This is a very niche post, which I write simply to help others who fall in the niche with me; the rest of you can move right along.

As I’ve mentioned before in passing, my RSS reader of choice these days is Reader, from the team that brought us Readwise (itself an estimable “highlight and catalog what you read” app in its own right).

Reader reads RSS feeds, but it’s also good, in an Instapaper-like way, of parsing HTML: you can use its browser extension to save an article you’re reading, for example, and it will show up in your Reader queue, in a nice distraction-and-advertising-free readable format.

There’s one aspect of Reader that’s mildly annoying: if an RSS feed is reported by another user as having a parsing issue, it can get marked on a system-wide level as “don’t use the RSS feed, but instead try scraping the HTML from the website.” In theory that’s a great and helpful fallback move, but, in practice, in can result a perplexingly presented feed.

Kottke.org is a feed of particular annoyance: if the site’s RSS feed is fed to Reader, then posts appear like this:

The body of the post is missing, and all that’s displayed is verbiage from the post footer about commenting.

I went back and forth with the (excellent) Reader user support desk a few times and that’s how I learned about the “helpful fallback move” that was causing it. Alas, as, again, this flag is set on a system level, it’s outside my control.

I found a fix, though: it seems like Reader tracks this flag using an exact URL match for the feed, so if I attach a random parameter to the RSS feed address, like this:

http://www.kottke.org/index.xml?feed

then this appears to Reader as a different RSS feed, and the workaround isn’t kicked in, so posts appear as you want them to:

I imagine that this kind of thing will be sorted by Reader at some point; in the meantime, this hack is useful.