Research and evidence standards
The claim should never outrun the evidence.
This is the public standard for research, benchmarks, specifications, and product evidence published by David Hurley. Publication on this site is not itself peer review, independent validation, or a customer result.
Name the artifact honestly
Working papers, research reports, benchmark reports, specifications, proposals, position papers, and framework papers are labeled by what they are. A work is described as peer-reviewed only when a review venue or equivalent public evidence can be cited.
Scope every quantitative claim
Measurements identify the dataset or workload, observation date, denominator, and important exclusions. Estimates are described as estimates. Reference or synthetic results are not presented as customer outcomes.
Make reproduction possible
When code, fixtures, or raw results can be published, the paper links to them. When they cannot, the limitation is disclosed rather than hidden behind a broad reproducibility claim.
Publish limitations and corrections
A result states what it does not establish. Material corrections are made in the public artifact and its surrounding copy rather than silently preserved in machine-readable summaries.
Current collection
The public library currently contains ten open-access works: four research reports, one benchmark report, one working paper, one protocol specification, one proposal, one position paper, and one framework paper. Their subject labels are descriptive; they do not imply indexing by arXiv or another repository.
Review the collection →