<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Alex</title>
    <description>The latest articles on DEV Community by Alex (@justonedev).</description>
    <link>https://dev.to/justonedev</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F2510020%2F19bdff13-e9a5-45d6-b3a0-b4f5e4991e20.jpg</url>
      <title>DEV Community: Alex</title>
      <link>https://dev.to/justonedev</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/justonedev"/>
    <language>en</language>
    <item>
      <title>How to create a cool looking explainer video for anything in 5 minutes</title>
      <dc:creator>Alex</dc:creator>
      <pubDate>Sun, 08 Nov 2026 11:36:11 +0000</pubDate>
      <link>https://dev.to/justonedev/how-to-create-a-cool-looking-explainer-video-for-anything-in-5-minutes-519p</link>
      <guid>https://dev.to/justonedev/how-to-create-a-cool-looking-explainer-video-for-anything-in-5-minutes-519p</guid>
      <description>&lt;p&gt;I needed a 52-second explainer for Pageree, the product I'm working on. It has no screen recording and was never opened in a video editor. It's a web page: HTML and CSS for the scenes, a tiny timeline script, headless Chrome to render the frames, and ffmpeg to put picture and sound together. An AI voice reads it, and the music and sound effects are generated too.&lt;/p&gt;

&lt;p&gt;  &lt;iframe src="https://www.youtube.com/embed/cH3LgAN7mSY" width="710" height="399"&gt;
  &lt;/iframe&gt;
&lt;/p&gt;

&lt;p&gt;You can make one like it for anything: your app, a feature launch, an internal process, a course. Your part takes about five minutes: fill in a few blanks in the prompt below, approve the storyboard, and look at the stills. Your AI coding assistant writes the code, and the render takes a couple of minutes on a laptop.&lt;/p&gt;

&lt;h2&gt;
  
  
  What you need
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;An AI coding assistant that can run commands on your machine: Claude Code, Cursor, Codex or similar.&lt;/li&gt;
&lt;li&gt;Node 22 or newer, Google Chrome, and ffmpeg (&lt;code&gt;brew install ffmpeg&lt;/code&gt; on a Mac).&lt;/li&gt;
&lt;li&gt;An ElevenLabs API key for the voice, sound effects and music. Without one you still get the video, just silent. Commercial use needs a paid ElevenLabs plan.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Step 1: paste this prompt
&lt;/h2&gt;

&lt;p&gt;Fill in the five things in [brackets], then give the whole prompt to your assistant in an empty folder.&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;Build me a short animated explainer video, rendered from a web page.

THE VIDEO
- Topic: [what it explains, e.g. "how our app sends invoice reminders"]
- Audience: [who watches it]
- Length: about [45] seconds, 16:9, 1920x1080, 30 fps
- Look: flat cards with thick dark borders and hard offset shadows, one accent colour [#693bda], big bold headlines, plenty of empty space. No stock photos, no people.
- End card: [product name] and [URL]

THE STORY
Write 5 to 7 scenes, each 4 to 9 seconds with one idea: a hook that names the viewer's problem, three or four scenes that show how it works, and the end card. For each scene give an on-screen headline (8 words at most) and one voiceover line (about 2 words a second at most). Show me the storyboard as a table and wait for my OK before building.

HOW TO BUILD IT
1. Make a folder video/ with:
   - video.json: the scenes (id, minimum seconds, voice line, when the line starts inside the scene), the voice, the sound effects (name, prompt, seconds), the cues (sound, scene, second) and a music prompt.
   - index.html: every scene in HTML and CSS on a 1600x900 CSS-pixel canvas, absolutely positioned. No CSS animations or transitions.
   - timeline.js: a tiny engine. at(selector, start, duration, {prop: [from, to]}, easing) adds a track; seek(t) sets every element to its state at second t: opacity, translate x and y, scale, rotate, width in %, and typed text taken from data-text. Before its first track starts, an element sits at that track's "from" value. Times are scene-relative: t(scene, offset) and end(scene, offset) read each scene's start and end from timing.js. index.html?t=12.5 shows that moment and index.html?play plays it.
   - make.mjs: plain Node 22, no dependencies, three commands:
     a) audio: if ELEVENLABS_API_KEY is set, generate one MP3 per voice line (POST https://api.elevenlabs.io/v1/text-to-speech/{voice_id} with model_id eleven_multilingual_v2, sending the neighbouring lines as previous_text and next_text), one per sound effect (POST /v1/sound-generation with duration_seconds) and an instrumental music bed (POST /v1/music with music_length_ms and force_instrumental true). Name each file after a hash of what made it, so unchanged lines are never generated again. Measure where each line's speech ends with ffmpeg silencedetect, ignoring trailing silence. A scene lasts max(its minimum, voice start + speech length + 0.5 s). Write the scene times to timing.js.
     b) stills 2,8,14: PNG frames at those seconds, so we can both check the layout.
     c) render: start Chrome headless with --remote-debugging-port=0, read the port from DevToolsActivePort in its profile folder, and drive it over the DevTools protocol with Node's built-in WebSocket. Set the viewport to 1600x900 at deviceScaleFactor 1.2. For every frame call seek(frame / 30), wait two requestAnimationFrames, and save a JPEG screenshot. Then one ffmpeg pass: the frames at 30 fps; each voice line and sound effect placed with adelay; amix with normalize=0; the music under everything, ducked with sidechaincompress whenever the voice speaks and faded out at the end; loudnorm to -14 LUFS; H.264 and AAC with faststart. Also write an .srt caption file from the voice lines and their measured times.
2. Animate like a motion designer. Something moves every half second to a second. Entrances take 0.3 to 0.6 s, staggers 0.12 s, pops have a small overshoot. Each scene holds its finished state for a second before it leaves. The voice may lead the picture by up to half a second, never trail it. A soft pop when a card lands, a whoosh on scene changes, typing sounds while text types.
3. Before rendering, make stills of every scene, look at them yourself, and fix overlaps, cut-off text and anything touching the edges. Then render and give me the path to the MP4.

If there is no ElevenLabs key, render without sound and tell me.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;h2&gt;
  
  
  Step 2: check it before it renders
&lt;/h2&gt;

&lt;p&gt;The assistant shows you a storyboard first. Changing the story costs nothing at this point, so read it. After it builds the page, open &lt;code&gt;index.html?play&lt;/code&gt; in Chrome to watch it silently, or &lt;code&gt;?t=12.5&lt;/code&gt; to look at one moment. Ask for changes in plain words: "the second scene is too busy", "make the hook faster", "use a warmer voice".&lt;/p&gt;

&lt;h2&gt;
  
  
  Step 3: render
&lt;/h2&gt;

&lt;p&gt;The assistant runs &lt;code&gt;node make.mjs audio&lt;/code&gt;, then &lt;code&gt;node make.mjs render&lt;/code&gt;, and you get an MP4 and an .srt for YouTube. If a voice line changes, only that line is generated again, and its scene stretches to fit.&lt;/p&gt;

&lt;h2&gt;
  
  
  How it works, briefly
&lt;/h2&gt;

&lt;p&gt;&lt;strong&gt;A timeline you can seek.&lt;/strong&gt; CSS animations run in real time. A renderer needs to ask what second 12.4 looks like, so every animation is data, and &lt;code&gt;seek(t)&lt;/code&gt; computes each element's state:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight javascript"&gt;&lt;code&gt;&lt;span class="kd"&gt;function&lt;/span&gt; &lt;span class="nf"&gt;seek&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;t&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="nx"&gt;state&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="k"&gt;new&lt;/span&gt; &lt;span class="nc"&gt;Map&lt;/span&gt;&lt;span class="p"&gt;();&lt;/span&gt;
  &lt;span class="k"&gt;for &lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="nx"&gt;tr&lt;/span&gt; &lt;span class="k"&gt;of&lt;/span&gt; &lt;span class="nx"&gt;tracks&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="c1"&gt;// sorted by start&lt;/span&gt;
    &lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="nx"&gt;s&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="nx"&gt;state&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;get&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;tr&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;el&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="o"&gt;??&lt;/span&gt; &lt;span class="p"&gt;{};&lt;/span&gt;
    &lt;span class="nx"&gt;state&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;set&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;tr&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;el&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nx"&gt;s&lt;/span&gt;&lt;span class="p"&gt;);&lt;/span&gt;
    &lt;span class="k"&gt;for &lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="nx"&gt;p&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="k"&gt;from&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nx"&gt;to&lt;/span&gt;&lt;span class="p"&gt;]]&lt;/span&gt; &lt;span class="k"&gt;of&lt;/span&gt; &lt;span class="nb"&gt;Object&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;entries&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;tr&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;props&lt;/span&gt;&lt;span class="p"&gt;))&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
      &lt;span class="k"&gt;if &lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;t&lt;/span&gt; &lt;span class="o"&gt;&amp;lt;&lt;/span&gt; &lt;span class="nx"&gt;tr&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;start&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="k"&gt;if &lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="o"&gt;!&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;p&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="nx"&gt;s&lt;/span&gt;&lt;span class="p"&gt;))&lt;/span&gt; &lt;span class="nx"&gt;s&lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="nx"&gt;p&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="k"&gt;from&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt; &lt;span class="k"&gt;continue&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
      &lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="nx"&gt;k&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="nx"&gt;tr&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;dur&lt;/span&gt; &lt;span class="p"&gt;?&lt;/span&gt; &lt;span class="nb"&gt;Math&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;min&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="mi"&gt;1&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;t&lt;/span&gt; &lt;span class="o"&gt;-&lt;/span&gt; &lt;span class="nx"&gt;tr&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;start&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="o"&gt;/&lt;/span&gt; &lt;span class="nx"&gt;tr&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;dur&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;1&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
      &lt;span class="nx"&gt;s&lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="nx"&gt;p&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="k"&gt;from&lt;/span&gt; &lt;span class="o"&gt;+&lt;/span&gt; &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;to&lt;/span&gt; &lt;span class="o"&gt;-&lt;/span&gt; &lt;span class="k"&gt;from&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="o"&gt;*&lt;/span&gt; &lt;span class="nx"&gt;tr&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;ease&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;k&lt;/span&gt;&lt;span class="p"&gt;);&lt;/span&gt;
    &lt;span class="p"&gt;}&lt;/span&gt;
  &lt;span class="p"&gt;}&lt;/span&gt;
  &lt;span class="k"&gt;for &lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="nx"&gt;el&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nx"&gt;s&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="k"&gt;of&lt;/span&gt; &lt;span class="nx"&gt;state&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="k"&gt;if &lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="s1"&gt;o&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="nx"&gt;s&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="nx"&gt;el&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;style&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;opacity&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="nx"&gt;s&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;o&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
    &lt;span class="nx"&gt;el&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;style&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;transform&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="s2"&gt;`translate(&lt;/span&gt;&lt;span class="p"&gt;${&lt;/span&gt;&lt;span class="nx"&gt;s&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;x&lt;/span&gt; &lt;span class="o"&gt;??&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="s2"&gt;px, &lt;/span&gt;&lt;span class="p"&gt;${&lt;/span&gt;&lt;span class="nx"&gt;s&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;y&lt;/span&gt; &lt;span class="o"&gt;??&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="s2"&gt;px) scale(&lt;/span&gt;&lt;span class="p"&gt;${&lt;/span&gt;&lt;span class="nx"&gt;s&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;s&lt;/span&gt; &lt;span class="o"&gt;??&lt;/span&gt; &lt;span class="mi"&gt;1&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="s2"&gt;)`&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
  &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;strong&gt;The voice sets the timing.&lt;/strong&gt; My first version had hard-coded times, and the voiceover came back longer than my guesses. With scene-relative times and scene lengths measured from the speech, a longer line stretches its scene instead of being cut off. Generated speech ends with 0.2 to 0.6 seconds of silence, so measure where the speech really stops (&lt;code&gt;ffmpeg -i line.mp3 -af silencedetect=noise=-45dB:d=0.1 -f null -&lt;/code&gt;).&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Frames from Chrome, sound from ffmpeg.&lt;/strong&gt; Chrome renders 1,557 frames for 52 seconds. One ffmpeg filter graph places every line and sound, ducks the music under the voice and sets the loudness. Watch for &lt;code&gt;amix&lt;/code&gt;: without &lt;code&gt;normalize=0&lt;/code&gt; it turns every input down and your voice comes out quiet.&lt;/p&gt;

&lt;h2&gt;
  
  
  What I'd tell myself before starting
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;The silent version looked finished. It wasn't. The voice, the music and a few pops and whooshes did more than any visual change.&lt;/li&gt;
&lt;li&gt;Make it scrubbable. &lt;code&gt;?t=&lt;/code&gt; was the most useful thing in the whole setup.&lt;/li&gt;
&lt;li&gt;Let the voice decide the pace, not your guesses.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;If you make a video with this prompt, I'd love to see it in the comments.&lt;/p&gt;

&lt;p&gt;This post was drafted with help from an AI assistant.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>webdev</category>
      <category>tutorial</category>
      <category>productivity</category>
    </item>
    <item>
      <title>Landing page builders don't make sense anymore</title>
      <dc:creator>Alex</dc:creator>
      <pubDate>Mon, 02 Nov 2026 13:45:16 +0000</pubDate>
      <link>https://dev.to/justonedev/landing-page-builders-dont-make-sense-anymore-139f</link>
      <guid>https://dev.to/justonedev/landing-page-builders-dont-make-sense-anymore-139f</guid>
      <description>&lt;p&gt;I build a product in this space, so read this with that in mind. I'll get to it at the end.&lt;/p&gt;

&lt;p&gt;For about fifteen years a landing page builder was the sensible way for a non-developer to put a page on the internet. You picked a template, dragged blocks around, typed your copy into boxes, connected the form to Mailchimp and pressed publish. The editor was the product. Hosting, forms and a bit of analytics came bundled, so you never had to think about them.&lt;/p&gt;

&lt;p&gt;The editor is the part that stopped making sense.&lt;/p&gt;

&lt;h2&gt;
  
  
  The page is the easy part now
&lt;/h2&gt;

&lt;p&gt;If you can describe what you sell, ChatGPT or Claude will write you a decent landing page in a minute: a headline, a few sections, a form, CSS that holds up on a phone. Ask for three versions and pick one. Say "shorter, less corporate, move pricing up" and it does that too. That is the job the drag-and-drop editor existed for, and a chat does it faster, with nothing to learn.&lt;/p&gt;

&lt;p&gt;What surprised me is that the assistants seem to agree.&lt;/p&gt;

&lt;p&gt;On 15 October we asked eight AI tools (the ChatGPT app, OpenAI's API with search, Claude, Perplexity, Google's AI Overviews and AI Mode, Grok and Mistral) 68 questions people ask when they need a landing page. Things like "What are the best AI landing page builders in 2026?", "How do I host an HTML file that ChatGPT made?" and "How do I add a working contact form to an HTML page ChatGPT wrote?". None of them named a product. We sell to people asking the second and third kind, so the set leans that way. Keep that in mind.&lt;/p&gt;

&lt;p&gt;Across about 1,560 answers, the two products named most often were not builders. They were hosts:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Vercel: 492 of 1,561 answers (31.5%)&lt;/li&gt;
&lt;li&gt;Netlify: 480 of 1,538 (31.2%)&lt;/li&gt;
&lt;li&gt;Framer: 360 of 1,560 (23.1%)&lt;/li&gt;
&lt;li&gt;Unbounce: 352 of 1,537 (22.9%)&lt;/li&gt;
&lt;li&gt;Wix: 331 of 1,560 (21.2%)&lt;/li&gt;
&lt;li&gt;Leadpages: 258 of 1,560 (16.5%)&lt;/li&gt;
&lt;li&gt;Carrd: 213 of 1,536 (13.9%)&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;It's one day's run, and these answers change a lot from one run to the next, so don't read anything into the decimals. The claim I'd make is a small one: when you ask an assistant how to get a page live, it no longer sends you to an editor. It assumes you have the HTML and tells you where to put it.&lt;/p&gt;

&lt;p&gt;(Our own product came up in 2 of 1,584 answers. Fair is fair.)&lt;/p&gt;

&lt;h2&gt;
  
  
  Where it actually breaks
&lt;/h2&gt;

&lt;p&gt;The trouble is everything after the HTML. I read a lot of posts from people trying to do exactly this, and the same three problems come up every day.&lt;/p&gt;

&lt;p&gt;Where does it go? The page is sitting in a chat window. Some assistants will share it, but the link lives inside their product and belongs to their account, not to you, and it's not something you'd print on a business card.&lt;/p&gt;

&lt;p&gt;The form goes nowhere. An assistant will happily write a form with a submit button. Nothing receives it. People find out when the first lead never shows up, or they never find out. "My contact form doesn't send anything" might be the sentence I read most often.&lt;/p&gt;

&lt;p&gt;Nobody knows what happened. A page you can't measure is a guess. Did anyone read past the headline? Which section do people leave at?&lt;/p&gt;

&lt;p&gt;One post last week summed it up. Someone asked Claude for "something to collect emails for launch" and got a full todo app instead of the one-page waitlist they needed. Writing that page would have taken the model seconds. What it couldn't know was that a waitlist means a page plus somewhere for the emails to go.&lt;/p&gt;

&lt;h2&gt;
  
  
  Where builders still win
&lt;/h2&gt;

&lt;p&gt;Builders aren't dead, and I'd be lying if I said they were. They still win when:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;several non-technical people edit the same pages every week and need a visual editor they all understand;&lt;/li&gt;
&lt;li&gt;an agency runs a brand system across dozens of clients, with approvals;&lt;/li&gt;
&lt;li&gt;you have enough traffic to A/B test properly, and the builder runs the test for you;&lt;/li&gt;
&lt;li&gt;you need a big catalog of integrations someone else maintains;&lt;/li&gt;
&lt;li&gt;you don't use an AI assistant at all, which is still most people.&lt;/li&gt;
&lt;li&gt;If that's you, keep your builder. This is about the growing group who already have the assistant open and pay for a builder only because it came with hosting and a form.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  What replaces it
&lt;/h2&gt;

&lt;p&gt;If the assistant writes the page, the rest of the stack gets small:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;somewhere to publish it under a real URL, with versions so you can undo a bad change;&lt;/li&gt;
&lt;li&gt;a form endpoint that delivers to an inbox or a CRM;&lt;/li&gt;
&lt;li&gt;measurement the same assistant can read, so it can say "people stop at the pricing section" and change one thing;&lt;/li&gt;
&lt;li&gt;a person approving what goes live.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;You can watch this shape form. OpenAI put hosting into ChatGPT with Sites. Leadpages, a builder company, launched HTML Pub, whose pitch is roughly "your AI builds it, we publish it". Convika describes itself as a git repo, a static host, a form service, an analytics script and DNS, replaced by one MCP server. A new YC company, Spacesheep, launched on Product Hunt this weekend doing publishing for agents. Everyone is heading the same way: the editor goes, and what's left is plumbing an assistant can drive.&lt;/p&gt;

&lt;p&gt;I'm building one of these, &lt;a href="https://pageree.com" rel="noopener noreferrer"&gt;Pageree&lt;/a&gt;. It's an MCP server, so ChatGPT, Claude, Cursor and others can publish a page, send its form to your email or CRM and read section-by-section visitor data back, and nothing goes live until you approve it. It isn't the only way to do this. If you're a developer, a static host plus a form service plus an analytics script, glued together by hand, gets you most of the way.&lt;/p&gt;

&lt;p&gt;The point I actually care about is narrower. If you're not a developer and you already write your page in a chat, paying for a visual editor you never open, just to get hosting and a form, doesn't make sense anymore.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>chatgpt</category>
      <category>saas</category>
      <category>startup</category>
    </item>
    <item>
      <title>Is Vibe Coding the Future of Development?</title>
      <dc:creator>Alex</dc:creator>
      <pubDate>Sat, 15 Mar 2025 07:11:56 +0000</pubDate>
      <link>https://dev.to/justonedev/is-vibe-coding-the-future-of-development-627</link>
      <guid>https://dev.to/justonedev/is-vibe-coding-the-future-of-development-627</guid>
      <description>&lt;p&gt;I hear a lot of opinions—both positive and negative—about vibe coding. After 20 years of coding hard, I want to share my take on it.&lt;br&gt;
Here are some of the most common concerns I’ve heard:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;❌ "If you use vibe coding, your app will be insecure, poorly written, unoptimized, and unmaintainable."&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;I strongly disagree. If you can’t analyze AI-generated code, use it effectively, and guide the AI in the right direction, the problem isn’t the tool—it’s you. Let’s be honest: in the right hands, AI writes better code than many developers.&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;❌ You’ll spend more time debugging AI-generated code than writing it yourself.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;Again, this is likely a matter of how you use it. Treat AI like a junior developer: give it small, well-defined tasks, break down complex problems, and provide guidance. You're the architect; AI is the tool. It can handle 80–90% of the work, but you still need to oversee complex logic and edge cases.&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;❌ Real developers don’t use AI; it makes you lazy and weakens your problem-solving skills.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;Using AI doesn’t mean you’re not a real developer. Writing every line of code manually is not a badge of honor—it’s just inefficient. Great developers know how to leverage the best tools available. AI can accelerate development, allowing you to focus on architecture, problem-solving, and innovation.&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;❌ AI can’t replace deep domain knowledge.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;True, but it doesn’t have to. AI is a productivity booster, not a replacement for expertise. It still requires a developer who understands the business logic, security concerns, performance trade-offs, and long-term maintainability.&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;❌ Vibe coding is just a fad.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;People once said the same thing about IDEs, version control, and even high-level programming languages. Vibe coding isn’t about replacing developers—it’s about making them more efficient. The tools will only get better, and developers who learn to use them effectively will stay ahead.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;🔥 So, is vibe coding the future?&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;I think it's already here. AI is rapidly changing the way we write software, and those who learn to work with it—not against it—will be the ones leading the future of development.&lt;br&gt;
What’s your take? Have you tried vibe coding yet? 🚀&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;🚀 Proof That Vibe Coding Works&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;Just look at &lt;a href="https://justcms.co" rel="noopener noreferrer"&gt;JustCMS&lt;/a&gt;—a fully distributed, AI-driven, headless CMS that’s 90% written by an AI agents under human guidance. It’s stable, scalable, and a real-world example of how AI can help developers build powerful software. The key isn’t replacing developers—it’s using AI to amplify what’s possible.&lt;/p&gt;

</description>
    </item>
    <item>
      <title>The 4-Day SaaS: Tools, Tips, and the Process Behind the Build</title>
      <dc:creator>Alex</dc:creator>
      <pubDate>Mon, 02 Dec 2024 17:46:40 +0000</pubDate>
      <link>https://dev.to/justonedev/the-4-day-saas-tools-tips-and-the-process-behind-the-build-35d4</link>
      <guid>https://dev.to/justonedev/the-4-day-saas-tools-tips-and-the-process-behind-the-build-35d4</guid>
      <description>&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F4prxvcg0be6eli6772b4.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F4prxvcg0be6eli6772b4.png" alt=" " width="800" height="533"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;In case you prefer to watch, I made a video version with &lt;a href="https://lazy2read.com" rel="noopener noreferrer"&gt;Lazy2Read&lt;/a&gt;. If not you can continue reading.&lt;/p&gt;

&lt;p&gt;  &lt;iframe src="https://www.youtube.com/embed/oHWf8OauDOc" width="710" height="399"&gt;
  &lt;/iframe&gt;
&lt;/p&gt;

&lt;p&gt;Hello everyone! My name is Alex, and I’m a technical lead with over 22 years of experience in web and mobile development. By day, I work on large-scale projects in my 9-to-5 job, but by night, I’m an indie hacker driven by a passion for building new apps and launching them into the world.&lt;/p&gt;

&lt;p&gt;Because of my full-time work, I’ve learned how to optimize every step of the development process, always looking for the quickest, smartest ways to turn an idea into a live, functioning product. In this article, I’ll share how I recently built and launched a SaaS product in just 4 days, along with the exact tools, strategies, and shortcuts that made it possible.&lt;/p&gt;

&lt;p&gt;If you’ve ever wondered how to create a product fast without sacrificing quality, you’re in the right place — let’s dive in!&lt;/p&gt;

&lt;h2&gt;
  
  
  The Stack is Important
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fn8gvju4cr53xjsf2hazk.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fn8gvju4cr53xjsf2hazk.png" alt="The Stack is Important" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Choosing the right stack is crucial, not just for scalability but also for development speed. The right stack simplifies complexity, speeds up the process, and works seamlessly with AI-generated code — saving you time spent debugging and fixing.&lt;/p&gt;

&lt;p&gt;For my projects, I use Nuxt.js on the frontend, NestJS on the backend, MongoDB as my database, and TypeScript everywhere. This combination strikes the perfect balance between speed, simplicity, and scalability. Let’s explore why I chose this stack.&lt;/p&gt;

&lt;h3&gt;
  
  
  TypeScript on Backend and Frontend
&lt;/h3&gt;

&lt;p&gt;TypeScript is an incredible choice for many reasons, but here’s why I use it across both the backend and frontend:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Code Reusability: With TypeScript on both sides, I can easily reuse code like validation schemas, types, and utility functions. This saves significant development time and reduces duplication.&lt;/li&gt;
&lt;li&gt;Reliability and Maintainability: TypeScript’s static typing helps catch errors early, making the codebase more reliable and easier to maintain — without adding extra effort.&lt;/li&gt;
&lt;li&gt;AI Compatibility: TypeScript (and JavaScript) are the most popular languages on GitHub, meaning AI tools have a deep understanding of them. This results in better AI-generated code and less time spent debugging.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;In short, TypeScript speeds up development while keeping the codebase clean, error-free, and future-proof.&lt;/p&gt;

&lt;h3&gt;
  
  
  Nuxt.js (Vue.js)
&lt;/h3&gt;

&lt;p&gt;Nuxt.js, built on Vue.js, strikes the perfect balance between simplicity and power. I still believe it’s one of the easiest frameworks to learn, which not only reduces the chances of human errors but also minimizes mistakes from AI-generated code.&lt;/p&gt;

&lt;p&gt;This ease of use saves a lot of time and effort during development, making it a fantastic choice for rapid project builds.&lt;/p&gt;

&lt;h3&gt;
  
  
  MongoDB as a Primary Database
&lt;/h3&gt;

&lt;p&gt;MongoDB is ideal for prototyping, thanks to its flexibility. Instead of designing 30 relational tables, you can create just 5 collections by nesting data directly within documents. This approach not only saves time but also simplifies development.&lt;/p&gt;

&lt;p&gt;Additionally, MongoDB’s straightforward schema structure makes it easier to provide context for AI, helping it generate better code with fewer corrections.&lt;/p&gt;

&lt;h2&gt;
  
  
  Use Boilerplates or Create Your Own
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Frflayogl1lnoghqqxqbu.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Frflayogl1lnoghqqxqbu.png" alt="Use Boilerplates or Create Your Own" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Boilerplates are a massive time-saver, and I highly recommend either purchasing one that fits your needs or creating your own. Personally, I use my own boilerplate because, over years of building projects, I’ve compiled a collection of reusable modules, components, and utilities that I can plug into any new app.&lt;/p&gt;

&lt;p&gt;Beyond boilerplates, I also suggest relying on UI libraries, open-source components, and pre-built solutions whenever possible. The less code you have to write from scratch, the faster you can ship your project.&lt;/p&gt;

&lt;h2&gt;
  
  
  Let AI Write 90% of Your Code
&lt;/h2&gt;

&lt;p&gt;As a developer, your job is no longer just about writing code. Think of yourself as the architect — the one who organizes the codebase, plans functionality, and prioritizes features. You’re the tech lead and product manager, not the coder.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fcwfslo7obcunwuet1od6.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fcwfslo7obcunwuet1od6.png" alt="Let AI Write 90% of Your Code" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Tools like Cursor, ChatGPT, and others should be treated as your developers. Your role is to ensure they produce the right code by providing clear instructions, reviewing their output, and refining as necessary. By leveraging AI to handle most of the coding, you can focus on the bigger picture, reduce repetitive tasks, and ship products faster than ever before. Here are the two tools I rely on the most:&lt;/p&gt;

&lt;h3&gt;
  
  
  1. ChatGPT
&lt;/h3&gt;

&lt;p&gt;It might sound obvious, but ChatGPT remains a game-changer. I use it for a variety of tasks: coding, bootstrapping project structures, writing documentation and promo content, and even analyzing data.&lt;/p&gt;

&lt;p&gt;One standout feature is GPTs — pre-defined AI agents that I customize to fit my stack and workflow. These agents save me from repeating the same prompts, allowing me to dive straight into productivity. Here’s how I use ChatGPT:&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Planning and Research:&lt;/strong&gt;&lt;br&gt;
Using the o1-preview model, I begin by researching and planning the structure of the project. This includes outlining features, deciding on architecture, and organizing priorities. ChatGPT helps me brainstorm ideas, validate assumptions, and create a roadmap.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Generating Code:&lt;/strong&gt;&lt;br&gt;
Once the plan is in place, I switch to gpt-4o for bootstrapping the code. It generates models, repositories, services, and other components to give me a head start. My custom GPT agent, trained on my tech stack, ensures I don’t waste time explaining my tools in every prompt—it already knows what I need.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Landing Pages:&lt;/strong&gt;&lt;br&gt;
I use it to create a structure and copy for all my landing pages.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Writing Texts:&lt;/strong&gt;&lt;br&gt;
From basic copy to Terms of Use and Privacy Policies, GPT makes it easy. I use my own pre-defined GPT with a set tone and rules to quickly create any text I need.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. Cursor AI
&lt;/h3&gt;

&lt;p&gt;Once the initial structure is ready, I switch to Cursor AI (using the Claude-3.5 Sonnet model) for the detailed implementation phase. It helps refine the codebase and handle deeper development tasks.&lt;/p&gt;

&lt;p&gt;You can also use Cursor for bootstrapping but I personally like ChatGPT for such kind of tasks a little bit more.&lt;/p&gt;

&lt;h3&gt;
  
  
  Why These Two?
&lt;/h3&gt;

&lt;p&gt;While there are many AI tools out there, ChatGPT and Cursor AI cover 99% of what I need, from ideation to deployment. Together, they allow me to spend less time coding and more time orchestrating the project like an architect.&lt;/p&gt;

&lt;h2&gt;
  
  
  Don’t Spend Time on Unimportant Parts
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fsokoic1dl2wdoehpjgsr.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fsokoic1dl2wdoehpjgsr.png" alt="Don’t Spend Time on Unimportant Parts" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Focus on what truly matters. Spend your time on features that solve user problems and deliver value — not on things like logos, design, or minor details. A great-looking logotype won’t matter if your project doesn’t address real pain points.&lt;/p&gt;

&lt;p&gt;Validate your idea first. If it works, you can always refine the extras later.&lt;/p&gt;

&lt;h2&gt;
  
  
  That’s it for now!
&lt;/h2&gt;

&lt;p&gt;I’ll be sharing more tips and insights in future articles, so if you found this helpful, give it a like or follow me here on Medium.&lt;/p&gt;

&lt;p&gt;Check out the project I built: &lt;a href="https://uniqmail.app" rel="noopener noreferrer"&gt;UniqMail&lt;/a&gt;. Feel free to explore it and let me know what you think!&lt;/p&gt;

&lt;p&gt;I also help people build cool projects. If you know anyone looking to create a great MVP, send them my way: &lt;a href="https://justonedev.co" rel="noopener noreferrer"&gt;JustOneDev&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Thanks for reading, and see you in the next article!&lt;/p&gt;

</description>
      <category>ai</category>
      <category>development</category>
      <category>saas</category>
      <category>productivity</category>
    </item>
  </channel>
</rss>
