Sometimes giving syndication feed readers good errors is a mistake
5 points by runxiyu
5 points by runxiyu
Feedly is genuinely the worst-behaving cloud feed reader I've ever encountered. Almost all cloud readers spoof their UA (e.g. inoreader does), but with Feedly i've seen huge pools of IPs used, no respect for rate limits, fetching the same feed twice concurrently. I'm not surprised they ignore the semantics of 302.
I too, see Feedly hit my blog with multiple requests, but I think I have it figured out why: some subscribed to the feed via http:, some via https: and some via a typo ("//index.atom" instead of "/index.atom"). Even though now I do a permanent redirect for http: to https: and "//index.atom" to "/index.atom", Feedly hasn't exactly gotten the message it can make one request for "/index.atom". I have not noticed it spoofing its UA, nor using a huge pool of IPs (just a /24). At least it does support conditional requests (once I set up my web server properly).
I am a Feedly user and moderately happy with the service, but somehow I am not surprised they pull stunts like this.