Fix links with invalid fragments

[chalin]

To see the current invalid links run:

$ ./scripts/double-check-refcache-4XX.mjs -f -m 0
Checking static/refcache.json for 4XX status URLs ...
Skipping 206: https://docs.spring.io/spring-framework/docs/current/reference/html/integration.html#jms (last seen 10/1/2025) INVALID FRAGMENT
Skipping 200: https://elixir.bootlin.com/linux/v6.8.7/source/include/uapi/linux/in.h#L28 (last seen 10/1/2025) INVALID FRAGMENT
...

Look for URLs reported to have "INVALID FRAGMENT". For each such URL:

• Find the correct URL, which might just need a fragment update, or maybe the entire URL
• Remove the invalid URL from the refcache.json file.

To test any external URLs with fragments for validity, run the command above with a non-zero argument to -m.

[vitorvasc]

@chalin is this issue still valid? I can help with it, just need to clarify what needs to be done and how to 😅

[svrnm]

@vitorvasc thanks! I think what needs to be done is go to the pages where the link is listed and verify if the fragment can be replaced with an alternative on the updated page or removed.

[chalin]

Yes it is still valid and ongoing. And help would be greatly appreciated. Let me refresh the links in the opening comment and explain a recommended process.

[vitorvasc]

As the changes from open-telemetry/semantic-conventions#2954 have already been released in #7863, I re-ran the checks using the ./scripts/double-check-refcache-4XX.mjs script. The output suggests we no longer have links with invalid fragments — but I'm not fully confident in the result, since it reports that only 28 URLs were processed.

❯ ./scripts/double-check-refcache-4XX.mjs -f -m 0
Checking static/refcache.json for 4XX status URLs ...
Updated 0 entries. Reach our max of 0, exiting.
Processed 28 URLs (1 with fragments, 0 are invalid)
Updated 0 entries.

Final HTTP status counts of entries:
  206 no frag: 3270
  200 no frag: 1510
  206 frag er: 570
  200 frag er: 158
  -30 no frag: 1
  200 frag ok: 3

Is that right?

[svrnm]

28 does not sound correct, yes 😆 @chalin would know best I guess