Explanatory guidance, worked examples and data-source references for
implementers calculating an SCI for Web score. This site is a
non-normative companion to the SCI for Web Specification, which extends
the Software Carbon Intensity (SCI) specification,
ISO/IEC 21031:2024, to web
applications. The specification states requirements (shall / should /
may); these guidelines explain how to fulfil them. They never create,
relax or tighten a requirement.
This project is a draft work item of the Green Software Foundation's
Standards Working Group (SSWG). It is not yet officially endorsed, and its
content should be expected to change as the SSWG resolves the open
questions tracked in docs/OpenQuestions.md.
This project follows the structure and conventions of the GSF's SWI Guidance project, which occupies the same role for the Software Water Intensity (SWI) specification.
The content of this site is derived from two source documents, kept in
source/:
source/SCI-Web-Proposed-Specification.md: the proposed lightweight SCI for Web Specification. Its normative text (Clauses 1–11) is what every guideline page explains, and its decoupling map (§3) records where each part of the original draft now lives.source/SCI-Web-Original-Draft.md: the original SCI for Web draft (v0.2.0-draft, public snapshot of 2026-05-07). Explanatory material, examples and templates migrated from it into these guidelines, with requirements that the proposed specification does not carry converted into advisory wording.
Every page opens with an HTML comment naming the proposed specification
clause(s) it is grounded in and the original draft section(s) it migrated
from, so each passage can be traced back to its source.
docs/Traceability.md renders the full decoupling
map.
Content in this repository is licensed under CC-BY-4.0. Code (the site itself) is licensed under MIT.
No patent rights are asserted over the contents of this repository.
See LICENSE for the code licence (MIT). Documentation content is CC-BY-4.0 as noted above.
See CONTRIBUTING.md for the branching model and review process. All contributions are subject to the Code of Conduct.
All guideline content lives under docs/. Each folder corresponds to a
component of the SCI for Web formula or a supporting topic. Each page
carries front matter (sidebar_position, title) that controls its place
in the sidebar, which is otherwise generated automatically from the
docs/ folder structure, so there is no sidebars.js list to maintain by
hand when adding a page.
Every page follows the same conventions:
- a provenance comment immediately after the front matter;
- a bold In short: opening paragraph;
- no normative language of its own: where a page relies on a requirement, it attributes it to the specification ("Clause 7.5 requires…");
- every number either cited (source and access date) or labelled as an illustrative placeholder;
- a feedback link at the foot of the page.
Run node scripts/check-pages.js to check these conventions before
opening a pull request.
yarn install
yarn startThis starts a local development server and opens a browser window. Most changes are reflected live without restarting the server.
yarn buildThis generates static content into the build directory, which can be
served by any static hosting service.
The site is built with Docusaurus 2. Configuration
lives in docusaurus.config.js (site metadata, navbar, footer) and
sidebars.js (autogenerated sidebar). The draft-state banner is added by a
swizzled src/theme/Navbar/index.js.
Three branches are used, following the SWI Guidance model:
dev: the working branch; all pull requests targetdev.main: the reviewed, official GSF version of the content.web: the branch that is actually deployed. Netlify is connected to this branch and builds and publishes it on every push, using the settings innetlify.toml(yarn build, publishing thebuilddirectory).