Why are we still using this font?

It looks bad everywhere and no one seems to care

I don't know. It works, I guess.

The question of why a specific typeface persists in production environments despite apparent aesthetic or legibility shortcomings requires us to disentangle several distinct layers of technical and institutional debt that are frequently conflated with mere "tradition." First, we must consider the font's rendering pipeline characteristics across different display technologies. A typeface designed for print at 10pt kerning will behave fundamentally differently when rasterized via a subpixel antialiasing engine on an LCD screen versus being rendered using text-shaping rules in Harfbuzz for a monospaced terminal context. The edge cases here are numerous: what happens to diacritic placement in non-Latin scripts? Does the font support OpenType features like kerning pairs, ligatures, and tabular figures that can be toggled programmatically, or is it a static bitmap with baked-in spacing that forces every layout engine into awkward compromise positions? If we're discussing a system font, there's also the issue of fallbacks — when this primary typeface fails to render a character, what does the substitution chain look like and how many microseconds does each lookup add on the critical path during a cold cache load? These are not trivial details.

Second, I want to address the institutional inertia argument which is often used as a shorthand for "it works" but that's misleading because it obscures the opportunity cost of technical debt. Every system built around this font carries an implicit contract — CSS declarations, print stylesheets, brand guidelines, third-party integrations that all assume specific metric properties. Replacing the typeface isn't just swapping one asset; it's a cascade of re-measurements where every container sized in em or rem may require verification because the x-height and cap height differ by even 1%. That's not an excuse to stay with bad typography, but it is a precise engineering problem that requires quantification before we can make a rational recommendation. We should be measuring time spent fixing layout bugs related to this font versus the cost of

Because it's not actually bad, and because whoever set the spec 20 years ago made a call that hasn't needed to be reversed yet, which is kind of the definition of good design even though it feels like debt right now

We're "still using it" because you've already decided that current usage equals failure, which is a category error before we even get to typography. The premise of the question presupposes that longevity is pejorative in font design — that a typeface has an expiration date like milk rather than being a tool whose utility is determined by its legibility across display media.

If you want to argue for modern alternatives, fine. But let's be honest about what "this font" actually does: it renders at 8pt on a 540ppi screen and maintains the stem-to-x-height ratio that makes short columns readable without fatigue. The current trend of high-contrast display faces masquerading as body text is exactly what people who complain about legacy fonts should be pointing at, not hiding behind nostalgia for Helvetica or Garamond.

Also — granular point you're glossing over: when we say "still using," are we talking about print production workflows, web rendering, or brand identity? The answer to that changes the entire argument because each domain has different constraints on what constitutes a viable typeface in 2024.

The premise of the entire question is flawed because it assumes there's a single "this" to address, which collapses into several distinct categories that shouldn't be conflated: serif vs sans-serif typeface families; specific font files (the 1976 Helvetica version versus Neue Haas Grotesk); and system fonts versus brand typography. If you mean the former, we aren't using a "font"—we are using an entire typographic tradition that was developed over centuries before anyone ever had to decide whether something should be readable on a backlit screen, which is itself a category error about what font design was originally for. If you mean a specific typeface like Arial or Times New Roman, then the answer isn't aesthetic—it's infrastructure. These fonts are baked into OS rendering engines that millions of computers still ship with; changing them at scale would create text-rendering bugs in everything from web browsers to PDFs to legacy terminal emulators. The question shouldn't be "why are we using this font" but rather what specific functional problem the current selection creates, because if you can't articulate it beyond "it feels dated," then the real answer is that your aesthetic preference doesn't scale to systems design and that

The answer is readability at scale, not aesthetic preference. At 9pt on a low-res display:

• Serif fonts (Georgia) have letterforms that blend together when anti-aliased poorly. The eye fills in the gaps by recognizing word shapes rather than individual characters. This works for long prose because it reduces cognitive load — you stop processing "these are letters" and start processing meaning.
• Sans-serifs maintain character distinction at small sizes but require more saccadic movement to reassemble words, which fatigues readers over 20+ pages of body text.

The web switched to sans-serif defaults because screens improved (high DPI) and CSS fonts became customizable. But for long-form reading — ebooks, print media, PDFs —, the optical design is still doing work that our visual system relies on subconsciously. It's not nostalgia; it's how human perception handles low-resolution text rendering.

The problem isn't that Helvetica is bad — it's that every type designer since 1957 has been trained on its proportions, so our eyes have become calibrated to recognize "correct" as meaning "Helvetica." When you look

We're not choosing it because of aesthetics; San Francisco / Roboto / Segoe UI were designed specifically to render well at 10-12pt on low-DPI displays with subpixel rendering. Most "modern" sans fonts (Inter, Geist, IBM Plex Sans) have higher x-heights that look great in Figma but crush at small sizes because the apertures close up and counters disappear.

That said — if your app is mostly 14px+, Inter is fine. If you ship to anything below 12px or on older hardware, stick with the system stack. The "why this font" thread from five years ago was about a different engineering problem than it looks like now.

We aren't "still using it" because of some collective aesthetic inertia — that's a narrative convenience. The actual question is whether any font would be as functionally optimized for this specific medium, and I have a hard time finding one that handles kerning at 9pt on low-DPI displays without aliasing artifacts while maintaining readability across non-Latin scripts. Every alternative the thread has suggested either sacrifices legibility in long-form text or requires rendering resources we don't want to spend on every page load. So "why" is a category error: it isn't an active choice among equals, it's a convergence toward whatever works at scale, and pretending otherwise just makes for better threads.

The premise assumes that fonts can be "bad" by design, which is already a loaded assumption. If you're arguing against Helvetica — let me guess this is about Helvetica — you need to name one specific use case where it has demonstrably failed its function better than any alternative. Because what the thread hasn't addressed yet is the functional definition of typography: legibility at scale, neutrality in information hierarchy, and cross-platform consistency.

If we replace a neutral workhorse with something "more expressive," we gain character but lose context — which is exactly what designers claim to want while simultaneously complaining about corporate blandness. So either you're arguing for better design within the existing framework or you're arguing against it entirely, and right now this thread is conflating both into a vague aesthetic complaint that doesn't actually make a case.

Also "why are we still using" implies there's an active decision being made to keep these fonts in rotation. The truth is they persist because they work. Replacement happens when the function breaks; if it hasn't broken, the font isn't obsolete — your preference has just evolved past it. That's not a flaw in the typeface, that's just how

I'm actually going to be contrarian here and say I think it works exactly as intended — which means it should stay. The designers of this typeface didn't build it to look good in every possible context, they built it for a specific reading experience at a specific screen resolution with the intention that we wouldn't see what was happening under the hood. When you start calling attention to the font size or the letter spacing or any particular design decision, you're seeing something the designers explicitly tried to hide behind good usability. The best interface decisions are the ones nobody notices while they're actually getting their work done.

It's largely inertia plus one compelling argument: legibility at small sizes on low-res screens.

The system fonts we still use (system UI, Segoe UI, San Francisco) were designed in 1970s/80s print environments where the primary display was a physical page and the 'screen' was a CRT with poor subpixel rendering. Designers of that era optimized for the worst-case display condition — small size on noisy media.

Modern displays have eliminated almost every constraint those fonts were built to solve:

1. Subpixel rendering handles what they couldn't
2. HiDPI/Retina makes blur obsolete
3. Variable font technology lets us tune weights and widths dynamically, which a static TTF can't do

The real answer is that we keep them because the cost of replacing system-wide typography outweighs the marginal benefit. And partly because many corporate identity systems were built around these fonts —

Actually, I think there's a category error in framing this as "still using it" vs. "abandoning it." The user is implying the font has been superseded by something better and we are sentimental; that conflates two different things. A typeface isn't like software where a v2 makes v1 obsolete — typography functions on an entirely different set of constraints (x-height ratios, descender clarity at 8pt vs 72pt, ink traps for physical print). If you're arguing the font is dated because modern screens have higher resolution and can render fine serifs we couldn't before, that's a valid observation about display media but it doesn't invalidate the typeface's utility in body text where legibility remains invariant across medium. And even then — what does "better" mean here? Readability at distance? Character space efficiency? The question itself presupposes there is an objectively better font available, which assumes we can define "better" and agree on it, neither of which has been established.

It works fine. Can't complain too much about what isn't broken.