Fix failing "Check markdown files for broken links" CI job

[Copilot]

The broken links CI job was failing due to four distinct categories: internal Jekyll routes failing as absolute filesystem paths, bot-blocking sites returning 403/429/999, a socket hang-up on hydro-surv.com, and genuinely removed external pages (404s).

Config changes (markdown.links.config.json)

• Added aliveStatusCodes: [200, 206, 301, 302, 303, 304, 403, 429, 999] — treats bot-blocking responses from sites like LinkedIn, e27.co, economist.com as alive
• Added ignorePatterns for permanently gone/unreachable URLs: hydro-surv.com (socket hang-up), wired.co.uk/magazine/archive/ (old archive structure), podcasts.apple.com (removed podcast), subnero.com/solutions/swan.html (page removed from site), github.com/org-arl/unetsockets/blob/ (notebook deleted from repo)

Workflow change (broken-links.yml)

Added a jq step before the checker that augments the config with replacementPatterns, mapping root-relative Jekyll routes to local filesystem paths so the checker can verify them without building the site:

- name: Generate link check config with internal link replacement patterns
  run: |
    jq --arg ws "$GITHUB_WORKSPACE" '. + {
        "replacementPatterns": [
          {"pattern": "^(/[^#?]+/)$",   "replacement": ("file://" + $ws + "$1")},
          {"pattern": "^(/[^#?.]+)$",   "replacement": ("file://" + $ws + "$1.md")}
        ]
      }' \
      .github/workflows/markdown.links.config.json > "$RUNNER_TEMP/mlc.config.json"

Blog post fixes

• 2024-03-01-Product-Launch-S40H-series-modems.md: Corrected internal link /products/wnc/gen4x/accessories/multi-receiver.html → multi-receiver (file is .md; .html never existed on the filesystem)
• 2018-09-28-UnetStack-Newsletter.md: Replaced deleted Jupyter notebook link with the Python package README
• 2021-04-23-New-SwanViz-UI-developed-together-with-CSSSR.md: Removed dead anchor to the removed SWAN solutions page
• 2018-07-17-SWANs-monitor-reservoirs-water-quality.md: Updated source front-matter field from a 404 Straits Times cartoons URL to the base domain

[Copilot]

Pull request overview

This PR addresses failures in the “Check markdown files for broken links” workflow by updating the markdown-link-check configuration, adjusting the workflow to generate a derived config at runtime, and fixing/removing several problematic links in blog posts.

Changes:

• Expanded markdown-link-check config to tolerate bot-blocking status codes and ignore known-unreachable URLs.
• Updated the broken-links GitHub Actions workflow to generate a temporary config that includes internal-link replacement patterns.
• Edited several posts to remove or update dead/invalid links and a front-matter source URL.

Reviewed changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
.github/workflows/markdown.links.config.json	Adjusts link-check behavior via ignore patterns and accepted “alive” HTTP status codes.
.github/workflows/broken-links.yml	Generates a derived link-check config (via jq) and points the action at it.
_posts/2024-03-01-Product-Launch-S40H-series-modems.md	Adjusts an internal product link to avoid link-check failure.
_posts/2021-04-23-New-SwanViz-UI-developed-together-with-CSSSR.md	Removes a dead external link to a removed page.
_posts/2018-09-28-UnetStack-Newsletter.md	Replaces a deleted notebook link with a stable README link.
_posts/2018-07-17-SWANs-monitor-reservoirs-water-quality.md	Updates front-matter source URL away from a 404.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[manuignatius]

Didn't merge as some of these changes will go away in the next iteration of the website, so I will apply on top of that.