Specs vs Implementations

The cluster debates the role, authority, bugs, and necessity of formal specifications compared to reference implementations in software development, particularly whether specs are infallible or superior to code.

➡️ Stable 0.7x Programming Languages
4,499
Comments
20
Years Active
5
Top Authors
#7296
Topic ID

Activity Over Time

2007
7
2008
20
2009
60
2010
100
2011
108
2012
130
2013
172
2014
213
2015
192
2016
234
2017
258
2018
221
2019
297
2020
296
2021
351
2022
456
2023
426
2024
365
2025
544
2026
49

Keywords

tilton.html e.g IETF crockford.com CPL TOML www.iso GPT ISO RFC spec specification code implementations specs implementation language implement specified bugs

Sample Comments

jstarks • Feb 19, 2021 • View on HN

Tell that to Annex K... Getting something into the spec isn't enough.

gearshift • Jun 29, 2019 • View on HN

Do you have docs that explain the spec so new implementations can be written without dissecting the code?

Kuinox • Dec 14, 2022 • View on HN

Specification are not immune to bugs.

eru • May 28, 2021 • View on HN

Why? The spec ain't God given.

tunesmith • Jul 31, 2025 • View on HN

Yeah this seems like the specification/implementation problem. One can perfectly implement a bad spec, but coming up with the perfect spec is a human problem.

HPsquared • Oct 28, 2025 • View on HN

Maybe the code itself is less important now, relative to the specification.

karmakaze • Mar 14, 2024 • View on HN

Still blame the ones defining the spec rather than implementing it.

foota • Feb 10, 2017 • View on HN

They have, but specs are a more authoritative standard than a reference implementation.

vbezhenar • Oct 18, 2017 • View on HN

Specs are not meant to be read by average developer. Average developer should read books. Specs are meant to clarify moot points between different implementations. Spec is like legal document, they are hard to read, but they serve their purpose.

ekianjo • Aug 23, 2025 • View on HN

We were talking about the spec, not the implementation.