rays/ipx
1
0
Fork 0
Code Issues 1 Pull Requests Actions Packages Projects Releases Wiki Activity

Four subscribed feeds fail on every scan: their publishers' addresses are dead #92

New Issue
Closed
opened 2026-09-29 08:02:49 -07:00 by rays · 3 comments
rays commented 2026-09-29 08:02:49 -07:00
Owner
Copy Link

Found by the production check on 2026-09-29. Every scan since Loki starts (2026-09-29 12:52 UTC, so how long before that is unknown) logs a `feed_error` for these four, and none is ipx's fault: curl gets the same answer with ipx's user agent and a browser's.

feed url what the server says
gizmodo http://feeds.gawker.com/gizmodo/full 404. Gawker's feed host is long gone; Gizmodo's feed is now https://gizmodo.com/feed
toddstashwick-com-devil-inside http://toddstashwick.com/cms/css_rss-comic.asp?ID=1 301 to https://www.toddstashwick.com/, which 308s to itself forever. The feed address no longer exists
pelgrane-press-ltd https://www.pelgranepress.com/?feed=rss2 403, also to a browser user agent; likely a bot wall in front of the site
daily-quests http://daily-quests.com/comic/?feed=atom 503 on every try

Each wants unsubscribing (`ipx rm `) or pointing at a new address. daily-quests and pelgrane may come back; the other two will not.

Query:

sum by (feed, msg) (count_over_time({container="iPX"} |= "\"target\":\"ipx::io\"" | json | __error__="" | ev="feed_error" [$range]))
Found by the production check on 2026-09-29. Every scan since Loki starts (2026-09-29 12:52 UTC, so how long before that is unknown) logs a \`feed_error\` for these four, and none is ipx's fault: curl gets the same answer with ipx's user agent and a browser's. | feed | url | what the server says | |---|---|---| | gizmodo | http://feeds.gawker.com/gizmodo/full | 404. Gawker's feed host is long gone; Gizmodo's feed is now https://gizmodo.com/feed | | toddstashwick-com-devil-inside | http://toddstashwick.com/cms/css_rss-comic.asp?ID=1 | 301 to https://www.toddstashwick.com/, which 308s to itself forever. The feed address no longer exists | | pelgrane-press-ltd | https://www.pelgranepress.com/?feed=rss2 | 403, also to a browser user agent; likely a bot wall in front of the site | | daily-quests | http://daily-quests.com/comic/?feed=atom | 503 on every try | Each wants unsubscribing (\`ipx rm <id>\`) or pointing at a new address. daily-quests and pelgrane may come back; the other two will not. Query: ``` sum by (feed, msg) (count_over_time({container="iPX"} |= "\"target\":\"ipx::io\"" | json | __error__="" | ev="feed_error" [$range])) ```
rays added the bug label 2026-09-29 08:02:49 -07:00
rays commented 2026-09-29 08:28:14 -07:00
Author
Owner
Copy Link

pelgrane-press-ltd, looked into on 2026-09-29: not Cloudflare. The 403 comes from Automattic's CDN (server-timing: a8c-cdn; the site is a Jetpack-connected WordPress, WordPress.com site 113967104). Every path on pelgranepress.com, robots.txt included, answers 403 with a "Checking your browser..." page that makes the visitor run a JavaScript proof of work. It does this whatever the user agent, so no feed reader can get through; a browser does, because it runs the script.

There is no other address serving the same RSS: pelgranepress.wordpress.com does not exist. WordPress.com's public API does answer, with every post:

https://public-api.wordpress.com/rest/v1.1/sites/pelgranepress.com/posts/?number=20

but it is JSON, not RSS, and ipx only reads RSS and Atom.

Options: ask Pelgrane to let feed readers through to /feed/ (it is almost certainly a bot rule they set, or Automattic's on their behalf), unsubscribe, or have ipx read a WordPress.com site through that API (an enhancement, only worth it if more feeds end up behind the same wall). The Ken and Robin podcast, the site's audio, has its own feed that works (ken-and-robin-talk-about-stuff).

pelgrane-press-ltd, looked into on 2026-09-29: not Cloudflare. The 403 comes from Automattic's CDN (`server-timing: a8c-cdn`; the site is a Jetpack-connected WordPress, WordPress.com site 113967104). Every path on pelgranepress.com, robots.txt included, answers 403 with a "Checking your browser..." page that makes the visitor run a JavaScript proof of work. It does this whatever the user agent, so no feed reader can get through; a browser does, because it runs the script. There is no other address serving the same RSS: pelgranepress.wordpress.com does not exist. WordPress.com's public API does answer, with every post: https://public-api.wordpress.com/rest/v1.1/sites/pelgranepress.com/posts/?number=20 but it is JSON, not RSS, and ipx only reads RSS and Atom. Options: ask Pelgrane to let feed readers through to /feed/ (it is almost certainly a bot rule they set, or Automattic's on their behalf), unsubscribe, or have ipx read a WordPress.com site through that API (an enhancement, only worth it if more feeds end up behind the same wall). The Ken and Robin podcast, the site's audio, has its own feed that works (ken-and-robin-talk-about-stuff).
rays commented 2026-09-29 08:43:15 -07:00
Author
Owner
Copy Link

pelgrane-press-ltd, more precisely: the site is their own WordPress install, hosted on Pressable (199.16.172.36 and 199.16.173.48, RDAP: Pressable), which is Automattic's managed WordPress hosting. Its DNS is on Cloudflare, but DNS only; the traffic doesn't go through Cloudflare. The "Checking your browser..." wall (it sets an _hcc cookie) is Pressable's edge, so the site's own settings can't lift it; Pelgrane would have to ask Pressable to exempt /feed/.

pelgrane-press-ltd, more precisely: the site is their own WordPress install, hosted on Pressable (199.16.172.36 and 199.16.173.48, RDAP: Pressable), which is Automattic's managed WordPress hosting. Its DNS is on Cloudflare, but DNS only; the traffic doesn't go through Cloudflare. The "Checking your browser..." wall (it sets an `_hcc` cookie) is Pressable's edge, so the site's own settings can't lift it; Pelgrane would have to ask Pressable to exempt /feed/.
rays commented 2026-09-29 09:14:28 -07:00
Author
Owner
Copy Link

Left as they are, by choice: these are the publishers' problems, and since e9f2832 (#93) the sidebar shows each one with a mark on its artwork, what is wrong in words and, after a day, greyed-out artwork. Unsubscribing or finding a new address is up to whoever subscribed.

Left as they are, by choice: these are the publishers' problems, and since e9f2832 (#93) the sidebar shows each one with a mark on its artwork, what is wrong in words and, after a day, greyed-out artwork. Unsubscribing or finding a new address is up to whoever subscribed.
rays closed this issue 2026-09-29 09:14:28 -07:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
bug
Something is not working
 duplicate
This issue or pull request already exists
 enhancement
New feature
 help wanted
Need some help
 invalid
Something is wrong
 question
More information is needed
 wontfix
This won't be fixed
No Label bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
rays
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: rays/ipx#92
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?