Straight from today’s Quotle — here’s who really said it, and the story behind it. Play again

Who really said it

There are only two hard things in Computer Science: cache invalidation and naming things

Phil Karlton — attested, never documented Phil Karlton 1947–1997 · engineer at Xerox PARC, DEC, Silicon Graphics and Netscape
Attributed Rights unconfirmed

The credit is almost certainly right and it rests entirely on people’s word. Phil Karlton was a Netscape engineer who died in a car accident in Italy on 20 June 1997. His son, David Karlton, wrote in 2017 that his father ‘touted’ the phrase during his Netscape years. Colleagues remember hearing it from him. But there is no email, no memo, no slide, no mailing-list post and no recording — and the earliest instances anyone has located online are from December 2005, eight years after his death. This one cannot be closed, and it is more interesting for that.

Last verified 21 Jul 2026 · how we verify · spot an error?

For presentations

Put it on a slide

Unconfirmed quote & credit

“There are only two hard things in Computer Science: cache invalidation and naming things” Attributed to Phil Karlton (1947–1997) of Netscape, on his son David's attestation and his colleagues' memories. The earliest written instances located date to December 2005; nothing survives in Karlton's own hand.

Can you use it?

We couldn’t verify where this wording comes from, so there’s no rights status to give. Safe to present as an unverified or anonymous line — just don’t assert a specific author or source.

Image direction — grounded in this quote’s real context; paste into your image tool

In context

Evoking the world of Phil Karlton — engineer at Xerox PARC, DEC, Silicon Graphics and Netscape. Theme: "There are only two hard things in Computer Science: cache invalidation and naming things". Atmospheric and restrained, cinematic directional light, a muted period-appropriate palette, fine grain and texture. Compose with generous empty space for the quote and a small credit line. No lettering in the image itself.

Minimalist

Minimalist editorial slide for the quote "There are only two hard things in Computer Science: cache invalidation and naming things". Clean background, a single muted accent color, generous negative space, subtle paper or linen texture, no literal illustration — let the words carry it. Reserve clear space for a serif quote and a small credit line. No lettering in the image itself.

Provenance

The source

Best record
His son’s attestation, 2017
Said at
Netscape, mid-1990s — heard by colleagues
Earliest located online
December 2005
Documentary proof
None — nothing in his own hand

There are only two hard things in Computer Science: cache invalidation and naming things. Netscape, mid-1990s

The canonical wording, which is also the wording of the earliest 2005 instances. There is no manuscript to check it against, so 'canonical' here means 'the form the industry settled on' rather than 'the form he used' — and the widespread joke variant, which adds off-by-one errors as a third item after listing two, is a later riff by other people.

David Karlton — 'Naming things is hard'

How we traced it

  • David Karlton is the anchor. Writing on 4 December 2017, he opens: ‘Years ago, when my dad Phil Karlton was working at Netscape, he touted a now-infamous phrase.’ Note what he does and does not claim — that his father repeated it, not that his father is provably the first person ever to say it his post ↗.
  • The earliest instances located online are from December 2005 — a post by Tim Bray, and one by Kellan Elliott-McCrea dated 23 December 2005 — both already crediting Karlton by name, and neither citing anything but memory. Bray has said he first heard the line around 1996 or 1997 the 2005 post ↗.
  • Martin Fowler, who collects this sort of thing carefully, wrote it up as a favourite saying for which he could not find a satisfactory URL — a neat description of the whole problem Fowler’s note ↗.
  • The dates leave a gap nobody has filled. Karlton was at Netscape from the mid-1990s and died in June 1997; the first written trace appears in 2005. Everything in between is oral transmission among people who worked with him.
  • One further claim we could not check: a commenter on David Karlton’s post says he heard Phil say it at Carnegie Mellon around 1970, which would push it back a quarter of a century. It is a single unverified recollection, and we record it as that rather than as a finding.

We give no rights status because there is no fixed, authored text to attach one to — the line exists as a spoken aphorism transmitted by other people, and a single sentence of this length would in most jurisdictions fall below the threshold of copyright protection anyway. Phil Karlton died in 1997, so anything he did write is protected in the UK and EU until 2067, and David Karlton’s 2017 blog post is his own. Quote the aphorism freely with the credit attached; quote the blog posts about it with attribution. This is not legal advice.

Fact-check

How good is the evidence?

Nobody seriously proposes a rival author, and the industry has credited Karlton consistently for twenty years. The honest problem is narrower: nothing written by him has ever been produced, and the first written trace postdates his death by eight years.

His son says so closest to primary

David Karlton’s 2017 post states plainly that his father used the phrase at Netscape. A family member with no incentive to invent it is good evidence — though he is attesting to a memory, not producing a document the post ↗.

The credit is contemporaneous with the spread no anonymous phase

Fabricated attributions usually appear late, after a saying is already famous. This one carries Karlton’s name in the earliest instances anyone can find, from people who worked in his orbit. That is the opposite pattern.

The eight-year gap no written trace

He died in June 1997. The earliest located written instances are from December 2005. Nobody has produced an email, a mailing-list post, a code comment or a slide, in an era that generated all four in quantity.

Nobody claims exactness reconstructed

The standard form may be a tidied version of something looser. Without a text there is no way to test it, and the people repeating it are honest about repeating it from memory.

'…and off-by-one errors' later joke

The famous variant that lists three things after promising two is a riff added by other people. It is a good joke and it is not Karlton’s.

Carnegie Mellon, unverified single recollection

A commenter on the family blog recalls hearing Phil say it at CMU around 1970. If true it moves the date back twenty-five years. We could not corroborate it and are not treating it as established.

Phil Karlton, on the testimony of his son and his colleagues — and on nothing else. No document in his hand has ever been produced, and the earliest written instances located date to December 2005, eight years after he died Fowler on the missing source ↗.

Context

Why it resonates

It is funny because it is a list of two things that sound trivial and are not. Caching is the standard way to make software fast, and knowing when a cached copy has gone stale is a problem with no general solution. Naming looks like typing and is really the act of deciding what a thing is.

There is also a joke inside the joke that programmers rarely say out loud: naming things is hard because the second problem is the first one in disguise. A name is a cache of your understanding, and it goes stale the moment the code changes underneath it.

A name is a cache of your understanding. It goes stale the moment the code moves.

The timeline, and the hole in it

1947: Philip Lewis Karlton born in Los Angeles, 28 February.

Career at Xerox PARC, Digital and Silicon Graphics; principal work on the X11 sample server.

Mid-1990s: at Netscape, working on SSL, listing himself as Principal Curmudgeon. This is when colleagues remember hearing the line.

20 June 1997: Phil and Jan Karlton are killed in a car accident in Italy. He was fifty.

December 2005: the earliest written instances anyone has located appear online, already crediting him.

4 December 2017: his son David publishes the attestation that is now the standard citation.

The gap from the mid-1990s to 2005 has never been filled by a document, and at this distance it probably never will be.

Explore

Filed under

More provenance-checked quotes on the same theme, each with its reuse status.

Knowledge Work

Go further

Dig deeper

Trusted places to keep pulling the thread — primary archives, fact-checks, and scholarship. We show how we used each one before you leave.

Naming things is hard — David Karltonkarlton.org

The son’s attestation, 4 December 2017 — the closest thing to a primary source that exists.
Two Hard Things — Martin Fowlermartinfowler.com

Fowler on the missing URL, Tim Bray’s account, and the variants the saying has spawned.
The saying online, 23 December 2005laughingmeme.org

One of the earliest located written instances, already crediting Karlton by name.
In Memory of Phil and Jan Karltonkarlton.hamilton.com

The memorial site put up by colleagues after the 1997 accident — the man behind the aphorism.

The credited author

About Phil Karlton

Phil Karlton

1947–1997 · engineer at Xerox PARC, DEC, Silicon Graphics and Netscape

Phil Karlton (1947–1997) spent his career on infrastructure other people built on: he was a principal contributor to the X11 sample server that commercial X implementations were based on, and at Netscape he worked on SSL — the encryption that became HTTPS. His own home page there listed his title as ‘Principal Curmudgeon’, which tells you roughly everything about the register the quotation is in. He and his wife Jan were killed in a car accident in Italy on 20 June 1997; he was fifty. He is now known to millions of programmers who could not name a single thing he built.

For writers & researchers

Cite this

Primary source — Chicago

Karlton, Phil. Spoken aphorism, Netscape, mid-1990s. Attested by David Karlton, ‘Naming things is hard’, karlton.org, 4 December 2017. Earliest located written instances December 2005. No document in Karlton’s own hand exists.

MLA (9th ed.)

Karlton, Phil. Spoken aphorism, Netscape, mid-1990s. Attested in David Karlton, “Naming things is hard,” karlton.org, 4 Dec. 2017.

APA (7th ed.)

Karlton, P. (n.d.). Spoken aphorism [Netscape, mid-1990s]. Attested in D. Karlton (2017, December 4), Naming things is hard.

This verification page

Quotle.info. “Who really said ‘There are only two hard things in Computer Science’?” Last verified 21 Jul 2026. https://quotle.info/who-said/there-are-only-two-hard-things-in-computer-science

Copied