.local/"><strong><em>http://umbrel.local</em></strong></a></p><p><strong>Bitcoin Knots isn’t installed by default</strong>, so head over to the app store and download the app, this would take a few seconds to minutes depending on internet speed, when complete the blockchain starts synchronising from the inital block.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_X-zdV0GfHTqgltP2uaNeA.png" /><figcaption>UmbrelOS App Store, download Bitcoin Knots.</figcaption></figure><p>A few caveats on why i used this method, I have no NVMe hat which would’ve simplified my setup, with an NVMe hat and an NVMe card of about 1TB, the setup would’ve been simplified because Umbrel has the Raspberry Pi Image but in the absence of an NVMe, hence this method.</p><p>Happy Hacking !</p><p>For More on Umbrel OS, check out their Github and there’s loads of videos on YouTube.</p><h4>Resources:</h4><ol><li><a href="https://www.raspberrypi.com/software/">Raspberry Pi Imager</a></li><li><a href="https://github.com/getumbrel/umbrel">UmbrelOS GitHub</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f2f5346cd56a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Breathing Port Harcourt: Air Pollution and Population in Nigeria’s Oil Hub]]></title>
            <link>https://medium.com/@modeleke61/breathing-port-harcourt-air-pollution-and-population-in-nigerias-oil-hub-07b838aaff30?source=rss-b67a1b0e065f------2</link>
            <guid isPermaLink="false">https://medium.com/p/07b838aaff30</guid>
            <category><![CDATA[nigeria]]></category>
            <category><![CDATA[environmental-impact]]></category>
            <category><![CDATA[port-harcourt]]></category>
            <category><![CDATA[gas-flaring]]></category>
            <category><![CDATA[environmental-issues]]></category>
            <dc:creator><![CDATA[Bosun Oyeleke]]></dc:creator>
            <pubDate>Wed, 25 Jun 2025 04:11:40 GMT</pubDate>
            <atom:updated>2025-06-25T04:12:49.275Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/526/1*lv5M4ItIDXUON1q7f9e8YA.gif" /></figure><p>This is the third city in my series on urban air quality in Nigeria, following Abuja and Lagos. This time, I turn my lens and satellite data to Port Harcourt, a city known globally for its oil wealth, but also grappling with environmental and public health challenges.</p><h3>What Is NO₂ and Why It Matters Here?</h3><p>Nitrogen Dioxide (NO₂) is a reddish-brown gas released primarily by combustion from car engines, gas flaring, generators, and industrial activity. It’s toxic to lungs, especially for children and vulnerable populations. In a city like Port Harcourt, where crude oil is being produced and urban life thrives, the risks are magnified.</p><p>Using satellite data from 2019 to 2025, I mapped NO₂ concentrations over Port Harcourt and linked it to land cover types and population density. I used data from Sentinel-5P, WorldPop, MODIS, and Google Earth Engine to analyse this complex relationship.</p><p><strong>Interpreting the Visual</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Z9Mc9vyI-Oif32D6czwD-Q.png" /></figure><p>In this scatterplot, I explored the relationship between NO₂ air pollution and population density across different land cover types in Port Harcourt, Nigeria, using a logarithmic scale on both axes. Each dot represents a location I sampled from satellite data collected between 2019 and 2025, with colours indicating their respective land cover classifications. I noticed that Urban and Built-Up areas (green dots) tend to cluster in the upper-right, signalling zones with both high population density and elevated NO₂ levels, likely due to emissions from vehicles, generators, and industrial activities in densely populated neighbourhoods. In contrast, land cover types such as Wetlands, Savannahs, and Cropland/Natural Vegetation Mosaics are mostly found in the bottom-left, where both human activity and pollution levels remain relatively low, indicating ecologically safer spaces. A few outliers stood out, where rural-looking areas exhibited unexpected spikes in NO₂ perhaps reflecting local industrial influence or encroaching urban sprawl. By using a log scale, I could better observe how even small increases in population density often lead to sharp rises in NO₂ exposure, emphasising the urgent need for better air quality planning in cities like Port Harcourt.</p><p><strong>Urbanisation Drives NO₂ Pollution, But Rural Hotspots and Urban Sprawl Raise Environmental Awareness and Policy Concerns</strong></p><p>From the data, the relationship is unmissable: urbanisation clearly correlates with worse NO₂ pollution. The green dots representing urban and built-up land dominate the upper-right of the chart, where population and NO₂ levels are both high. This reinforces what many residents already experience firsthand: traffic congestion, diesel generators, and industrial emissions are all part of daily life in Port Harcourt’s urban core. But what surprised me were some non-urban zones showing elevated pollution levels, areas classified as wetlands, or cropland mosaics. These are supposed to be cleaner, greener spaces. So why the spike? It could be gas flaring near rural communities, unregulated industrial activity. These outliers are a red flag and a reason to investigate further.</p><p><strong>What This Means for Port Harcourt?</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7M0D7uL9eJNnjPB_bSGcCQ.png" /></figure><p>Port Harcourt’s people are breathing in a legacy of fossil fuel dependence and poor planning. With rising population comes higher pollution. The data isn’t just numbers, it reflects daily reality, acid rains, dilapidation of the roofing materials of the city, illegal bunkering, unregulated industrial activities and just bare lungs absorbing it all.</p><p><strong>Resources &amp; GitHub:</strong></p><ul><li>📁 🔗 GitHub link with notebook + plots: <a href="https://github.com/olatunbosunoyeleke94/phc-no2">https://github.com/olatunbosunoyeleke94/phc-no2</a></li><li>🔗 Dataset credits (<a href="https://developers.google.com/earth-engine/datasets/catalog/COPERNICUS_S5P_NRTI_L3_NO2">Sentinel-5P</a>, <a href="https://www.worldpop.org/">WorldPop</a>, <a href="https://developers.google.com/earth-engine/datasets/catalog/MODIS_006_MCD12Q1">MODIS</a>)</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=07b838aaff30" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What NO₂ Says About Lagos: Mapping Air Pollution and Population Exposure with Google Earth Engine]]></title>
            <link>https://medium.com/@modeleke61/what-no%E2%82%82-says-about-lagos-mapping-air-pollution-and-population-exposure-with-google-earth-engine-4fa84d5614c4?source=rss-b67a1b0e065f------2</link>
            <guid isPermaLink="false">https://medium.com/p/4fa84d5614c4</guid>
            <category><![CDATA[lagos]]></category>
            <category><![CDATA[urban-planning]]></category>
            <category><![CDATA[health]]></category>
            <category><![CDATA[youth-development]]></category>
            <category><![CDATA[air-pollution]]></category>
            <dc:creator><![CDATA[Bosun Oyeleke]]></dc:creator>
            <pubDate>Wed, 25 Jun 2025 02:56:46 GMT</pubDate>
            <atom:updated>2025-06-25T03:00:05.402Z</atom:updated>
            <content:encoded><![CDATA[<p>This article is the second in a series exploring how air pollution intersects with urban growth in Nigerian cities.</p><p>Following my first report on Abuja, where I examined how traffic, urban sprawl, and topography influence nitrogen dioxide (NO₂) levels, this piece turns to Lagos, Nigeria’s bustling megacity. Using satellite data and Python-based geospatial analysis, I map the places where Lagos breathes clean air and the places where it doesn’t.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*hzc8-U-uyqYsq2Jj2hW1cQ.gif" /></figure><p><strong>You might be asking, <em>what exactly is NO₂?</em></strong><br><strong>Nitrogen Dioxide (NO₂)</strong> is a reddish-brown gas and a major air pollutant It belongs to a group of highly reactive gases known as nitrogen oxides (NOₓ). It’s primarily produced by Combustion engines (cars, trucks, generators), Industrial facilities, Power plants, Burning of fossil fuels and biomass.</p><p><strong>Wondering how dangerous it is to human health?</strong><br>NO₂ causes respiratory irritation, can trigger asthma attacks, and is especially harmful to children and the elderly. Long-term exposure is linked to reduced lung function and increased hospital admissions for bronchitis, asthma, and other pulmonary diseases.</p><h3>Indirect &amp; Environmental Impact of NO₂</h3><p>NO₂ contributes to the formation of ground-level ozone (O₃) and fine particulate matter (PM₂.₅), both of which are longer-lasting and more harmful pollutants that worsen cardiovascular diseases and may affect brain health over time; it also contributes to acid rain, damages vegetation and biodiversity, and reduces visibility, especially in traffic-congested urban areas.</p><p>Using data from <strong>Sentinel-5P, WorldPop, </strong>and<strong> ESA</strong> land cover classifications, I investigated:</p><ul><li>How NO₂ pollution is distributed across Lagos’ diverse urban and ecological zones</li><li>Which population areas, experience the greatest exposure</li><li>What land use types dominate in areas with the highest pollution load</li></ul><p>The analysis was conducted using Google Earth Engine, Python, and geospatial analysis libraries like geemap, with visualisations helping to tell the story of how Lagos breathes, and who’s most at risk.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*t5JhlvIj9mBS23I9lmSd7w.png" /></figure><p>In the top-right corner, orange dots representing urban areas dominate. These are zones of high population density and high NO₂ levels. The correlation is sharp, as people cluster in Lagos’ built-up areas, so does pollution. This is no coincidence. Traffic congestion, constant generator use, and industrial activity all pour nitrogen dioxide into the air and the people living there breathe it in daily. These urban zones are your exposure hotspots.</p><p>In stark contrast, the bottom-left of the chart tells a different story. Here savannas, wetlands, and grasslands quiet, have green spaces with lower human density and cleaner air. These are not just low-exposure zones; they are also potential buffers against urban sprawl and could be priority areas for conservation.</p><p>Some dots defy the pattern outliers. These may represent peri-urban zones where pollution has crept in before people fully have, or vice versa growing populations in areas not yet classified as urban, but already showing signs of degradation.</p><p>Importantly, the logarithmic scale used in both axes reveals a hidden truth, small increases in population density can trigger exponential rises in NO₂ levels. In other words, pollution isn’t just about the size of a population, it’s about the intensity and clustering of urban life.</p><p>This chart doesn’t just visualise data, it maps the lived air quality experience across Lagos, and offers a window into how urban planning, energy access, and mobility policy shape the health of a megacity.</p><p>In my next report, I would be looking into Port Harcourt, a city in the Southern Region of Nigeria, I can’t wait to see the what the NO₂ looks like.</p><h3>Resources &amp; Credits</h3><p><strong>GitHub Repository:</strong><br> Access the full notebook, and visualisation assets:<br> 🔗 <a href="https://github.com/olatunbosunoyeleke94/lagos-no2-population-exposure">https://github.com/olatunbosunoyeleke94/lagos-no2-population-exposure</a></p><p><strong>Data Sources:</strong></p><ul><li><strong>Sentinel-5P NO₂ (TROPOMI):</strong><br> <a href="https://developers.google.com/earth-engine/datasets/catalog/COPERNICUS_S5P_NRTI_L3_NO2">Google Earth Engine Dataset</a></li><li><strong>WorldPop Population Data:</strong><br> <a href="https://www.worldpop.org/">https://www.worldpop.org/</a></li><li><strong>ESA WorldCover Land Use Classification:</strong><br> <a href="https://developers.google.com/earth-engine/datasets/catalog/ESA_WorldCover_v200">ESA WorldCover v200</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4fa84d5614c4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What NO₂ Says About Abuja: Mapping Air Pollution and Population with Python & Earth Engine.]]></title>
            <link>https://medium.com/@modeleke61/what-no%E2%82%82-says-about-abuja-mapping-air-pollution-and-population-with-python-earth-engine-f8b33613ad13?source=rss-b67a1b0e065f------2</link>
            <guid isPermaLink="false">https://medium.com/p/f8b33613ad13</guid>
            <dc:creator><![CDATA[Bosun Oyeleke]]></dc:creator>
            <pubDate>Sun, 22 Jun 2025 01:19:37 GMT</pubDate>
            <atom:updated>2025-06-22T01:32:43.137Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hn-YpIjI_uVYhREhXO8W-g.png" /></figure><p>Air quality has become a quiet but critical concern in Nigeria’s fast-growing cities. This short study focuses on Abuja, using satellite data to explore nitrogen dioxide (NO₂) levels between 2019 and June 2025. I took a look at how NO₂ varies across land cover types, especially urban zones and whether population density correlates with pollution.</p><p><strong>Data &amp; Method Summary:</strong></p><p>Using data from Google Earth Engine, combining:</p><ul><li>NO₂ data from the Sentinel-5P satellite</li><li>Land cover data (MODIS MCD12Q1)</li><li>Population density from the GPWv4 dataset (last updated in 2020)</li></ul><p>Focusing on <strong>Abuja</strong>, 500 random points were sampled, and analysed NO₂ concentrations in relation to land type and population levels.</p><p><em>Link to GitHub:</em><br> 📎 <em>See full code and notebook: </em><a href="https://github.com/olatunbosunoyeleke94/UrbanExpansionStudy/">https://github.com/olatunbosunoyeleke94/UrbanExpansionStudy/</a></p><p>Over five years of data (Jan 2019 — June 2025), seasonal fluctuations in NO₂ were observed. Here’s an animated GIF showing how levels vary monthly:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/554/1*OHCG2TCnR-WB7wNfYKRzVg.gif" /></figure><p><strong>Key Findings:</strong></p><p>Below is a scatterplot of NO₂ concentrations vs population density, coloured by land cover class. Log scales have been used for clearer interpretation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5CjLx9NKnoq3zZSsEFlpsA.png" /></figure><ul><li><em>Urban zones show clustering of high NO₂ and high population.</em></li><li><em>Land cover plays a major role, croplands and savannas generally show lower NO₂.</em></li><li><em>Correlation between NO₂ and population exists, but is stronger in urbanised land types.</em></li></ul><p>When you look at the scatterplot above, NO₂ concentrations versus population density, coloured by land cover, you might notice something odd:</p><p>Population density is in the thousands, while NO₂ concentrations look tiny, almost like errors, just 0.00001 or so. But don’t worry, it’s not a mistake. It’s just science being subtle.</p><h4>On the X-axis: People per Square Kilometre</h4><p>Population density is pretty intuitive. In urban Abuja, we’re talking hundreds to tens of thousands of people per km². In log scale, this shows up as values between:</p><ul><li>1⁰³ (1,000 people/km²) → suburban areas.</li><li>1⁰⁵ (100,000 people/km²) → dense neighbourhoods.</li></ul><h4>On the Y-axis: Trace Gases in the Air</h4><p>Now NO₂, that’s a bit different.</p><p>I’ve used satellite data (from Sentinel-5P) that measures <strong>NO₂ in moles per square meter</strong> of atmosphere, not at ground level, but across the full atmospheric column overhead.</p><p>Since NO₂ is a <strong>trace gas</strong>, even in polluted cities, we’re only seeing:</p><ul><li>Around <strong>0.00001 to 0.0001 mol/m²</strong></li><li>That’s <strong>10⁻⁵ to 10⁻⁴ mol/m²</strong></li><li>Hence, log₁₀(NO₂) falls between <strong>−5 and −3</strong></li></ul><p>This isn’t an error. It’s just the real chemistry of a clean sky mixed with a busy city.</p><h4><strong>Limitations:</strong></h4><p>The Population data (GPWv4) is static, based on estimates from <strong>2020</strong>. MODIS land cover is also <strong>from 2020</strong>, so newer land-use changes may not be reflected. Cloud contamination and orbital gaps may affect some NO₂ readings.</p><h4>Final Thoughts</h4><p>This short remote sensing experiment highlights how easy it is to blend Earth Observation and Python to study urban health, especially air quality. As cities grow, tools like this can help us identify environmental risks in real time.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f8b33613ad13" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Over the past few weeks, I’ve observed growing concern in Nigeria’s tech and fintech community over…]]></title>
            <link>https://medium.com/@modeleke61/over-the-past-few-weeks-ive-observed-growing-concern-in-nigeria-s-tech-and-fintech-community-over-b5f291a55465?source=rss-b67a1b0e065f------2</link>
            <guid isPermaLink="false">https://medium.com/p/b5f291a55465</guid>
            <category><![CDATA[regulation]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[fintech]]></category>
            <category><![CDATA[bitcoin-news]]></category>
            <category><![CDATA[bitcoin]]></category>
            <dc:creator><![CDATA[Bosun Oyeleke]]></dc:creator>
            <pubDate>Tue, 27 May 2025 12:36:01 GMT</pubDate>
            <atom:updated>2025-06-22T03:10:56.611Z</atom:updated>
            <content:encoded><![CDATA[<p>Over the past few weeks, I’ve observed growing concern in Nigeria’s tech and fintech community over a possible shift in how the Securities and Exchange Commission (SEC) Nigeria classifies Bitcoin. Some suggest that Bitcoin may soon be regulated as a security under the Investment and Securities Act (ISA).</p><p>As a citizen, technologist, and Bitcoin advocate, I want to clearly state:</p><p>Bitcoin is not a security. It is a digital asset and should be treated as such.</p><p>What Is Bitcoin, Really?</p><p>Bitcoin is not a company. It’s not an investment contract. There is no CEO, no office, and no promises of profit from any central party.</p><p>Bitcoin is:</p><ul><li>A decentralised digital asset with no issuer.</li><li>A store of value, like digital gold.</li><li>A medium of exchange, used by millions for cross-border payments, remittances, and inflation-resistant savings.</li></ul><p>Treating it as a security is like calling cash or gold bars a stock certificate. It doesn’t fit legally or economically.</p><p>What the World Is Doing</p><p>Most global regulators agree:</p><ul><li>United States: The Commodity Futures Trading Commission (CFTC) treats Bitcoin as a commodity, not a security.</li><li>Germany: BaFin classifies Bitcoin as private money or property.</li><li>Japan: The Financial Services Agency recognises it as a means of payment.</li><li>FATF &amp; IMF: Define Bitcoin as a virtual asset, not an investment security.</li></ul><p>Even in Africa, countries like South Africa and Kenya are exploring payment-focused frameworks, not security-based ones, for Bitcoin.</p><p>What the Nigerian Law Says?</p><p>Nigeria’s Investment and Securities Act (ISA) defines a security as:</p><p>“Any share, debenture, bond, or instrument issued or to be issued by a body corporate…”</p><p>Bitcoin fails this test:</p><ul><li>It was not issued by any body corporate.</li><li>There is no central entity whose success drives its value.</li><li>It does not represent equity, debt, or profit-sharing.</li></ul><p>Even the SEC’s 2020 Statement on Digital Assets said cryptocurrencies like Bitcoin would only be treated as securities if they were part of a collective investment scheme. Regular peer-to-peer BTC use for remittances or payments, is far from that.</p><p>Why It Matters for Nigeria</p><p>If we label Bitcoin a security:</p><ul><li>Fintech apps and wallets using BTC may face unnecessary licensing burdens.</li><li>Innovation in remittances and cross-border commerce could be stifled.</li><li>Everyday Nigerians who use Bitcoin to hedge against inflation could be locked out or criminalised.</li></ul><p>At a time when Nigeria leads Africa in crypto adoption, this would be a major setback.</p><p>My Appeal to the SEC</p><p>I urge the Securities and Exchange Commission Nigeria to:</p><ol><li>Publicly clarify that Bitcoin is not a security.</li><li>Reaffirm the 2020 position distinguishing digital assets from tokenised securities. And I would quote that document here, <strong><em>“The position of the Commission is that virtual crypto assets are securities, unless proven otherwise. Thus, the burden of proving that the crypto assets proposed to be offered are not securities and therefore not under the jurisdiction of the SEC, is placed on the issuer or sponsor of the said assets.” </em></strong>and in this case, Bitcoin.</li><li>Convene a stakeholder roundtable with fintech startups, legal experts, and civil society to co-develop appropriate digital asset guidelines.</li></ol><p>We can regulate responsibly without stifling freedom, innovation, or economic inclusion.</p><p>Let’s Build, Not Ban</p><p>Bitcoin is not a security. It’s an open protocol for economic freedom. Let’s treat it with the nuance and clarity it deserves.</p><p>If you’re a policymaker, developer, entrepreneur, or citizen who believes in responsible innovation, let’s talk. Let’s collaborate. Let’s keep Nigeria leading, not lagging.</p><ol><li>Nigerian Gov Passes Law Recognising Bitcoin As A Security: <a href="https://www.forbes.com/sites/digital-assets/2025/04/30/nigerian-gov-passes-law-recognizing-bitcoin-as-a-security/">https://www.forbes.com/sites/digital-assets/2025/04/30/nigerian-gov-passes-law-recognizing-bitcoin-as-a-security/</a></li><li>Reaffirm the 2020 position distinguishing digital assets from tokenised securities.: <a href="https://sec.gov.ng/statement-on-digital-assets-and-their-classification-and-treatment/">https://sec.gov.ng/statement-on-digital-assets-and-their-classification-and-treatment/</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b5f291a55465" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>