DEV Community

Albert
Albert

Posted on

Why I built an open-source benchmark reference UI - and what I refused to add

http://benchmarkwatcher.online/

https://github.com/alikatgh/benchmarkwatcher

This project started as a negative space exercise.

I wanted to build a UI that displays historical benchmark observations clearly — and then deliberately not add everything people expect once numbers are on a screen.

No predictions.
No signals.
No recommendations.
No interpretive language.

Just recorded values, changes, and timestamps — presented as reference material.

BenchmarkWatcher is an open-source web interface for displaying commodity benchmark observations.

It focuses on:

  • clarity of displayed data
  • neutral terminology
  • consistent visual encoding
  • accessibility and keyboard navigation

Every label, badge, and tooltip is written to describe what is shown, not what it means.

Early on, I wrote explicit UI vocabulary rules for myself.

Certain words never appear in the interface — not because they’re inaccurate, but because they imply interpretation:

  • no predictive language
  • no evaluative language
  • no advisory framing

This wasn’t about legal defensiveness.

It was about honesty.

If a UI presents historical observations, it shouldn’t quietly suggest conclusions.

Throughout development, I consciously declined features that would have pushed the UI into analysis:

  • trend labeling
  • performance comparisons
  • alerts or signals
  • directional advice
  • interpretive summaries

Each one would have been easy to add.
Each one would have changed what the tool is.

From a technical perspective, the project emphasizes:

  • strict separation between data and presentation
  • client-side state stored with explicit defaults
  • accessibility-first modal and control design
  • theme consistency without runtime surprises

The hardest part wasn’t writing code.
It was keeping the scope intact.

This is useful if you want:

  • a clean reference display
  • neutral historical context
  • a UI that doesn’t editorialize

It is not designed for:

  • decision-making workflows
  • forecasting
  • comparative evaluation

I’m sharing this primarily for feedback from people who care about:

  • UI semantics
  • wording discipline
  • scope control in open-source projects

If that resonates, I’d appreciate a look or a comment.

http://benchmarkwatcher.online/

https://github.com/alikatgh/benchmarkwatcher

Top comments (0)