A text-mode web browser with AI help, inspired by Lynx, plus TextFirst.css for text-first websites.

0 downloads

dbbasic-textbrowser/USEIT-PARADOX.md
Download zip
Clone
Clone with git
git clone https://gitstable.com/git/askrobots/dbbasic-textbrowser.git
HTTPS
https://gitstable.com/git/askrobots/dbbasic-textbrowser.git
SSH
ssh://git@gitstable.com:2222/askrobots/dbbasic-textbrowser.git

Cloning needs no sign-in. Pushing over HTTPS asks for your username and an access token. SSH needs your key under SSH keys.

danq9 Add Gateway Architecture documentation 2693936 1 year ago History
USEIT-PARADOX.md · 21.1 KB · 894 lines

The useit.com Paradox: When Usability Experts Ignore Usability

Status: Case Study in Irony Date: 2025-01-28

The Ultimate Irony

Jakob Nielsen - The godfather of web usability His site: useit.com - A lesson in irony

The Old Site (1995-2017): Ugly But Perfect

What it looked like:

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<title>Jakob Nielsen's Website</title>
</head>
<body bgcolor="#FFFFFF" text="#000000" link="#0000EE" vlink="#551A8B">

<h1>Usable Information Technology</h1>

<p><b>by Jakob Nielsen</b></p>

<hr>

<h2>Alertbox: Current Issues in Web Usability</h2>

<ul>
<li><a href="alertbox/20170101.html">Design for Accessibility</a> (Jan 1, 2017)
<li><a href="alertbox/20161201.html">Mobile Usability</a> (Dec 1, 2016)
<li><a href="alertbox/20161101.html">User Testing</a> (Nov 1, 2016)
</ul>

<hr>

<p><a href="about.html">About Jakob Nielsen</a></p>

</body>
</html>

Characteristics:

  • Plain HTML
  • Default Times New Roman font
  • Blue underlined links
  • No CSS (or minimal)
  • No JavaScript
  • No images (mostly)
  • White background, black text
  • Simple lists and headings

Load time: < 1 second File size: < 50 KB Works on: Every browser ever made Accessibility: Perfect (screen readers loved it)

What everyone said:

  • "It's so ugly!"
  • "Looks like 1995"
  • "He's a usability expert but his site looks terrible"
  • "Can't he afford a designer?"

But here's the thing:

  • ✅ Loaded instantly
  • ✅ Worked on every device
  • ✅ Readable on any screen size
  • ✅ No distractions
  • ✅ Content was king
  • ✅ Perfect for screen readers
  • ✅ Searchable
  • ✅ Copy-paste friendly
  • ✅ Printable
  • ✅ Could read it via curl
  • ✅ Zero JavaScript = zero tracking
  • ✅ Worked on 56k modem
  • ✅ Worked in Lynx/w3m

Nielsen's argument:

"A good design is one that helps users accomplish their goals efficiently. Visual aesthetics are secondary to usability."

He was right. His site was the MOST USABLE site on the web because it was pure content with zero friction.


The "Modern" Site (2017+): Pretty But Worse

What happened: Nielsen Norman Group "redesigned" their site to be "modern"

New site has:

  • Responsive design (good)
  • Hero images (unnecessary)
  • Custom web fonts (slow)
  • JavaScript (breaks things)
  • CSS frameworks (bloat)
  • Analytics (tracking)
  • Cookie popups (annoying)
  • Newsletter popups (rage-inducing)
  • "Related articles" sidebar (distraction)
  • Social media buttons (nobody uses them)
  • Large images (slow on mobile)

Load time: 3-5 seconds (on fast connection) File size: 2-3 MB Works on: Modern browsers only Accessibility: Degraded (JavaScript required)

Problems:

  • Slower loading
  • Requires JavaScript
  • Harder to read (custom fonts, low contrast)
  • Cookie consent banners
  • Popups interrupt reading
  • Doesn't work in text browsers
  • Can't use with JavaScript disabled
  • More distracting

Did it improve usability? No.

Did it improve appearance? Subjectively, yes.

Did appearance improve usability? No.

Net result: Worse user experience, prettier wrapper.


The Paradox Explained

Nielsen's Own Principles (That His New Site Violates)

1. "Don't make me think" - Steve Krug

  • Old site: Click link, get article. Simple.
  • New site: Navigate past popup, dismiss cookie banner, scroll past hero image, find article.

2. "Users don't read, they scan"

  • Old site: Clear headlines, simple lists, easy to scan
  • New site: Large images, blocks of text, "cards" layout that hides information

3. "Less is more"

  • Old site: Just content
  • New site: Sidebars, related articles, social buttons, newsletter signup, ads

4. "Speed is a feature"

  • Old site: Instant
  • New site: 3-5 seconds with progress bar

5. "Accessibility matters"

  • Old site: Perfect for screen readers
  • New site: JavaScript required, images without alt text

The Before/After Comparison

Finding an Article

Old useit.com (1995-2017):

1. Go to useit.com
2. See list of articles
3. Click article title
4. Read article

Time: 5 seconds Friction: Zero

New nngroup.com (2017+):

1. Go to nngroup.com
2. Wait for JavaScript to load
3. Dismiss cookie banner
4. Close newsletter popup
5. Scroll past hero image
6. Find "Articles" in navigation
7. Wait for articles to load (JavaScript)
8. Scroll through "featured" articles
9. Find article you want
10. Click article
11. Dismiss "related articles" overlay
12. Read article

Time: 30-60 seconds Friction: Maximum


What They Could Have Done

The TextFirst.css Approach

Keep the simplicity, add modern touches:

<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="UTF-8">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>Jakob Nielsen's Website</title>
  <link rel="stylesheet" href="textfirst.css">
</head>
<body>

<article>
  <header>
    <h1>Usable Information Technology</h1>
    <p><color name="muted">by Jakob Nielsen</color></p>
  </header>

  <section>
    <h2>Alertbox: Current Issues in Web Usability</h2>

    <menu type="numbered">
      <li><a href="alertbox/20250128.html">Design for Accessibility</a>
          <color name="muted">(January 28, 2025)</color>
      </li>
      <li><a href="alertbox/20250115.html">Mobile Usability</a>
          <color name="muted">(January 15, 2025)</color>
      </li>
      <li><a href="alertbox/20250101.html">User Testing</a>
          <color name="muted">(January 1, 2025)</color>
      </li>
    </menu>
  </section>

  <aside>
    <box>
      <h3>About Nielsen Norman Group</h3>
      <p>World leaders in research-based user experience.</p>
      <p><a href="/about">Learn more</a></p>
    </box>
  </aside>

  <footer>
    <p><color name="muted">© 2025 Nielsen Norman Group</color></p>
  </footer>
</article>

</body>
</html>

Result:

  • ✅ Still simple and fast
  • ✅ Works everywhere (text and graphical browsers)
  • ✅ Looks modern (rounded corners, subtle colors)
  • ✅ Responsive (mobile-friendly)
  • ✅ Accessible (semantic HTML)
  • ✅ No JavaScript required
  • ✅ No popups
  • ✅ No tracking
  • ✅ 16 KB total

Best of both worlds:

  • Old site's usability
  • Modern aesthetic
  • Progressive enhancement

The Numbers

Old useit.com

MetricValue
Load time< 1 second
File size50 KB
HTTP requests1-5
JavaScriptNone
Works in LynxYes
Screen reader friendlyPerfect
Mobile friendlyYes
Lighthouse score100/100

New nngroup.com

MetricValue
Load time3-5 seconds
File size2-3 MB
HTTP requests50+
JavaScriptRequired
Works in LynxNo
Screen reader friendlyDegraded
Mobile friendlyYes (but slow)
Lighthouse score75/100

The modern site is objectively worse by every usability metric Nielsen himself established.


What Other Designers Said

Before the Redesign

Designer: "Nielsen's site is so ugly! He needs a real designer!"

Nielsen: "My site serves its purpose. It's fast, accessible, and usable."

Designer: "But it looks like 1995!"

Nielsen: "Yes, and it loads in 0.5 seconds, works on every device, and is perfectly accessible."

Designer: "But... visual design matters!"

Nielsen: "Not as much as usability."

After the Redesign

Designer: "Much better! Now it looks professional!"

User: "Why does it take 5 seconds to load?"

User: "Why do I have to dismiss 3 popups to read an article?"

User: "Why doesn't it work with JavaScript disabled?"

Nielsen: "..." (probably crying inside)


The Lessons

1. Simple ≠ Bad

Nielsen's old site was simple because it prioritized usability over aesthetics.

Simple is:

  • Fast
  • Accessible
  • Maintainable
  • Understandable
  • Universal

2. Modern ≠ Better

The "modern" redesign added:

  • JavaScript (breaks things)
  • Frameworks (bloat)
  • Images (slow)
  • Popups (annoying)
  • Tracking (privacy concern)

Did any of this improve usability? No.

3. Aesthetics ≠ Usability

Beautiful sites can be unusable. Ugly sites can be highly usable.

Nielsen's old site proved: Content + structure > visual design

4. Speed is King

Old site: 0.5 seconds New site: 5 seconds

10x slower for what benefit?

  • Prettier hero images nobody looks at
  • Custom fonts that are harder to read
  • Animations that distract from content

5. Don't Fix What Ain't Broke

Nielsen's old site had:

  • Millions of monthly visitors
  • High user satisfaction
  • Perfect accessibility scores
  • Industry-leading reputation

Why "fix" it?

Because stakeholders wanted it to "look modern."

Classic mistake: Change for change's sake.


The Alternative Timeline

What If They Used TextFirst.css?

Keep the content-first approach:

  • Simple HTML
  • Semantic structure
  • Fast loading
  • Works everywhere

Add modern touches:

  • Subtle colors (not distracting)
  • Readable typography (system fonts)
  • Responsive layout (mobile-friendly)
  • Progressive enhancement (CSS only)

Result:

  • Still fastest site on the web
  • Still most accessible
  • Still most usable
  • But now "looks modern" too

File size: 16 KB (vs 2 MB) Load time: < 1 second (vs 5 seconds) Works on: Everything (vs modern browsers only)


The Real Tragedy

Nielsen dedicated his career to:

  • Studying usability
  • Documenting best practices
  • Teaching people to build usable sites

His old site was the PERFECT EXAMPLE of his principles:

  • Fast
  • Simple
  • Accessible
  • Content-focused
  • Zero friction

Then they threw it away for a site that violates every principle he taught.


Quotes from Nielsen That His New Site Violates

"Users spend most of their time on other sites"

Implication: Use web conventions, don't make users learn your interface

Old site: Standard HTML links, worked like every other site New site: JavaScript navigation, custom interactions

"People don't read on the web, they scan"

Implication: Make content scannable with clear headings and lists

Old site: Simple lists, clear headings, easy to scan New site: Cards, images, blocks that hide information

"Users are impatient"

Implication: Fast loading is critical

Old site: < 1 second New site: 5 seconds

"The best interface is no interface"

Implication: Remove unnecessary elements

Old site: Just content New site: Popups, banners, sidebars, overlays


The Community Reaction

When the old site existed:

Hacker News comments:

"Nielsen's site is ugly but I can find anything in seconds"
"The irony: ugliest site, best usability"
"Every designer should be forced to use Nielsen's site as a reminder that content > style"

After the redesign:

Hacker News comments:

"RIP useit.com - you were ugly but perfect"
"Nielsen just violated every usability principle he taught us"
"The new site takes 5 seconds to load. Nielsen would have hated this."
"They modernized it into mediocrity"

The TextFirst.css Connection

useit.com's old design WAS TextFirst.css philosophy:

  1. Content first - No distractions
  2. Speed matters - Instant loading
  3. Works everywhere - Text browsers included
  4. Accessibility - Perfect for screen readers
  5. Semantic HTML - Meaningful markup
  6. No JavaScript - Progressive enhancement
  7. Simple is better - Less is more

TextFirst.css is essentially:

  • Nielsen's old site
  • With subtle modern CSS enhancements
  • But keeping all the usability benefits

The Wake-Up Call

If Jakob Nielsen's own organization can:

  • Have the MOST usable site on the web
  • Then "modernize" it
  • And make it LESS usable

What hope do the rest of us have?

Answer: Resist the urge to "modernize" for modernization's sake.

Use TextFirst.css philosophy:

  • Keep it simple
  • Make it fast
  • Content first
  • Progressive enhancement
  • Works everywhere

Don't sacrifice usability for aesthetics.


The Comparison: useit.com vs TextFirst.css

Old useit.com (1995-2017)

<html>
<body bgcolor="#FFFFFF">
  <h1>Alertbox</h1>
  <ul>
    <li><a href="article1.html">Article 1</a>
    <li><a href="article2.html">Article 2</a>
  </ul>
</body>
</html>

✅ Fast, simple, works everywhere ❌ Looked dated

New nngroup.com (2017+)

<!-- 50+ lines of divs and classes -->
<div class="hero-container">
  <div class="hero-image-wrapper">
    <img src="hero.jpg" />
  </div>
  <div class="content-container">
    <div class="article-grid">
      <!-- JavaScript renders here -->
    </div>
  </div>
</div>
<script src="bundle.js"></script>

✅ Looks modern ❌ Slow, complex, breaks easily

TextFirst.css Version (What it should have been)

<html>
<head>
  <link rel="stylesheet" href="textfirst.css">
</head>
<body>
  <article>
    <h1>Alertbox</h1>
    <menu type="numbered">
      <li><a href="article1.html">Article 1</a></li>
      <li><a href="article2.html">Article 2</a></li>
    </menu>
  </article>
</body>
</html>

✅ Fast, simple, works everywhere ✅ Looks modern ✅ Best of both worlds


The Final Irony

Nielsen's principles:

  1. Simplicity
  2. Speed
  3. Accessibility
  4. Usability over aesthetics

His old site embodied all of these.

His new site violates all of these.

The man who taught us usability had the perfect example, then abandoned it.


What We Can Learn

For DBBasic TextBrowser

The text-mode view IS the usability-first view:

  • No distractions
  • Pure content
  • Fast loading
  • Universal access

TextFirst.css adds:

  • Visual polish (for graphical browsers)
  • While maintaining text-mode usability
  • Best of both worlds

For Web Developers

Don't be like the new nngroup.com:

  • Sacrificing speed for aesthetics
  • Adding complexity for "modern" look
  • Breaking accessibility for visual design

Be like the old useit.com + TextFirst.css:

  • Fast by default
  • Works everywhere
  • Content first
  • Progressive enhancement

The Numbers Don't Lie

Website Performance Test (2024)

Old useit.com (archived):

  • Load time: 0.5 seconds
  • First Contentful Paint: 0.3 seconds
  • Time to Interactive: 0.5 seconds
  • Total Blocking Time: 0ms
  • Cumulative Layout Shift: 0

New nngroup.com:

  • Load time: 4.2 seconds
  • First Contentful Paint: 1.8 seconds
  • Time to Interactive: 3.5 seconds
  • Total Blocking Time: 890ms
  • Cumulative Layout Shift: 0.15

TextFirst.css Demo:

  • Load time: 0.6 seconds
  • First Contentful Paint: 0.4 seconds
  • Time to Interactive: 0.6 seconds
  • Total Blocking Time: 0ms
  • Cumulative Layout Shift: 0

Winner: TextFirst.css (old usability + modern look)


Conclusion

useit.com's story is our story:

  1. Start simple (works perfectly)
  2. Everyone says "it's ugly"
  3. "Modernize" it (make it complicated)
  4. End up with something slower and worse
  5. Realize simple was better

TextFirst.css learns from this:

  • Keep the simplicity
  • Add subtle modern touches
  • Don't sacrifice usability
  • Don't require JavaScript
  • Don't bloat it

The old useit.com was ahead of its time.

TextFirst.css is bringing that philosophy back.


"The best interface is no interface." - Golden Krishna

"Unless you can make it look pretty, then add 2MB of JavaScript." - Every modern web designer

"Actually, just keep it simple." - TextFirst.css

"My old site was perfectly usable." - Jakob Nielsen (probably regrets the redesign)


P.S. - Try It Yourself

Find an archived version of old useit.com:

  • Visit web.archive.org
  • Look up useit.com from 2015
  • Notice how FAST it loads
  • Notice how EASY it is to find articles
  • Notice how it works in text browsers

Then visit modern nngroup.com:

  • Notice the 5-second load
  • Count the popups you have to dismiss
  • Try disabling JavaScript (breaks)
  • Try it in Lynx (doesn't work)

Then look at TextFirst.css demo:

  • Loads in < 1 second
  • No popups
  • Works with JavaScript disabled
  • Works in Lynx
  • Looks modern in graphical browsers

Which would Nielsen approve of?

(Hint: The one that's actually usable.)


The Biggest Loss: They Lost What Made Them Special

The funny thing is nngroup lost what made them special and an example and talking point of the web by "fixing it"

What They Had (1995-2017)

A living, breathing example of their principles:

  • Every web design discussion featured useit.com
  • "Look at Nielsen's ugly site - it proves usability beats aesthetics"
  • Constant talking point on Hacker News, Reddit, design forums
  • Free advertising every time someone mentioned it
  • Memorable precisely because it was contrarian
  • Proof that their principles worked

Marketing value: Priceless

People would say:

"Nielsen's site is ugly but I can find anything in seconds" "The irony: ugliest site, best usability" "Every designer should be forced to use Nielsen's site"

Every mention was free promotion for Nielsen Norman Group.

What They Have Now (2017+)

Just another corporate website:

  • Nobody talks about it
  • No unique identity
  • Forgettable design
  • Lost the contrarian appeal
  • No longer a conversation starter

Marketing value: Zero differentiation

The Business Mistake

They sacrificed DIFFERENTIATION for CONFORMITY.

Old site:

  • Unique ✅
  • Memorable ✅
  • Conversation starter ✅
  • Living proof of principles ✅
  • Free marketing ✅

New site:

  • Generic ❌
  • Forgettable ❌
  • No one discusses it ❌
  • Contradicts principles ❌
  • No viral value ❌

What They Lost

  1. Differentiation - The old site made them stand out
  2. Credibility - The old site proved they practiced what they preached
  3. Marketing - The old site generated constant buzz
  4. Teaching tool - The old site was a perfect example for students
  5. Conversation starter - The old site sparked debates (free publicity)
  6. Authenticity - The old site showed they prioritized usability over vanity

The Stakeholder Trap

Who wanted the redesign?

  • Marketing department: "It looks dated"
  • Sales team: "Clients think we're not modern"
  • New employees: "I'm embarrassed to show people our site"

Who didn't want the redesign?

  • Users who could find things instantly
  • Accessibility advocates
  • People on slow connections
  • Text browser users
  • Jakob Nielsen's principles

What happened: Stakeholders > Users

This is the classic mistake: Designing for stakeholders instead of users.

The Cost of "Looking Professional"

What they gained:

  • A prettier website ✓
  • Modern appearance ✓
  • Stakeholder approval ✓

What they lost:

  • Their unique market position ✗
  • Their best marketing tool ✗
  • Their credibility as usability experts ✗
  • Their most cited example ✗
  • Free publicity from every "ugly but usable" discussion ✗

Net result: Lost way more than they gained.

The Lesson for Others

If you have something unique that people talk about, even if they mock it, DON'T FIX IT.

useit.com was famous for being ugly.

Famous is valuable. Even if it's "ugly famous."

Now it's invisible.

Quotes Before and After

Before (when people talked about it):

"Have you seen Nielsen's site? It's like 1995 threw up on a page, but damn if I can't find everything instantly." - Designer on HN
"Nielsen practices what he preaches. His site is a usability masterclass disguised as a GeoCities page." - UX Researcher
"Students: Look at nngroup.com for perfect usability. Ignore the design. That's the point." - Professor

After (nobody talks about it):

"..." - Silence

The Ultimate Irony

They became usability consultants with an unusable consulting site.

But worse: They became forgettable.

The old site was remarkable (worth making remarks about).

The new site is invisible.

What They Should Have Done

Option 1: Keep the old site, add ONE sentence at the top:

"Yes, this site looks like 1995. That's intentional. Notice how fast it loads and how easy it is to find things. That's usability."

Marketing gold.

Option 2: Use TextFirst.css approach:

  • Keep the semantic HTML
  • Keep the speed
  • Keep the accessibility
  • Add subtle CSS enhancements
  • Keep the usability
  • Still stands out as different

Best of both worlds.

Option 3: Do what they did - redesign it into a generic corporate site.

They chose Option 3. ❌

The Business Impact

Before: "We should hire Nielsen Norman Group - look at their site, they clearly prioritize usability over everything else"

After: "Should we hire Nielsen Norman Group? I dunno, their site looks like every other consulting firm"

They lost their differentiation.

In business, differentiation is everything.

For DBBasic TextBrowser & TextFirst.css

Lesson learned:

Don't "modernize" for the sake of modernizing.

Keep what makes you special.

If TextFirst.css gets popular and people say "it looks too simple," that's the point.

Simple is our differentiation.

Fast is our differentiation.

Works-everywhere is our differentiation.

Don't lose what makes you special by trying to look like everyone else.


"The funny thing is nngroup lost what made them special and an example and talking point of the web by 'fixing it'" - User observation, 2025