Roku Direct Publisher

[toddaa]

So I am attempting to get a channel going by making a json feed.  Everything validates, but the channel feed status claims only shows a vague message of "The feed format is invalid. We are unable to parse it".  The funny thing is that I'm using a feed supplied by Roku.  There's a topic called "Direct Publisher Channel Showcase" which holds the link:  https://api.myjson.com/bins/4tu28  I dumped the contents of that file on my server and the Direct Publisher cant parse it??

It seems that the Direct Publisher will not work with HTTPS urls.  Using one returned the "Unable to download from feedUrl." message.  I even tried the test url (https://api.myjson.com/bins/4tu28) and got the message.  

Fortunately, after switching the test url to just HTTP (http://api.myjson.com/bins/4tu28) it did ingest the feed.  Not sure what else would be different on my server.  CORS is enabled.  MIME is correct for a JSON file.  Sadly the error messages are too vague to be able to properly diagnose.

Can anyone verify any of this?

It seems there are many similar issues in this forum requesting this very same thing?  Has anyone been able to get to the bottom of this?

[toddaa]

Further diagnostics revealed that a ModSecurity rule was preventing the Direct Publisher from getting a response from my server.  The rule was defined as:

SecRule HTTP_User-Agent "Mozilla/(4|5)\.0$" "id:2000064"

Looking at my logs, Mozilla/5 is fairly normal for a user-agent, but a browser puts a whole ton of extra stuff to further define who it is.  Not knowing what Roku has checking for content on their back end made this hard to duplicate/diagnose.  There should be some some extra UA text added to their feed checker to define it to all the different servers out there it is requesting feeds from.  For now, disabling the rule has fixed my issue, but it is unknown how this opens up my server security wise.

[stubbstarbuck]

We serve our feed at Github pages (using the gh-pages branch). It works just fine. Just make sure you provide a robots.txt file with disallow or you will make your feed content publicly searchable.