Comment 52W0Q Re: Problem with Pipedot

Story

Centrelink Releases Welfare Receipient's Personal Information to try to save face

Preview

Problem with Pipedot "Feed" page (Score: 1)

by carguy@pipedot.org on 2020-04-25 14:02 (#52N97)

Hello,

I've enjoyed using your "Feed" page for years now. The other day, XKCD stopped updating, also the "XKCD" header above the list disappeared. Tried deleting that feed and using "+" to start over, but it's not working.
* There used to be a big pick list to the right of: "Use existing feed:" and that is currently empty.
* "Or add new feed:" didn't work when I tried https://xkcd.com/rss.xml

Maybe it's just a reboot or something easy?

Thanks!!

Re: Problem with Pipedot (Score: 1)

by bryan@pipedot.org on 2020-04-30 11:47 (#52W0Q)

Ya, I have a few bugs that cause the feed pages to act improperly at times.

The first is that I'm using a strict XML library to parse the feed. If the RSS/Atom page has any syntax error (for example: a stray tag or quote) the page wont parse correctly and will return blanks for everything. Pipedot's feed page uses the "title" tag returned from the RSS page as the name of the feed. This is why you sometimes see the header bar of a feed become blank. This is also why some feeds (like ExtremeTech) have weird names: because that's how their source title tag of their RSS page is set!

Another bug happens when adding dual protocol feeds. When I wrote the feed logic, I made each feed be unique based off the RSS URL. This lets a site have multiple feeds, as long as the RSS links have different URLs. The bug happens when you create two feeds to the same location, one using "http" and one using "https". A human knows that "http://xkcd.com/atom.xml" and "https://xkcd.com/atom.xml" are the same, but the site see's them as two unique URLs! (current xkcd feed link is using the atom version over http)

P.S. I also recently upgraded to Ubuntu 20.04 that was released last week. (may have some lingering issues)

Junk Status

Not marked as junk