<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/">
    <channel>
        <title>mhaas.eth</title>
        <link>https://paragraph.com/@mhaas</link>
        <description>undefined</description>
        <lastBuildDate>Sat, 04 Apr 2026 19:19:06 GMT</lastBuildDate>
        <docs>https://validator.w3.org/feed/docs/rss2.html</docs>
        <generator>https://github.com/jpmonette/feed</generator>
        <language>en</language>
        <copyright>All rights reserved</copyright>
        <item>
            <title><![CDATA[The system-wallet protocol]]></title>
            <link>https://paragraph.com/@mhaas/the-system-wallet-protocol</link>
            <guid>L06Gpq4cZ8hTbZOBB9lf</guid>
            <pubDate>Wed, 12 Jul 2023 08:09:33 GMT</pubDate>
            <description><![CDATA[Hi!IntroAs many of you probably know, I am the lead dev at ethOS, a new crypto native mobile OS. I created the system-level wallet protocol which I will explain here.The current problem with mobile3As of right now, many mobile dapps use WalletConnect to let users connect their wallet. I believe that WalletConnect is a nice protocol, if you have your wallet and dapp on seperate devices, but it’s not the best for mobile dapps that want to connect to a local wallet. Reason being, the WalletConne...]]></description>
            <content:encoded><![CDATA[<p>Hi!</p><h2 id="h-intro" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Intro</h2><p>As many of you probably know, I am the lead dev at ethOS, a new crypto native mobile OS. I created the system-level wallet protocol which I will explain here.</p><h2 id="h-the-current-problem-with-mobile3" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">The current problem with mobile3</h2><p>As of right now, many mobile dapps use WalletConnect to let users connect their wallet. I believe that WalletConnect is a nice protocol, if you have your wallet and dapp on seperate devices, but it’s not the best for mobile dapps that want to connect to a local wallet. Reason being, the WalletConnect protocol makes network requests and establishes a connection between the wallet and the dapp (I am over-simplifying). Now that’s fine for two seperate devices, but on mobile it should be faster just by communicating locally with the wallet. Coinbase Wallet uses intent-based communications afaik, which is better because it is local, but it is still wonky and fails sometimes. Truth is, current mobile OS just aren’t built with wallets in mind. That’s the reason we built the system-wallet protocol.</p><h2 id="h-the-system-wallet-protocol" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">The system-wallet protocol</h2><p>On Android there are services. Any app can deploy one to do updates in the background or other tasks. These services can get killed by the OS for many reasons, for example high cpu usage. But there is another type of service in the Android world, and that is a system service. The system service is a service dispatched by the OS itself, and it is always on and reachable by design. Any app can connect to a system-service and execute certain methods (As long as the policy allows it). The system-wallet is 2 system-services. One public, and one private system service. These two system services are connected via the SystemUI, where the UI is shown and the user can accept or decline the request.</p><h2 id="h-why-its-better" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Why it’s better</h2><p>No network requests means much faster execution of requests and much better UX. All the communication happens on the device which increases privacy over other protocols.</p><p>And now here is the best part, the system-wallet protocol is setup so any other OEM developer can adopt it. It is not ethOS specific, and any Android flavor can add the system-level wallet protocol if they want to. Even base-Android can add it if they want to. Everything is open-source, and it will always be this way. Everyone can inspect the code and add improvements, or add new wallet-methods to the system-level protocol.</p>]]></content:encoded>
            <author>mhaas@newsletter.paragraph.com (mhaas.eth)</author>
        </item>
        <item>
            <title><![CDATA[Self-sovereign computing]]></title>
            <link>https://paragraph.com/@mhaas/self-sovereign-computing</link>
            <guid>wJQPMfSsz3BCzebJgSx3</guid>
            <pubDate>Fri, 09 Jun 2023 05:33:21 GMT</pubDate>
            <description><![CDATA[The 2010s have been plagued by a curse called the Server-Client architecture. In theory it’s very simple. You have a client that interacts with a server to achieve some kind of result. I believe that in the future there will be much more self-sovereign computing, instead of delegated computing, and I will start by defining self-sovereign computing. I define the term “Self-sovereign computing” as the act of not having to be dependant on a server, but only dependant on the willingness to execut...]]></description>
            <content:encoded><![CDATA[<p>The 2010s have been plagued by a curse called the Server-Client architecture. In theory it’s very simple. You have a client that interacts with a server to achieve some kind of result.</p><p>I believe that in the future there will be much more self-sovereign computing, instead of delegated computing, and I will start by defining self-sovereign computing.</p><p>I define the term “Self-sovereign computing” as the act of not having to be dependant on a server, but only dependant on the willingness to execute something.</p><p>For example:</p><ul><li><p>Instead of being dependant on the Banking system to server you, you can run your own Ethereum node and compute self-sovereingly.</p></li><li><p>Instead of being dependant on Facebook to serve you, you can run your own Farcaster hub and compute self-sovereignly.</p></li><li><p>Instead of being dependant on Google to serve you, you can run your own Llama model and compute self-sovereingly.</p></li></ul><p>Of course the Llama model will not 100% give you factually results all the time, but this could change in the future with advances in the field, and it does get the birthdays of famous US presidents correct.</p><p>I hope for a future where most of Internet traffic will be only bits that *actually* cannot be computed on your own. Everything else should be computed on your own computing device, either completely local or if its connected then through a P2P network of nodes. I want ethOS to have a node for messaging (Waku), one for money/smart contracts (Ethereum) and all other things that people use their phone for. And if people ask why they would want that, the answer is that its the best thing they will ever use.</p>]]></content:encoded>
            <author>mhaas@newsletter.paragraph.com (mhaas.eth)</author>
            <enclosure url="https://storage.googleapis.com/papyrus_images/3d5d3a016946b10a7789ad857684ccf219bff3a2371847c9c62afce541bb120d.jpg" length="0" type="image/jpg"/>
        </item>
    </channel>
</rss>