<?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: Korean student Lee</title>
    <description>The latest articles on DEV Community by Korean student Lee (@koreanstudentlee).</description>
    <link>https://dev.to/koreanstudentlee</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fthepracticaldev.s3.amazonaws.com%2Fi%2F99mvlsfu5tfj9m7ku25d.png</url>
      <title>DEV Community: Korean student Lee</title>
      <link>https://dev.to/koreanstudentlee</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/koreanstudentlee"/>
    <language>en</language>
    <item>
      <title>I built 15 developer tools that never touch a server — here is what the local-only parts actually cost</title>
      <dc:creator>Korean student Lee</dc:creator>
      <pubDate>Mon, 16 Nov 2026 01:53:23 +0000</pubDate>
      <link>https://dev.to/koreanstudentlee/i-built-15-developer-tools-that-never-touch-a-server-here-is-what-the-local-only-parts-actually-2ag0</link>
      <guid>https://dev.to/koreanstudentlee/i-built-15-developer-tools-that-never-touch-a-server-here-is-what-the-local-only-parts-actually-2ag0</guid>
      <description>&lt;p&gt;I am a student in Korea, and I kept running into the same small annoyance: every time I needed to compress an image for a blog post, decode a JWT, or check a color contrast ratio, the fastest search result was a site covered in ads that wanted me to &lt;em&gt;upload&lt;/em&gt; the file first.&lt;/p&gt;

&lt;p&gt;Uploading a screenshot of an internal dashboard to a random image compressor is not something I want to do, and it is not something a company should do either. So I started building the tools I actually use, with one rule: &lt;strong&gt;nothing goes over the network&lt;/strong&gt;. Everything runs in the browser tab, on the user's own machine.&lt;/p&gt;

&lt;p&gt;I have been shipping one tool a day, and the hub now has 15 of them (16 pages counting the standalone builds). This post is about the parts that were harder than they looked, rather than a feature list.&lt;/p&gt;

&lt;h2&gt;
  
  
  The rule, and where it breaks
&lt;/h2&gt;

&lt;p&gt;The constraint is simple: a tool may use the DOM, Canvas, Web APIs, and the user's own CPU. It may not POST a file anywhere. No backend, no analytics endpoint that receives content, no "free tier" API key baked into the page.&lt;/p&gt;

&lt;p&gt;That works surprisingly well until you hit one of three walls.&lt;/p&gt;

&lt;h3&gt;
  
  
  1. Anything that "just needs a library" also needs a bundle
&lt;/h3&gt;

&lt;p&gt;A regex tester is easy. A &lt;em&gt;good&lt;/em&gt; regex tester needs a highlighting engine, and pulling a 200 kB syntax highlighter into a page that otherwise ships 20 kB of inline JavaScript is a bad trade for a tool that should feel instant on a mid-range Android phone. I ended up writing the tokenizer for the SQL formatter and the diff engine by hand instead of importing one:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;SQL formatter&lt;/strong&gt;: a small tokenizer that classifies strings, quoted identifiers, comments, and keywords, then re-emits them with indentation rules. Static syntax hints come from the same token stream, so there is no second pass.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Diff checker&lt;/strong&gt;: line-level LCS for the aligned view, then a word-level pass inside each changed line so a one-word edit does not paint the whole line red.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Both are a few hundred lines. That is less code than the configuration needed to tree-shake a library down to the same behavior.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. Text encoding is not &lt;code&gt;btoa&lt;/code&gt;
&lt;/h3&gt;

&lt;p&gt;This one bit me on the Base64 tool. &lt;code&gt;btoa()&lt;/code&gt; throws on anything outside Latin-1, which means it fails on Korean text and on emoji — the two things a Korean user will paste first.&lt;/p&gt;

&lt;p&gt;The fix is to stop thinking in strings and start thinking in bytes:&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;const&lt;/span&gt; &lt;span class="n"&gt;bytes&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;TextEncoder&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="nf"&gt;encode&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;text&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="n"&gt;binary&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="nb"&gt;Array&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="n"&gt;bytes&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;b&lt;/span&gt; &lt;span class="o"&gt;=&amp;gt;&lt;/span&gt; &lt;span class="nb"&gt;String&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;fromCharCode&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;b&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="nf"&gt;join&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="kd"&gt;const&lt;/span&gt; &lt;span class="n"&gt;b64&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="nf"&gt;btoa&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;binary&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;Decoding needs the mirror image (&lt;code&gt;atob&lt;/code&gt; → &lt;code&gt;Uint8Array&lt;/code&gt; → &lt;code&gt;TextDecoder&lt;/code&gt;), plus tolerance for what people actually paste: data URLs with the &lt;code&gt;data:image/png;base64,&lt;/code&gt; prefix, wrapped lines at 64 characters, stray whitespace, and the URL-safe alphabet with padding stripped. Tolerating that input is most of the work; the conversion itself is four lines.&lt;/p&gt;

&lt;h3&gt;
  
  
  3. Big files are a memory problem, not a CPU problem
&lt;/h3&gt;

&lt;p&gt;An image compressor is trivially "draw to canvas, call &lt;code&gt;toBlob('image/webp', q)&lt;/code&gt;". On a phone, with a 12 MP photo, the failure mode is the tab being killed while three copies of the bitmap are alive: the decoded &lt;code&gt;&amp;lt;img&amp;gt;&lt;/code&gt;, the canvas backing store, and the encoded blob.&lt;/p&gt;

&lt;p&gt;What helped:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;decode once, draw once, then release the source before encoding&lt;/li&gt;
&lt;li&gt;reset &lt;code&gt;canvas.width = 0&lt;/code&gt; after use so the backing store is freed rather than waiting for GC&lt;/li&gt;
&lt;li&gt;show the encoded size against the original &lt;em&gt;before&lt;/em&gt; offering the download, because "smaller" is the only reason anyone opens this tool&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The Base64 tool has the same shape with a different ceiling: a data URL is roughly 4/3 the size of the file, and pasting a 5 MB image into a textarea produces a 6.7 MB string that some editors handle badly. So the file path uses drag-and-drop and &lt;code&gt;FileReader&lt;/code&gt; instead of the textarea, and the page tells you the size increase before you commit.&lt;/p&gt;

&lt;h2&gt;
  
  
  What a JWT decoder should warn you about
&lt;/h2&gt;

&lt;p&gt;Decoding a JWT is base64url plus JSON, so the interesting part is not the decoding. It is telling the user something useful:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;code&gt;alg: none&lt;/code&gt; is accepted by some badly written verifiers, so flagging it matters more than pretty-printing the claims&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;exp&lt;/code&gt; in the past is common in bug reports, so a live countdown is worth more than a timestamp&lt;/li&gt;
&lt;li&gt;Korean claims arrive as escaped Unicode, so &lt;code&gt;\uXXXX&lt;/code&gt; sequences need normalizing before display or the payload looks like noise&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;I added HS256 verification as an optional step — the signature check happens in the browser with the pasted secret, and the secret is never transmitted. Verification that requires a server would have broken the rule, so it stays optional and clearly labeled.&lt;/p&gt;

&lt;h2&gt;
  
  
  A note on contrast and accessibility checks
&lt;/h2&gt;

&lt;p&gt;WCAG contrast is a formula, not a service: relative luminance with gamma correction, then &lt;code&gt;(L1 + 0.05) / (L2 + 0.05)&lt;/code&gt;. The part people get wrong is the threshold table — normal text needs 4.5:1, large text 3:1, and UI components 3:1 — so the tool shows all three verdicts side by side instead of a single pass/fail.&lt;/p&gt;

&lt;h2&gt;
  
  
  The honest limits
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;No collaboration or history.&lt;/strong&gt; Everything is per-tab. Reload and it is gone, which is the price of "we never receive your data".&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;No server-side parsing.&lt;/strong&gt; PDF and Office file input is basically out of scope; I would rather skip a tool than fake it badly.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Browser support.&lt;/strong&gt; &lt;code&gt;OffscreenCanvas&lt;/code&gt; and WebP encoding in canvas are fine in current Chrome, Edge, Firefox, and Safari, but older Android WebViews fall back to JPEG.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Where it is going
&lt;/h2&gt;

&lt;p&gt;One tool a day, chosen by what annoyed me that week. Next up is a UUID/ULID/SHA-256 generator, then whatever breaks most in the feedback.&lt;/p&gt;

&lt;p&gt;If you have a utility you keep reaching for that only exists as an ad-covered upload page, I would genuinely like to hear it — that is how the list has been growing.&lt;/p&gt;

&lt;p&gt;The hub is here, and every tool is free, with no signup: &lt;strong&gt;&lt;a href="https://lee-know-share.blogspot.com/p/dailytools.html" rel="noopener noreferrer"&gt;https://lee-know-share.blogspot.com/p/dailytools.html&lt;/a&gt;&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;&lt;em&gt;No affiliate links, no tracking pixel on the tools. If a tool uploads anything, that is a bug.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>showdev</category>
      <category>webdev</category>
      <category>javascript</category>
      <category>security</category>
    </item>
  </channel>
</rss>
