<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
<title>Luca's blog</title>
<link>https://structwafel.dev/</link>
<description>Another fun blog</description>
<atom:link href="https://structwafel.dev/rss.xml" rel="self" type="application/rss+xml"/>
<lastBuildDate>Fri, 29 Aug 2025 00:00:00 +0000</lastBuildDate>
<item>
<title>Yoga Pro 7 14ASP9 (14&quot; AMD)</title>
<link>https://structwafel.dev/reviews/yoga-pro-7/</link>
<guid>https://structwafel.dev/reviews/yoga-pro-7/</guid>
<pubDate>Fri, 29 Aug 2025 00:00:00 +0000</pubDate>
<category>review</category>
<category>laptop</category>
<content:encoded xmlns:content="http://purl.org/rss/1.0/modules/content/"><![CDATA[<p>Actual reviews:</p>
<ul>
<li><a href="https://www.ultrabookreview.com/69904-lenovo-yoga-pro7-review/">ultrabookreview</a></li>
<li><a href="https://www.youtube.com/watch?v=8750qjtaXeo">Just Josh</a></li>
</ul>
<h3 id="context">Context<a href="#context" aria-label="Link to heading 'Context'" data-heading-content="Context" class="anchor"></a></h3>
<p>Using the laptop to program, browse, watch shows.</p>
<h3 id="setup">Setup<a href="#setup" aria-label="Link to heading 'Setup'" data-heading-content="Setup" class="anchor"></a></h3>
<p>Arch + sway. <a href="https://github.com/structwafel/stuff">Config</a></p>
<h2 id="the-good">The good<a href="#the-good" aria-label="Link to heading 'The good'" data-heading-content="The good" class="anchor"></a></h2>
<ul>
<li>Screen is great
<ul>
<li>Oled is fantastic, resolution is great, little bit brighter would be nice</li>
</ul>
</li>
<li>Performance is great</li>
</ul>
<h2 id="the-bad">The bad<a href="#the-bad" aria-label="Link to heading 'The bad'" data-heading-content="The bad" class="anchor"></a></h2>
<ul>
<li>Absolute abyssmal power button
<ul>
<li>I don't know who decided it, but they need to touch grass and use a laptop more than on a desk. I have probably pressed the power button by accident more often than with intent.</li>
</ul>
</li>
<li>Battery life is fine
<ul>
<li>If I would need to go through 7-8 hours I will bring the charger with me. But if i know only 4-5 hours I won't stress about the battery.</li>
</ul>
</li>
</ul>
<h2 id="conclusion">Conclusion<a href="#conclusion" aria-label="Link to heading 'Conclusion'" data-heading-content="Conclusion" class="anchor"></a></h2>
<p>Picking a non mac laptop is a pain. Think it went fine with this one.</p>
<h2 id="future">Future<a href="#future" aria-label="Link to heading 'Future'" data-heading-content="Future" class="anchor"></a></h2>
<p>The only improvements would be brighter display and better battery life. Both which have been the must improve of all non mac laptops the past 15 years.</p>
]]></content:encoded>
</item>
<item>
<title>AI: The SaaS killer</title>
<link>https://structwafel.dev/articles/ai-the-saas-killer/</link>
<guid>https://structwafel.dev/articles/ai-the-saas-killer/</guid>
<pubDate>Fri, 27 Jun 2025 00:00:00 +0000</pubDate>
<category>ai</category>
<content:encoded xmlns:content="http://purl.org/rss/1.0/modules/content/"><![CDATA[<blockquote>
<p>This article will be about the idea that things like SaaS will completely dissapear in the next 5-20 years.</p>
</blockquote>
<p>AI CEO's keep telling me that either the programming world will dissapear, or that they are still needed but much more critical/specfic jobs. I probably agree with this someway.</p>
<p>However I haven't seen anyone yet mention the dangers which each SaaS is facing, if AI is getting so great at coding. Why would I pay a whole service for something instead of letting an AI create a service for me?</p>
<p>Why pay for managed hosting services, when AI can make a program (using opensource software to speed up the creation) to do it for me.</p>
<p>Why pay for authirization services, when you can have a reliable service for yourself?</p>
<p>Why pay for all these things, when you can have a little farm plus a week to create your own. With the features you need.</p>
<h2 id="what-if">What if<a href="#what-if" aria-label="Link to heading 'What if'" data-heading-content="What if" class="anchor"></a></h2>
<p>What if not tho, perhaps we don't reach a point in next 20 years. Or we find a way to make software even more convoluted.
Perhaps adding AI slob to everything is not the answer (why do people think it is). Perhaps we will slowly simplify our lives. Probably not.</p>
<p>Yet I can still not use AI's today for things you would actually need? (like what?)</p>
]]></content:encoded>
</item>
<item>
<title>Homerow is bad</title>
<link>https://structwafel.dev/articles/homerow-is-overrated/</link>
<guid>https://structwafel.dev/articles/homerow-is-overrated/</guid>
<pubDate>Sun, 04 May 2025 00:00:00 +0000</pubDate>
<category>opinion</category>
<category>desktop</category>
<content:encoded xmlns:content="http://purl.org/rss/1.0/modules/content/"><![CDATA[<blockquote>
<p>This will be short.</p>
</blockquote>
<p>When getting my <a href="https://www.moergo.com/">glove80</a>, I started with stock qwerty.
This however was not something desirable, as the separate layer heights made it such that typing letters which were not meant for that finger means crossing a height barrier.</p>
<p>I was quickly looking for alternatives.</p>
<p>One such alternative was <a href="https://sunaku.github.io/moergo-glove80-keyboard.html#layers">sunaku's moergo layout</a>, which at the time was engram layout.
And had multiple layers for different things, symbols, numbers, function, mouse etc.</p>
<p>sunaku's layout was great, engram is an extremely ergonomic layout. It makes a lot of sense to me.
And the homerow mods, woaw that was an eye-opener. Perfect combination of everything.</p>
<p>After some weeks when I could type fine on engram (40wpm?) something started to bother me. The homerow mods are made me somewhat slow?</p>
<p>The whole idea of homerow is that you can press a homerow mod for <code>Super, Alt, Ctrl, Shift</code> when you press it and then something else afterwards it does the combination you want.</p>
<p>But the problem is timing, the delay you need to press for homerow to function ,let's say 40ms, makes it such that you need to press the homerow key for atleast 40ms before you can press something else.
This makes you do a pause when typing.</p>
<p>After years of typing you probably press things like: <code>Ctrl+c</code> super quick, to the point where you basically press them at the same time.
You can of course do this aswell with homerow mods, just lower the ms needed to press down the key. But then you encounter issues where you always need to type as fast as you can.
As soon as you start relaxing, the homerow won't engage.</p>
<h2 id="fin">fin<a href="#fin" aria-label="Link to heading 'fin'" data-heading-content="fin" class="anchor"></a></h2>
<p>Perhaps I couldn't configure it correctly, but to me any delay is weird when typing. I would prefer for when i press something for it to work immediately.
Having the <code>Super, Alt, Ctrl, Shift</code> to keys down of homerow is perfectly fine on my glove80. Using 4 rows of keys is fine and doesn't require much more effort than only using 3 rows.</p>
<p>If you want to use homerow please do, and tell me how and if you got a setup working where you can type as fast as you want. And still have perfect modifiers actions.</p>
]]></content:encoded>
</item>
<item>
<title>AI: The new mediocre</title>
<link>https://structwafel.dev/articles/ai-the-new-mediocre/</link>
<guid>https://structwafel.dev/articles/ai-the-new-mediocre/</guid>
<pubDate>Wed, 30 Apr 2025 00:00:00 +0000</pubDate>
<category>ai</category>
<content:encoded xmlns:content="http://purl.org/rss/1.0/modules/content/"><![CDATA[<h3 id="things-to-discuss">things to discuss<a href="#things-to-discuss" aria-label="Link to heading 'things to discuss'" data-heading-content="things to discuss" class="anchor"></a></h3>
<p>two possibilites:</p>
<ul>
<li>
<p>things become more expected</p>
<p>now that everyone can do much more, we want better cleaner faster solutions.</p>
</li>
<li>
<p>things become cheaper and shittier</p>
<p>now that everyone can do much more, we don't want to put effort into going to the 100%.</p>
<p>Instead of getting 95% now with effort, we go to 80% with 0 effort which is &quot;good enough&quot;</p>
<p>think of lower quality of products since industrial age</p>
</li>
</ul>
<h2 id=""><a href="#" aria-label="Link to heading ''" data-heading-content="" class="anchor"></a></h2>
<p>I think the effort/productivity curve just shifted. And this doesn't neccesarily mean we will have better products.</p>
<p>As corporations (people) get greedier, I'm not so certain the promised AI-land will make things actually better for us.</p>
<h2 id="unfiltered-thoughts">Unfiltered Thoughts<a href="#unfiltered-thoughts" aria-label="Link to heading 'Unfiltered Thoughts'" data-heading-content="Unfiltered Thoughts" class="anchor"></a></h2>
<ul>
<li>If AI is so good now, why are we still writing js in the backend?</li>
<li>If AI is so good now, why are you still paying for vercel etc?</li>
</ul>
]]></content:encoded>
</item>
<item>
<title>Go Needs Immutable Pointers</title>
<link>https://structwafel.dev/articles/go-needs-immutability/</link>
<guid>https://structwafel.dev/articles/go-needs-immutability/</guid>
<pubDate>Fri, 25 Apr 2025 00:00:00 +0000</pubDate>
<category>go</category>
<category>programming</category>
<content:encoded xmlns:content="http://purl.org/rss/1.0/modules/content/"><![CDATA[<blockquote>
<p>To modify, or not to modify, that is the question</p>
</blockquote>
<blockquote>
<p>This will be more of a rant, and if motivation permits a go utility to have immutable pointers in go</p>
</blockquote>
]]></content:encoded>
</item>
</channel>
</rss>
