Tap a claim on the ring to put it at the centre.
← A performance test without an explanation of its limiting factor…
17 connected korrents · 14 moments from 18 Oct 2007 to 8 Sept 2026.
Everything filed under benchmarks
benchmarks
Everything filed under documentation
documentation
Everything filed under compilers
compilers
Everything filed under mental models
mental models
Everything filed under measuring intelligence
measuring intelligence
Everything filed under management
management
Everything filed under demographics
demographics
Everything filed under software performance
software performance
Everything filed under developer tools
developer tools Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Read this korrent: A performance test without an explanation of its limiting factor should be treated as unanalyzed and possibly bogus. A performance test without an explanation of its limiting factor should be treated as unanalyzed and possibly bogus. Last stated a year ago
3 Aug 2025
BG
Brendan Gregg — holds since 2025-08-03 — tap for who they are
Same subject: Improvement whose limit is unknown is wrongly taken for unlimited improvement — tap to centre the map on it Improvement whose limit is unknown is wrongly taken for unlimited improvement
TM
Thomas Robert Malthus — holds since c. 1798 — tap for who they are
Same subject: That IQ tests are unbiased across groups is highly implausible and empirically unsupported. — tap to centre the map on it That IQ tests are unbiased across groups is highly implausible and empirically unsupported. Last stated 19 years ago
18 Oct 2007
CS
Cosma Rohilla Shalizi — holds since 2007-10-18 — tap for who they are
Same subject: Numbers offer a false precision: which metrics you choose to look at, and whether a five percent rise is good, are both art rather than science. — tap to centre the map on it Numbers offer a false precision: which metrics you choose to look at, and whether a five percent rise is good, are both art rather than science. Last stated a year ago
21 Sept 2025
JZ
Julie Zhuo — holds since 2025-09-21 — tap for who they are
Same subject: Running a model five times and keeping the best answer buys a higher benchmark score without buying a better model. — tap to centre the map on it Running a model five times and keeping the best answer buys a higher benchmark score without buying a better model. Last stated 4 months ago
26 Jun 2026
NB
Noam Brown — holds since 2026-06-26 — tap for who they are
Same subject: Ruling out every measurable cause because none explains all the variation, and putting an unmeasured one in their place, is not an explanation. — tap to centre the map on it Ruling out every measurable cause because none explains all the variation, and putting an unmeasured one in their place, is not an explanation. Last stated 3 months ago
20 Aug 2026
LS
Lyman Stone — holds since 2026-08-20 — tap for who they are
Same subject: A benchmark result should be reported under a stated budget, or as a curve against test-time compute — never as a single number. — tap to centre the map on it A benchmark result should be reported under a stated budget, or as a curve against test-time compute — never as a single number. Last stated 4 months ago
26 Jun 2026
NB
Noam Brown — holds since 2026-06-26 — tap for who they are
Same subject: The evidence from the largest software companies says programmers should take performance more seriously than they do, not less. — tap to centre the map on it The evidence from the largest software companies says programmers should take performance more seriously than they do, not less. Last stated 4 years ago
26 Apr 2023
CM
Casey Muratori — holds since 2023-04-26 — tap for who they are
Same subject: Performance test results are highly dependent on the specific machine, library versions, and other components used, making them hard to compare over time. — tap to centre the map on it Performance test results are highly dependent on the specific machine, library versions, and other components used, making them hard to compare over time. Last stated 3 months ago
14 Aug 2026
DS
Daniel Stenberg — holds since 2026-08-14 — tap for who they are
Same subject: A book as a physical object has no aura, but its text does. — tap to centre the map on it A book as a physical object has no aura, but its text does. Last stated 6 months ago
16 May 2026
VC
Vajra Chandrasekera — holds since 2026-05-16 — tap for who they are
Same subject: A company that deviates from open standards in its data exports without documenting the deviation creates an artificial vendor lock-in. — tap to centre the map on it A company that deviates from open standards in its data exports without documenting the deviation creates an artificial vendor lock-in. Last stated 2 years ago
3 Mar 2025
TS
Tobias Sjösten — holds since 2025-03-03 — tap for who they are
Same subject: A design proposal should describe the feature as though it already existed — once as a tutorial and once as reference documentation. — tap to centre the map on it A design proposal should describe the feature as though it already existed — once as a tutorial and once as reference documentation. Last stated 6 months ago
20 May 2026
AR
Alice Ryhl — holds since 2026-05-20 — tap for who they are
Same subject: Documentation should be written for AI coding agents to understand and modify software, not only for humans. — tap to centre the map on it Documentation should be written for AI coding agents to understand and modify software, not only for humans. Last stated 3 months ago
22 Jul 2026
SS
Salvatore Sanfilippo — holds since 2026-07-22 — tap for who they are
Same subject: Letting customers pin old runtime versions would undo part of the promise of serverless: that developers should not have to worry about updating their stack. — tap to centre the map on it Letting customers pin old runtime versions would undo part of the promise of serverless: that developers should not have to worry about updating their stack. Last stated 5 years ago
19 Oct 2021
KV
Kenton Varda — holds since 2021-10-19 — tap for who they are
Same subject: The harness, skills and context matter alongside the model; a missing doc can matter more than the model choice. — tap to centre the map on it The harness, skills and context matter alongside the model; a missing doc can matter more than the model choice. Last stated 2 months ago
8 Sept 2026
AK
Aman Khan — holds since 2026-09-08 — tap for who they are
Same subject: A compiler that also serves an IDE has to be lazy and deferred throughout — a different way of building compilers from the one the textbooks teach. — tap to centre the map on it A compiler that also serves an IDE has to be lazy and deferred throughout — a different way of building compilers from the one the textbooks teach. Last stated 6 months ago
13 May 2026
AH
Anders Hejlsberg — holds since 2026-05-13 — tap for who they are
Same subject: A compiler that names every place a change breaks turns refactoring into fixing errors until the compiler stops shouting. — tap to centre the map on it A compiler that names every place a change breaks turns refactoring into fixing errors until the compiler stops shouting. Last stated 6 months ago
20 May 2026
AR
Alice Ryhl — holds since 2026-05-20 — tap for who they are
Same subject: A programming language is not a compiler but an experience: editing, running, debugging and the runtime library all have to fit together as one cycle. — tap to centre the map on it A programming language is not a compiler but an experience: editing, running, debugging and the runtime library all have to fit together as one cycle. Last stated 6 months ago
13 May 2026
AH
Anders Hejlsberg — holds since 2026-05-13 — tap for who they are
At the centre
A performance test without an explanation of its limiting factor should be treated as unanalyzed and possibly bogus.
Holds Read this korrent →
Improvement whose limit is unknown is wrongly taken for unlimited improvement
Last stated c. 1798
Holds
That IQ tests are unbiased across groups is highly implausible and empirically unsupported.
Holds
Numbers offer a false precision: which metrics you choose to look at, and whether a five percent rise is good, are both art rather than science.
Holds
Running a model five times and keeping the best answer buys a higher benchmark score without buying a better model.
Holds
Ruling out every measurable cause because none explains all the variation, and putting an unmeasured one in their place, is not an explanation.
Holds
A benchmark result should be reported under a stated budget, or as a curve against test-time compute — never as a single number.
Holds
The evidence from the largest software companies says programmers should take performance more seriously than they do, not less.
Holds
Performance test results are highly dependent on the specific machine, library versions, and other components used, making them hard to compare over time.
Holds
A book as a physical object has no aura, but its text does.
Holds
A company that deviates from open standards in its data exports without documenting the deviation creates an artificial vendor lock-in.
Holds
A design proposal should describe the feature as though it already existed — once as a tutorial and once as reference documentation.
Holds
Documentation should be written for AI coding agents to understand and modify software, not only for humans.
Holds
Letting customers pin old runtime versions would undo part of the promise of serverless: that developers should not have to worry about updating their stack.
Holds
The harness, skills and context matter alongside the model; a missing doc can matter more than the model choice.
Holds
A compiler that also serves an IDE has to be lazy and deferred throughout — a different way of building compilers from the one the textbooks teach.
Holds
A compiler that names every place a change breaks turns refactoring into fixing errors until the compiler stops shouting.
Holds
A programming language is not a compiler but an experience: editing, running, debugging and the runtime library all have to fit together as one cycle.
Holds